GALVANIZED STEEL REINFORCEMENT IN CONCRETE
Elsevier Internet Homepage- http://www.elsevier.com Consult the Elsevier homepage for full catalogue information on all books, journals and electronic products and services including further information about the publications listed below. Elsevier titles of related interest Books SOMIYA et al. Handbook of Advanced Ceramics 2003. ISBN: 0 12 654640 1 ELSSNER et al. Ceramics and Ceramic Composites: Materialographic Preparation ISBN: 0 444 10030 X Elsevier author discount Elsevier authors (of books and journal papers) are entitled to a 30% discount off the above books and most others. See ordering instructions below. Journals Sample copies of all Elsevier journals can be viewed online for FREE at www.sciencedirect.com, by visiting the journal homepage. Cement and Concrete Composites Cement and Concrete Research Ceramics International Journal of the European Ceramics Society To contact the Publisher: Elsevier welcomes enquiries concerning publishing proposals: books, journal special issues, conference proceedings, etc. All formats and media can be considered. Should you have a publishing proposal you wish to discuss, please contact, without obligation, the publisher responsible for Elsevier’s material science programme: Emma Hurst Publishing Editor Elsevier Science Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB, UK
Phone: +44 1865 843629 Fax: +44 1865 843920 E.mail:
[email protected]
General enquiries, including placing orders, should be directed to Elsevier’s Regional Sales Offices - please access the Elsevier homepage for full contact details www.elsevier.com
GALVANIZED STEEL REINFORCEMENT IN CONCRETE E DITED BY
S TEPHEN R. Y EOMANS U NIVERSITY OF N EW S OUTH W ALES , C ANBERRA , ACT, A USTRALIA
Prepared for the International Lead Zinc Research Organization, Inc. Research Triangle Park, Raleigh Durham North Carolina, USA
2004
Amsterdam – Boston – Heidelberg – London – New York – Oxford Paris – San Diego – San Francisco – Singapore – Sydney – Tokyo
ELSEVIER B.V. Radarweg 29 P.O. Box 211, 1000 AE Amsterdam, The Netherlands
ELSEVIER Inc. 525 B Street, Suite 1900 San Diego, CA 92101-4495 USA
ELSEVIER Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB UK
ELSEVIER Ltd 84 Theobalds Road London WC1X 8RR UK
q 2004 Elsevier Ltd., All rights reserved. This work is protected under copyright by Elsevier Ltd., and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit educational classroom use. Permissions may be sought directly from Elsevier’s Rights Department in Oxford, UK: phone (þ44) 1865 843830, fax (þ44) 1865 853333, e-mail:
[email protected]. Requests may also be completed on-line via the Elsevier home-page (http://www.elsevier.com/locate/ permissions). In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (þ1) (978) 7508400, fax: (þ1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (þ44) 20 7631 5555; fax: (þ44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier’s Rights Department, at the fax and e-mail addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.
First edition 2004 Library of Congress Cataloging in Publication Data A catalog record is available from the Library of Congress. British Library Cataloguing in Publication Data A catalogue record is available from the British Library.
ISBN: 0-08-044511-X 1 The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). W Printed in The Netherlands.
Contents
Preface Contributing Authors’ Biographical Notes
xi xiii
Chapter 1. Galvanized Steel in Concrete: An Overview Stephen R. Yeomans 1.1. Introduction 1.2. Corrosion Prevention in Reinforced Concrete 1.3. Coated Steel Reinforcement — General Principles 1.4. Galvanized Reinforcing Steel 1.4.1. Behaviour of Zinc in Concrete 1.4.2. Chloride Tolerance 1.4.3. Behaviour of the Coating and Corrosion Products 1.4.4. Experience with Galvanized Steel in Concrete 1.4.5. Economics of Galvanized Reinforcement 1.5. Corrosion Model for Galvanized Steel in Concrete 1.5.1. Initiation Stage 1.5.2. Protection Stage 1.5.3. Propagation Stage 1.5.4. Comparison of Models 1.6. Applications of Galvanized Reinforcement 1.7. Summary References
1 1 4 5 6 8 8 9 11 14 15 16 16 17 17 18 26 27
Chapter 2. Design for Durability with Galvanized Reinforcement R. Narayan Swamy 2.1. Introduction 2.2. Current Design Specifications 2.3. The Ubiquitous Nature of Exposure 2.3.1. Environmental Considerations 2.3.2. Geomorphological Considerations 2.3.3. Effects on Materials and Concrete 2.4. The Need for Protection of Steel in Concrete
31 31 33 34 34 36 36 36
vi
Contents 2.4.1. The Nature of Concrete 2.4.2. Deterioration of RC 2.5. Design Strategy for Structural Integrity 2.6. Design Implications of Galvanized Reinforcement 2.6.1. Galvanizing and Galvanized Bars 2.6.2. Cracking of Galvanized Coating 2.6.3. Mechanical Properties 2.6.4. The Galvanized Bar – Concrete Bond 2.7. Durability of Concrete with Galvanized Bars 2.7.1. Resistance to Carbonation 2.7.2. Resistance to Chloride Attack 2.7.3. Long-Term Field Tests 2.7.4. Engineering Implications 2.8. In Situ Performance of Galvanized Reinforcement 2.9. Concluding Remarks References
37 38 41 42 42 43 45 46 48 49 51 55 58 60 63 65
Chapter 3. Corrosion of Metals in Concrete Alberto A. Sagu¨e´s 3.1. The Steel Environment in Concrete 3.2. The Initiation – Propagation Model for Corrosion of Steel 3.2.1. Initiation Stage: The Chloride Concentration Threshold 3.2.2. Initiation Stage: Combined Factors 3.2.3. Corrosion Propagation Stage 3.3. Integrated Corrosion Forecasting 3.4. Strategies to Improve Durability 3.4.1. Concrete Quality and Cover Thickness 3.4.2. Concrete Surface Treatments and Chloride Barriers 3.4.3. Corrosion Inhibitors 3.4.4. Alternative Reinforcement Materials References
71 71 72 73 74 75 76 78 78 79 79 80 83
Chapter 4. Zinc Materials for Use in Concrete Thomas J. Langill, Barry Dugan 4.1. Introduction 4.2. The Hot-Dip Galvanizing Process 4.2.1. Surface Preparation 4.2.2. Fluxing 4.2.3. Galvanizing 4.3. Metallurgy of the Galvanized Coating and Alloy Structure 4.3.1. Coating Structure 4.3.2. Coating Appearance 4.4. Fabrication of Galvanized Products
87 87 91 91 92 92 93 93 95 98
Contents
vii
4.4.1. Bending 4.4.2. Abrasion and Impact Resistance of Galvanized Coatings 4.4.3. Corner and Edge Protection 4.4.4. Welding of Galvanized Steel 4.4.5. Repair of Galvanized Coatings 4.5. Zinc Inhibitors 4.6. Codes of Practice and Standards 4.7. Field Handling Techniques References
99 101 101 101 103 104 105 107 108
Chapter 5. Electrochemical Aspects of Galvanized Reinforcement Corrosion Carmen Andrade, Cruz Alonso 5.1. Introduction 5.2. Galvanized Reinforcement as a Protection Method 5.2.1. Corrosion of Galvanized Steel in Alkaline Media 5.2.2. Evolution of Hydrogen During Cement Setting 5.2.3. Influence of Alkali Content on Passivation of Galvanized Steel 5.2.4. Influence of Galvanized Coating in the Passivation Process 5.3. Behavior of Galvanized Steel in Aggressive Media 5.3.1. Corrosion of Galvanized Steel in Carbonated Concrete 5.3.2. Corrosion of Galvanized Steel in the Presence of Chlorides 5.4. Conclusions References
111 111 111 111 118 120 123 128 128 132 141 141
Chapter 6. Laboratory and Field Performance of Galvanized Steel in Concrete Stephen R. Yeomans 6.1. Introduction 6.2. Some Early Research 6.3. The 1970s 6.4. The 1980s 6.5. The 1990s 6.6. 2000 and Beyond 6.7. Summary References
145 145 147 149 153 169 183 189 192
Chapter 7. The Bermuda Experience: Leading the Way on Galvanized Reinforcement Neil D. Allan 7.1. Introduction 7.2. The Islands of Bermuda 7.3. Previous Investigations 7.3.1. Sponsored Research — 1970s 7.3.2. Bermuda MW&E Surveys — 1991 7.4. Longbird Bridge Revisited — 1995
199 199 199 201 201 203 206
viii
Contents 7.5. Hamilton — Old Bus Depot 7.6. Performance of Black Steel in Bermuda 7.6.1. Sargasso Fish Processing Facility 7.6.2. Ordinance Island Bridge 7.6.3. Commentary 7.7. The Watford Bridge — A Case Study 7.7.1. The Old Watford Bridge 7.7.2. The New Watford Bridge 7.8. Economic Case for Galvanized Reinforcement 7.9. Practical Considerations 7.9.1. Alternate Coatings 7.9.2. Special Practices 7.9.3. Damage to Coating 7.9.4. Coating Mass 7.9.5. Bond Strength 7.10. Continued Use of Galvanized Reinforcement in Bermuda 7.11. Conclusions References
207 207 208 208 209 209 210 212 215 216 216 216 217 218 218 218 220 220
Appendix A. Evaluation of Galvanized Reinforcing Steel in the Longbird Bridge, Bermuda David Stark A.1. Introduction A.2. Scope of Work A.3. Results of the Investigation A.3.1. Petrographic Examination A.3.2. Chloride Contents A.3.3. Metallurgical Analysis A.4. Summary and Conclusions References
222 222 222 223 223 224 225 226 227
Chapter 8. Bond of Steel in Concrete and the Effect of Galvanizing Obada Kayali 8.1. The Significance of Bond 8.2. Failure in Bond 8.3. Evaluating Bond Strength 8.4. The Bond Stress-Slip Relationship 8.5. Evaluation of Bond in a Flexural Member 8.6. New Developments in Concrete 8.6.1. Higher Strength Concretes 8.6.2. Structural Lightweight Concrete 8.6.3. Fibre-Reinforced Concrete 8.6.4. The Role of Galvanized Reinforcement 8.7. Coated Reinforcement and Bond
229 229 235 238 241 243 246 246 247 248 249 249
Contents 8.7.1. Fusion Bonded Epoxy Coating 8.7.2. Hot-Dip Galvanizing 8.8. Effect of Rib Geometry and the Coating 8.9. Design Standards and Bond 8.10. Concluding Remarks References
ix 249 250 261 262 263 264
Chapter 9. Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective John P. Broomfield 9.1. A Brief History 9.2. The Corrosion of Steel in Concrete and Protection Systems 9.3. Design for Durability 9.4. The Performance of Galvanized Reinforcement 9.4.1. Durability in Carbonated Concrete 9.4.2. Durability in Chloride-Contaminated Concrete 9.4.3. Consequences of Installing Galvanized Reinforcement 9.5. Conclusions References
271 271 273 275 280 280 281 283 284 285
Colour Plates Index
287 293
This Page Intentionally Left Blank
Preface
Galvanizing is an established means of protecting steel reinforcement and other embedded components against corrosion in concrete. It has been quite widely used for some 70-80 years in many countries for a range of concrete construction in both mild-to-moderate and aggressive environments. After a period of quite restricted use during the late 1970s and early 1980s, the last decade in particular has seen a growing interest in and market demand for galvanized reinforcing steel. This has come about largely as a result of the publication over this period of a significant body of laboratory-based research and a growing database of results on the long-term field performance of galvanized reinforced concrete structures. A significant shift in thinking has also followed the relaxation in restrictions on the use of galvanized reinforcement in concrete highways and bridge decks in the USA especially. With the quantity of literature that had become available, a pressing need existed for the publication of an authoritative and critical assessment of the characteristics and use of galvanized steel in concrete. Apart from a number of small surveys and reviews interspersed through the technical literature over the last few years, the most comprehensive single document covering this topic was Galvanized Reinforcement for Concrete - II, published by the International Lead Zinc Research Organization in 1981. A thorough updating of this reference source was long overdue and the publication of this book is intended to bridge that gap. In the preparation of this book, I have indeed been fortunate to have as contributors a group of eminent international figures from academia, research and industry, each well known for their work in the concrete, corrosion and corrosion protection areas. They bring to the book their particular expertise and experience, all aimed at allowing the reader to make an informed technical decision on the merits of galvanizing for corrosion protection in concrete construction. Implicit in this is a high level of technical detail as well as a completeness and accuracy in reporting the facts. I have every confidence that the contributors have all achieved this aim and collectively presented a balanced and fair record. This book is aimed at a wide audience covering industry groups, suppliers, specifiers, architects and engineers associated with reinforced concrete
xii
Preface
construction, as well as students, teachers and researchers in the areas of civil and construction engineering, building science, materials and corrosion engineering. For each, varying levels of technical information and descriptive material have been presented, which I believe are sufficiently detailed to develop a deeper understanding of this diverse topic. It has taken several years to complete this book and my sincere thanks are due to all the contributors for their willing cooperation, their patience and the care and attention they have taken in the preparation of their chapters. I also gratefully acknowledge the continuing support of the International Lead Zinc Research Organization, in particular Dr Safaa Alhassan and Dr Frank Goodwin, without whose guidance and assistance it is unlikely that this project would have been completed. Sincere thanks are also due to friends and colleagues in the galvanizing and fabrication industries, both in Australia and elsewhere, who have provided technical advice and assistance for my work. In particular, I would like to mention my good friends Mr Tom Kinstler, Mr Mike Dennett, Mr John Robinson, Dr Geoff Fronhsdorff and the late Dr Jim Clifton, who have all, in different ways and at different times, helped to bring this book to fruition. Thanks also are due to the staff and graduate students of the School of Civil Engineering, University of New South Wales, both at the main campus in Sydney and at the Australian Defence Force Academy, with whom I have worked so closely for many years. Finally, but most importantly, my love and gratitude goes to Susan, Nicola and Julia for their unfailing support and encouragement. Dr Stephen R Yeomans Canberra, Australia June, 2004
Contributing Authors’ Biographical Notes
Mr Neil D Allan Neil Allan graduated with a BSc (Honours) from Glamorgan University and MBA from the Edinburgh Business School. He is a chartered civil engineer with over 15 years experience in the construction industry gained in the design and management of projects in UK, Africa and the Caribbean. Over the period 19871995, he was a Senior Engineer with the Ministry of Works and Engineering in Bermuda, where he managed many Government projects, including the New Civil Air Terminal and the award-winning Tynes Bay Waste-to-Energy Plant. He has published widely in the areas of management and engineering construction, including a number of papers on his experiences in Bermuda at international conferences, workshops and technical seminars and in the US Roads and Bridges Magazine. In 1997, he was awarded the Institution of Civil Engineers (UK) Gold Medal for the best paper on management in the construction industry. Neil worked at the University of Bristol over the period 1996-2002 and is presently Senior Lecturer and Director of Studies for Engineering Management, Bath University, UK. Ms Cruz Alonso Cruz Alonso is a graduate in Chemistry from the University of Zaragoza (1982) and holds a PhD in Chemistry from the University Complutense of Madrid (1986). She has received a number of Awards and Fellowships, including the National Development Plan for Junior Researchers at the Eduardo Torroja Institute in Madrid (1988), a British Council Scholarship at the Corrosion and Protection Centre, UMIST (1985-1966) and Research Fellowship at the Eduardo Torroja Institute in Madrid (1987). Most of her work has been devoted to studies of the corrosion and protection of steel in concrete and the microstructure and durability of concrete. She has published in excess of 50 papers in international and Spanish journals and chapters in internationally published books and has presented some 80 papers in local and international conferences.
xiv
Contributing Authors’ Biographical Notes
Professor Carmen Andrade Perdix Carmen Andrade holds a Doctorate in Industrial Chemistry and is a Research Professor at the Consejo Superior de Investigaciones Cientificas, Madrid, Spain. She is also Director of the Institute of Construction Sciences Eduardo Torroja. Carmen is known internationally for her work in corrosion science, where her domain of expertise is the corrosion of steel in concrete and the mechanisms of degradation of concrete. She has published extensively in these and related fields. Carmen holds a number of high-profile positions, including that of Vice President of RILEM, President of UEATC (The Eurpoean Union of Technical Agreement in Construction) and President of the Sectorial Commission “Construction” on ENAC (Entidad Nacional de Acreditacion de Espana). She is Vice-Chair of the European COST-509 Action, Corrosion and Protection of Metals in Concrete, and Chair of RILEM TC154-EMC, Electrochemical Techniques for Measuring Metallic Corrosion in Concrete. In 1986, Carmen was the recipient of the prestigious Robert L’Hermite Award and in 1996 was a RILEM Fellow. Dr John P Broomfield John Broomfield has a BSc from Sussex University in Chemical Physics and PhD from Linacre College, Oxford. He is a registered European Engineer, Fellow of the Institute of Corrosion and the Institute of Materials and a NACE International Corrosion Specialist. John has worked on the corrosion of steel in concrete for the past 20 years. He has authored over 80 publications, including a book on the subject. John is also active in Standards writing committees in the USA, UK and Europe. Since spending 3 years in Washington, DC as Technical Program Manager for the SHRP structures research project, John has run his own consultancy based in London, England. Recent projects include the National War Memorial in Wellington New Zealand, a jetty in Zanzibar, Tanzania and the Post Office Tower and Battersea Power Station in London. Mr Barry Dugan Barry Dugan has a BSc in Physics from Duquesne University in Pittsburgh, Pennsylvania, USA, along with numerous courses in metallurgy and corrosion engineering. He is currently Technical Service Manager for the Zinc Corporation of America, USA. He started in the zinc industry in 1978 working for St. Joe Minerals Corporation doing zinc and lead alloy development work. For the past two decades, he has been primarily involved in technical support to the
Contributing Authors’ Biographical Notes
xv
after-fabrication hot-dip galvanizing industry. He is a frequent speaker at conferences and seminars on galvanizing and related topics. Much of his work deals with the examination and evaluation of failures of galvanized products. He is also noted for his work on process improvements to the galvanizing industry. Dr Obada Kayali Obada Kayali holds a BE in Civil Engineering from the American University of Beirut and MSc and PhD from Strathclyde University. He is a Chartered Professional Engineer and Member of the Institution of Engineers Australia and is currently a Senior Lecturer in the School of Civil Engineering, University of New South Wales at the Australian Defence Force Academy, Canberra. He worked initially as a structural engineer in North Africa in the rehabilitation and design of railway and highway bridges. His postgraduate studies were in structural engineering and concrete durability. After obtaining his doctorate, Dr Kayali entered academic life and through this time has been regularly involved in consultancies to industry and the engineering profession in matters of reinforced concrete design, analysis and durability. Dr Kayali has over 50 published papers in international journals and conference proceedings. His current research interests cover a number of areas, including the design and analysis of reinforced concrete slabs, the production of high-strength and ultra-high-strength concrete, structural lightweight concrete, the behaviour of reinforced concrete exposed to fire and other adverse effects, the bond characteristics of steel reinforcement in concrete, high-volume fly-ash concrete and fibre-reinforced concrete. Dr Thomas J Langill Thomas Langill graduated with BS (Cum Laude) and MS in Physics from John Carroll University. He went on to complete a PhD in Materials Science and Engineering at Northwestern University in 1980. He has been with the American Galvanizers Association as its Technical Director for nearly 10 years. Over this time he has been active in shaping research programmes for galvanizing and has provided technical support to engineers and specifiers about the processing and use of galvanized steel. He has contributed widely to ASTM Committees that deal with the specifications used by the galvanizing industry. In his capacity as Technical Manager, he writes regular features for the AGA Magazine and has presented numerous technical papers at conferences and other meetings. He has also authored and presents a seminar series on Processing Details in the Hot-Dip Galvanizing Industry.
xvi
Contributing Authors’ Biographical Notes
Professor Alberto A Sagu¨e´s Alberto Sagu¨e´s is a Licentiate from the National University of Rosario and holds a PhD in Metallurgy from Case Western Reserve University. He is a Distinguished University Professor in the Department of Civil and Environmental Engineering at the University of South Florida and is a registered Professional Engineer in Florida. From 1997 to 2002, Professor Sagu¨e´s served by Presidential appointment as a member of the United States Nuclear Waste Technical Review Board. Previously, he has held positions at Columbia University, the Ju¨lich Nuclear Research Center, Argonne National Laboratory, and the Kentucky Center for Energy Research Laboratory. He has 20 years of teaching experience and more than 140 technical publications to his credit. His research interests are in corrosion of reinforcing steel in concrete, durability forecasting of civil infrastructure and engineering materials for energy applications. Professor R Narayan Swamy Narayan Swamy holds the degrees of BEng, MEng and PhD. He is a Chartered Engineer and also FICE, FIStructE, FIMechE, FASCE and FACI. He is Professor and member of the Structural Integrity Research Institute and the Centre for Cement and Concrete Research at the University of Sheffield. His research interests cover a wide range of cement and concrete-related topics, including cement materials, admixtures, microstructure and properties, durability, long-term exposure and repair and rehabilitation of structures. He has also undertaken significant research on the durability of steel in concrete and protective coatings for steel and concrete. He has over 200 refereed publications in national and international journals and conferences and has edited a number of books, including five in the Concrete Technology and Design Series, and authored multiple chapters in books. He is the founder and editor of the Journal of Cement and Concrete Composites and the International Journal of Cement Composites and Lightweight Concrete. Through his long career, Professor Swamy has been the recipient of a number of awards and prizes, including the ICE George Stephenson Gold Medal, the IStructE Henry Adams Diploma, the ACI/Concrete Research Council Robert E Phileo Award and several ACI/CANMET International Awards. Professor Stephen R Yeomans Stephen Yeomans holds a BSc (Honours I) and PhD in Metallurgy from the University of New South Wales. He is a Chartered Professional Engineer and
Contributing Authors’ Biographical Notes
xvii
Member of the Institution of Engineers, Australia. He joined the academic staff of the University of NSW in 1974 and is an Associate Professor presently located at the Australian Defence Force Academy in Canberra where he was Head, School of Civil Engineering from 1998 to 2003. He has also worked at the Alcan Research Laboratories in Canada, the University of Sheffield and the National Institute of Standards and Technology in the USA. He has more than 90 publications in journals, international and local conferences, seminars and workshops. He is active in several Standards Australia committees and has been a member of both ISO and ASTM Committees concerned with reinforcing materials and galvanizing. His principal research interests concern the corrosion of metals in concrete and, in particular, the use of coated reinforcement for corrosion protection. He has undertaken long-term research on these topics with major support from the International Lead Zinc Research Organisation, USA. He has lectured widely on these topics at seminars in Australia, Canada, the USA and throughout SE Asia. He has also presented his work at NACE and CANMET/ACI conferences and in other international forums.
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 1
Galvanized Steel in Concrete: An Overview Stephen R. Yeomans University of New South Wales, Canberra, Australia
1.1. Introduction Reinforced concrete is one of the most widely used modern materials of construction. It is comparatively cheap and readily available and has a range of attractive properties and characteristics that makes it suitable for a variety of building and construction applications. It is also used in a range of exposure conditions. The long-term performance of reinforced concrete is usually assessed against two main criteria, serviceability and durability. Serviceability relates to structural integrity, i.e. the ability of the element to sustain loads safely throughout its life and to perform its intended function. Durability, on the other hand, refers to the ability of the concrete to resist changes in its microstructure and properties, particularly where such changes may adversely affect the serviceability of the element. Perhaps the most obvious consequence of a lack of durability in reinforced concrete is the corrosion of the steel reinforcement, a topic that has been widely studied and reported [1 –7]. Steel embedded in concrete (or cement mortar) is normally protected from corrosion due to the presence of a passive film on the surface of the metal. This film forms in the highly alkaline environment of hydrated cement, with a pH in excess of about 13 and, as long as the passive state is maintained, the steel will not corrode. To ensure long-term corrosion protection to the steel, the concrete mass must be sufficiently impermeable so as to limit the transport of species such as water, chloride ions, oxygen, carbon dioxide and other gases through the concrete to the depth of the reinforcement. The presence of critical (or socalled threshold) levels of these species, which are usually carried into the concrete in solution in water, either change the nature of the concrete or alter the condition of the embedded steel. In either case, corrosion of the steel can then initiate.
2
S.R. Yeomans
For example, the presence of chloride ions above threshold concentrations has the effect of depassivating the steel, even if the pH of the adjacent concrete remains high. On the other hand, carbon dioxide and other gases that penetrate the concrete mass in aqueous solution react with the alkali-rich pore water and may lower the pH of concrete below the value at which the steel can remain passivated. This process, known as carbonation, specifically refers to the neutralization of the calcium hydroxide saturated pore solution by carbon dioxide dissolved in water (carbonic acid). In aqueous conditions, the reduction of oxygen at cathodic sites in the corrosion cell is an essential component of the corrosion process and so the rate of supply of oxygen is also important. Should corrosion of embedded steel in concrete occur, physical damage to the concrete mass is likely to follow. Steel corrosion products are quite voluminous (with an expansion factor of 2 – 10 times) and typically precipitate at the interface between the steel and the concrete. The swelling caused by this generates stresses of sufficient magnitude (about 3 – 4 MPa) to exceed the tensile capacity of the concrete and, as a result, the concrete cracks in tension. Such cracks usually run from the bar to the nearest adjacent surface, which may be the edge of a column or precast element or the surface of a slab or beam. Once cracking has occurred, unsightly rust staining of the surface is often observed and further swelling caused by the laying down of more corrosion product usually leads to delamination of the element or spalling of pieces of concrete from the surface. By this stage, the structure would be in a serious state of distress and remedial action would be necessary to extend its life. Some examples of corrosion-induced damage in reinforced concrete are shown in Fig. 1. Corrosion-induced damage to reinforced concrete often necessitates early repair and occasionally complete replacement of the structure or element well before its design life is reached. Worldwide, the costs associated with such remedial work are massive and are expected to increase in the future at an alarming rate. Estimates of these costs vary widely. For example, as early as 1977 in the USA, the cost of major repair and replacement of damaged bridge decks, which was rapidly becoming a systemic problem, was about US$ 23b [8]. In Australia at about the same time, the cost of building repairs was estimated at A$ 50m, increasing to A$ 200m by 1990 [9] while, in the mid-1980s in Hong Kong, the replacement cost of 26 public housing blocks less than 20 years old was HK$ 800m (as reported in the South China Morning Post, 1986). Broomfield [6] reported that, in the late 1980s, the UK Department of Transport estimated repair costs of some £616m for motorway bridges in England and Wales. Over the same period in the USA, annual costs associated with bridge deck repairs due to deicing salts alone was in the range US $50– 200m, plus an additional US $100m for substructures and US$ 50–100m on multi-level parking garages [10].
Galvanized Steel in Concrete: An Overview
3
Figure 1: Effects of reinforcement corrosion in concrete.
Clearly, such costs are an enormous drain on finances and resources that could otherwise be allocated to new construction. What has also become evident is that, while the repair of reinforced concrete may make good the surface deterioration of the structure, it usually does not address the root causes of the problem. The circumstances that lead to the initial onset of corrosion often survive in adjacent or more deeply buried regions and may reveal themselves at some time in the future.
4
S.R. Yeomans
1.2. Corrosion Prevention in Reinforced Concrete There can be little doubt that the most cost-effective way to minimize the risk of corrosion in reinforced concrete is to ensure that the cover to the reinforcement is of adequate thickness and that the concrete itself is dense and impermeable. The primary function of the concrete cover is to protect the reinforcing steel and it can only do so if the structure and the mix has been properly designed and appropriate materials chosen to suit the expected exposure conditions, if the reinforcement has been accurately located in the formwork so as to achieve the required cover and if the mix has been well placed, compacted and cured. These are all issues directly related to good practice in concrete construction, which, if properly implemented, should ensure that reinforcement corrosion should not be a major and recurring problem. While this appears to be a simple solution to the issue of reinforcement corrosion, it is an unfortunate fact that the deterioration of concrete due to corrosion is not uncommon. This may be the result of poor design or the use of concrete of inadequate quality in aggressive conditions. Equally, however, it may be due to simple deficiencies in the concrete such as insufficient cover to the reinforcement, porosity and cracking as a result of poor site practices (workmanship) in placing concrete even in mild-to-medium exposure conditions. In order to mitigate these effects, additional measures of corrosion protection are available to the engineer and builder. These choices include, but are not limited to: * * * * *
* *
the use of membrane-type coatings applied to the surface of concrete; the painting of concrete; the impregnation of concrete with materials intended to reduce its permeability; the addition of corrosion inhibitors to concrete; the use of corrosion-resisting materials (e.g. stainless steels) as replacement for conventional steel reinforcement; cathodic protection of the reinforcement; and/or the application of coatings to the reinforcement itself.
Of these methods, the use of coated steel reinforcement in particular has been widely accepted as an economical and convenient means of providing corrosion protection in many types of concrete construction. The two most common coatings applied to steel reinforcement are fusion-bond epoxy coatings and hot-dip galvanizing. Both systems have been used in moderate-to-severe exposure conditions such as marine and coastal construction, industrial plant, water treatment and chemical processing facilities, power generation and bridge and highway construction. They have also been used in less severe applications in building and construction for both cast-in-place and precast concrete elements.
Galvanized Steel in Concrete: An Overview
5
The use of fusion-bonded epoxy-coated reinforcement has been widely reported in the literature and technical press and will not be dealt with further here. The use of galvanized reinforcement has also been widely reported over a period of more than 30 years but a major review of this topic has not been undertaken from some time. The most thorough treatment of the early work was that published in 1981 by the International Lead Zinc Research Organization [11]. Since this time, however, a large amount of research and field-based studies has been undertaken and a significant body of new scientific evidence and understanding of the performance of galvanized reinforcement in concrete has emerged. This has been reviewed to an extent in several more recent publications [12,13] and is reflected in the publication of a number of national standards and industry guides for the use of zinc-coated (i.e. galvanized) reinforcement as discussed in Chapter 4 of this book.
1.3. Coated Steel Reinforcement — General Principles The coating of conventional reinforcement [13,14] is designed to provide the steel with corrosion protection beyond that afforded by the cover concrete. In the first instance, the coating material, be it metallic or non-metallic, provides barrier-type protection to the steel by isolating it from the local environment. The adhesion and continuity of the coating over the surface, as well as its reactivity in the environment to which it is exposed, are key features of the level of protection provided by the coating. Some examples include organic coatings, such as paints and fusion bonded powders and noble metal coatings on steel, such as tin, chromium, copper and stainless steel. On the other hand, active metal coatings on steel, such as zinc, cadmium and aluminium, provide not only simple barrier protection but they also provide additional cathodic protection in that the coating acts as a sacrificial anode in the event that the underlying steel is exposed. As might be expected, the long-term performance of active metal coatings on steel also depends on the quality and method of application (e.g. adhesion and thickness) of the coating and the reactivity of the coating metal with the local environment. Apart from this, however, it is important to note that the use of coated reinforcement for corrosion control should not be at the expense of using the best quality concrete available and appropriate to the intended application. With this in mind, it can be expected that, providing due care is taken in the specification of the concrete materials and the mix design as well as attention to good workmanship and supervision of the concreting practice, the coating of reinforcement offers a number of advantages over conventional black steel
6
S.R. Yeomans
reinforcement, viz.: *
* * *
*
an increased time to the initiation of corrosion of the steel and a reduced risk of cracking, rust staining and spalling of the concrete; an increase in the service life of the concrete structure or component; a reduction in the frequency and extent of repairs to the concrete; greater tolerance of both the natural variability of the concrete and the presence of inferior quality or poorly placed concrete; and corrosion protection of the steel prior to it being embedded in concrete.
For metal-coated steel reinforcement, a number of coating systems, including zinc, copper, nickel and stainless steel, have been used in concrete. Zinc, in the form of a hot-dip galvanized coating, is by far the most common coating metal for this purpose and research and practical experience with galvanized reinforcement has been widely reported [11,15–17]. In the sections and chapters of this book that follow, a detailed review is given of the properties and behaviour, processing and application of galvanized steel reinforcement for use in concrete construction.
1.4. Galvanized Reinforcing Steel In the broadest sense, galvanizing is a process of coating steel with zinc for the purpose of providing corrosion protection. Zinc can be applied to the surface of steel in a variety of ways but, for structural steelwork (generally .5 mm thick), hot dipping is the preferred and most widely used method [18]. Hot-dip galvanizing involves the immersion of cleaned steel in a bath of molten zinc at about 450 8C, allowing a metallurgical reaction to occur between the steel and the zinc. This reaction produces a coating on the steel made up of a series of iron–zinc alloy layers (gamma, delta and zeta), which grow from the steel/zinc interface with a layer of essentially pure zinc (eta) at the outer surface. The typical morphology of a hot-dipped galvanized coating on steel sections greater than 5 mm thick is shown in Fig. 2. In this case, the total alloy-layer thickness is about 180 mm. A full description of the galvanizing process and the alloy-layer structure of the resultant coating is given in Chapter 4 and will not be dealt with further here. What distinguishes galvanizing from other types of coatings on steel, such as electroplated coatings, powder coatings or paints, is that the coating is metallurgically bonded to the steel. As a result, galvanizing produces a tough and adherent coating which resists abrasion and fairly heavy handling and which can be fabricated by bending without substantial damage to the coating and with little or no effect on its corrosion resistance [18,19]. In most situations, galvanized
Galvanized Steel in Concrete: An Overview
7
Figure 2: Freshly galvanized steel with 180 mm thick alloy layer coating (200 £ ).
bar can be treated and transported as would be the case for conventional black steel bar. In particular, it does not require any special precautions to protect the coating against superficial damage that may occur during fabrication and transport to site [19,20]. Further, the design and construction of reinforced concrete utilizing galvanized reinforcement is essentially the same as that for conventional steel reinforcement and best practice when utilizing galvanized reinforcement is to use appropriately designed and placed concrete, such as would normally be used in general construction. Galvanizing has been used since the 1930s for corrosion protection in many types of reinforced concrete structures and elements exposed to a range of environmental conditions. Evidence from field applications, supported by a growing body of experimental data, has demonstrated that galvanizing extends the life of reinforcement in concrete and provides a safeguard against premature cracking and rust staining of the concrete [11,17,21]. The corrosion protection afforded by galvanizing is due to a combination of beneficial effects, as considered below. Of primary importance is the substantially higher chloride threshold for zinc-coated steel in concrete compared with conventional (uncoated) steel [22,23]. In addition, galvanized reinforcement is resistant to the effects of carbonation of the concrete mass [17]. The net effect of the presence of the zinc coating is that it not only delays the initiation of the corrosion process but it continues to provide barrier protection during the ensuing period when the coating is reacting (i.e. dissolving) but remains intact. What has become clear from the considerable body of research that has been undertaken is that the life of the galvanized coating and thus the reliability of
8
S.R. Yeomans
the corrosion protection it provides, depends on many factors. These include the morphology and thickness of the coating, the quality of the concrete in which it is placed and the severity of the environment to which the concrete is exposed [16,24]. In addition, in circumstances where the underlying steel is exposed, such as at breaks in the coating or at cut edges, the zinc sacrificially protects the steel, thereby further extending the life of the reinforcement [23]. These and other issues are broadly addressed in the following sections and in greater detail in other chapters of this book, mainly Chapters 5 and 6.
1.4.1. Behaviour of Zinc in Concrete Zinc in solution reacts with both strong acidic and strong bases, the attack being most severe below pH 6 and above pH 13. Between these values, the rate of attack is very slow due to the formation of protective layers on the zinc surface [25]. Zinc in concrete is passivated for pH values between about 8 and 12.5 due to the formation of a protective surface film of corrosion product that is relatively insoluble below pH 12.5 [16]. Zinc reacts quite vigorously with wet cement but this reaction effectively ceases once the concrete has hardened. The result of these reactions is the formation of a barrier layer of calcium hydroxyzincate accompanied by the evolution of hydrogen. In ordinary concrete, uncoated steel depassivates once the pH level drops below about 11.5, though in chloride-contaminated concrete this depassivation occurs at higher pH levels. In contrast, zinc-coated steel in concrete remains passivated to pH levels of about 9.5 thereby offering substantial protection against the effects of carbonation of concrete. Zinc-coated reinforcement can also withstand exposure to chloride ion concentrations several times higher than causes corrosion of black steel reinforcement [22,23]. A detailed treatment of the electrochemistry and corrosion behaviour of zinc and the galvanized coating, well beyond the scope of the summary here, is given by Andrade and Alonso in Chapter 5 of this book.
1.4.2. Chloride Tolerance Accelerated corrosion studies in chloride contaminated concrete have revealed the improved corrosion behaviour of galvanized reinforcement over that of conventional steel [23]. Under identical exposure conditions, galvanized reinforcement resisted chloride levels in concrete at least 2.5 times higher than black steel and delayed the time to the onset of corrosion of the underlying steel by some 4 – 5 times. These results have been recently confirmed in other work [26]
Galvanized Steel in Concrete: An Overview
9
where it was also demonstrated that galvanizing had a higher chloride threshold relative to bare steel and delayed the onset of corrosion. In chloride-contaminated concrete, galvanizing increases the time to depassivate the reinforcement, thereby significantly extending the service life of concrete structures in environments where there is a low-to-moderate chloride presence. In concrete with very high chloride levels, the life of the zinc coating may be somewhat reduced due to early depassivation of the zinc. In these circumstances, however, although the longevity of the galvanized coating may be reduced, the overall life of the reinforcement would still be somewhat longer than that of conventional steel in equivalent concrete and exposure conditions due to the inherently higher chloride tolerance of the zinc coating.
1.4.3. Behaviour of the Coating and Corrosion Products An understanding of the reaction mechanism of the zinc alloy coating when placed in concrete and the characteristics of the corrosion products so formed is fundamental to a full appreciation of the corrosion protection afforded by the galvanizing of reinforcement. Considerable work has been done over many years [16,24,27] to investigate these effects, including the reaction of the various coating alloy layers when in contact with wet cement, the nature of the corrosion products that form when zinc reacts with cement and the mixing of the corrosion products into the concrete matrix. This research has indicated that, when the galvanized coating first comes in contact with wet cement and is initially passivated, about 10 mm of zinc is dissolved from the pure zinc (eta) layer of the coating. This effect is shown in Fig. 3a for a galvanized steel with an initial coating thickness of 180 mm (refer Fig. 2) embedded in non-chloride-contaminated concrete for a short period. The average thickness of the coating remaining at this stage is 164 mm and the coating retains a smooth and bright surface. Studies of galvanized bars recovered from field structures indicate that the coating remains in this condition for extended periods of time provided the conditions in the concrete do not change significantly. In such circumstances, very little further metal loss will occur until the zinc is depassivated and active corrosion commences. Once active corrosion of the zinc initiates, usually due to the accumulation of high levels of chloride at the depth of the reinforcement, continued dissolution of the eta alloy layer occurs, followed by progressive dissolution of the underlying alloy layers, as shown in Fig. 3b. This form of attack results in the formation of deep tunnels and holes in the alloy layers, particularly around and through the delta phase, which comprises the bulk of the coating. Although the coating appears to be disintegrating at this stage, a dense layer of both the gamma and delta phases
10
S.R. Yeomans
Figure 3: Changes in the galvanized coating with exposure to concrete (200 £ ). (a) Galvanized bar exposed to fresh concrete showing partial loss of outer pure zinc layer. Remaining coating , 164 mm thick. (b) Long-term exposure to chloride-contaminated concrete showing loss of pure zinc layer with intrusions around underlying alloy layers. Base layers of coating remain intact. Average coating thickness remaining , 110 mm.
remains intact at the bar surface and this affords ongoing corrosion protection to the underlying steel. Once the coating is completely lost from small areas of the bar surface, the zinc continues to provide sacrificial protection over distances of a few mm [23].
Galvanized Steel in Concrete: An Overview
11
Considerable work has also been done to identify the nature of the corrosion products produced and the effect of these on the integrity of the concrete mass [24,28]. A number of minerals have been identified in the corrosion products but they are primarily zinc oxide and zinc hydroxide. A unique feature of these products is that they are friable (loose and powdery) minerals, they are less voluminous than iron-rich corrosion products and they are able to migrate away from the bar and into the adjacent concrete matrix where they fill voids and microcracks. These effects are shown in Fig. 4, in which the plume of zinc-rich corrosion products appears white against the grey calcium-rich cement matrix.1 In contrast to the situation encountered when iron-rich corrosion products form when steel corrodes in concrete, the presence of the zinc corrosion products cause very little physical disruption to the surrounding matrix. This assists in maintaining the integrity of the cover concrete itself, even though active corrosion of the coating may be occurring [24,29]. There is also a suggestion that the presence of these corrosion products and the filling of the pore space in the matrix may create a barrier in the matrix of reduced permeability, which not only increases the adhesion of the matrix to the bar [30] but may also reduce the transport of aggressive species such as chlorides through the matrix to the coating surface.
1.4.4. Experience with Galvanized Steel in Concrete Practical experience and research over many years has clearly demonstrated the benefits of galvanizing for corrosion protection of steel reinforcement in many types of environments, including high chloride-exposure situations [16,32–35]. Galvanizing has been shown to extend the time-to-corrosion of reinforcement and reduce the risk of physical damage to concrete structures such as delamination, cracking and spalling. A detailed treatment of this laboratory and field-based research is given in Chapter 6. Considerable research has been done in the USA, in particular, to investigate the use of galvanized reinforcement for concrete bridge and highway construction exposed to high levels of accumulated chlorides due to the application of deicing salts or in marine exposure. In the case of top and bottom mat reinforcement for bridge decks, for example, when both top and bottom mat bars were galvanized, very low corrosion current densities resulted compared with black steel and the extent of corrosion on the galvanized bars was significantly less with no ferrous corrosion products (i.e. red rust) apparent. It has been shown [35] that, when galvanized bars were used in the top mat only with black steel bottom mats, 1 It has been reported [31] that in some mortars with high acidic contents, the highly expansive corrosion product zinc hydroxychloride may form. This product, with an expansion factor of 2.6 £ , about twice that of zinc oxide for example, may crack the cover concrete if it were to form.
12
S.R. Yeomans
Figure 4: SEM images of interfacial zone between galvanized bar and matrix showing presence of zinc corrosion products [24]. (a) Showing partial dissolution of the galvanized coating (left) and plume of zinc-rich corrosion product (centre) migrating into the cement matrix (right) (1000 £). (b) Migration of zinc-rich corrosion products away from the bar/matrix interface and into the cement matrix. Large particles are fine sand (100 £).
Galvanized Steel in Concrete: An Overview
13
significant corrosion of the zinc occurred, although with very much less red rust corrosion compared with black bars in equivalent conditions. Other work [36] indicated that, for a 0.5 w/c (water/cement ratio) concrete, galvanized bars performed better than black bars although, in a 0.4 w/c concrete, there was similar behaviour for both black and galvanized bars after 8 years cyclic exposure and meaningful comparisons could not be made. It was also noted that the worst-case corrosion occurred when top mat galvanized bars in high chloride concrete were coupled to black steel bars in relatively chloride-free concrete at the bottom of the slab; the best case was when galvanized bars were used in both the top and bottom mats. Other data have also verified the enhanced field performance of galvanized reinforcement in both marine and bridge deck applications [35,38,39]. Surveys of many structures at various ages of exposure with varying concrete quality (high w/c and low cover) and high-to-extreme chloride levels (up to 10 times recommended ACI levels) at the reinforcement have consistently revealed that galvanized steel outperforms black steel where meaningful corrosion comparisons were possible. For example, in 1991, a survey [40] of a number of bridges in Iowa, Florida and Pennsylvania was undertaken to compare the performance of galvanized and uncoated reinforcement in decks exposed year round with humid marine conditions or deicing salts in winter. This survey complemented earlier surveys in 1974 – 1976 and 1981 of many of the same bridges. After periods of up to 24 years exposure, it was found that, generally, the galvanized bars had suffered only superficial corrosion in sound, uncracked concrete, even when the chloride levels were high. Although the chloride levels had increased since the 1981 survey, no major change in the galvanized bars was detected, the average thickness of zinc remaining on the reinforcement had not significantly changed since 1981 and it was still well in excess of that required by ASTM A767 for new material (refer to Chapter 4 for details of relevant Standards). Similar data reported from Bermuda have also verified the long-term durability of galvanized reinforced concrete in marine environments. This topic is covered in detail in Chapter 7. Commencing shortly after WW2, a number of docks, jetties and other infrastructure were constructed using a mix of galvanized and bare-steel bars. A major survey of these structures in 1991 showed that the galvanizing was providing continuing corrosion protection to reinforcement at chloride levels well in excess of threshold levels for bare-steel corrosion [41]. Follow-up examination [42] confirmed the findings of the earlier survey and revealed that the galvanized bars maintained a residual zinc-coating thickness at a structure age of 42 þ years, well in excess of the minimum ASTM requirement. Detailed SEM examination of concrete cores from these structures confirmed the previously mentioned observations that the zinc corrosion products had migrated a considerable distance
14
S.R. Yeomans
(380– 500 mm) beyond the surface of the coating and into the adjacent concrete matrix with no visible effect on the concrete mass. Studies such as these clearly indicate that galvanized bar, when properly used as the exclusive reinforcing material, can provide enhanced corrosion protection compared with black steel in equivalent concrete and exposure conditions. What is clear is that, in good quality concrete that is well compacted, cured and of adequate cover, galvanized bar survives for extended periods of time and offers a costeffective method of corrosion protection. In poor quality concrete, however, particularly those with high w/c ratio and low cover to the reinforcement, galvanizing will delay the onset of chloride-induced corrosion of the reinforcement, but this may be of limited benefit.
1.4.5. Economics of Galvanized Reinforcement When the costs and consequences of corrosion damage to a reinforced concrete building are analysed, the extra cost of galvanizing is seen as a small investment in corrosion protection. While the initial cost of galvanizing may add up to 50% to the cost of the reinforcement, depending on the country of origin and the availability and access to galvanizing plants within the country, the cost of using galvanized reinforcement as a percentage of total building cost is always significantly less than this. The overall cost depends, of course, on the nature and location of construction and the extent to which galvanized bar is used throughout the structure. For example, it is rarely necessary for the structural core or internal elements of large, reinforced concrete structures such as high rise buildings or the deeply embedded components of large abutments and foundations to be galvanized. Cost analysis for building construction [15] reveals that the galvanizing of reinforcement increases the overall cost of reinforced concrete as placed by about 6 – 10%. The actual value will vary depending on many factors, including the type of bar and the galvanizing price, the amount of steel used per cubic meter of concrete poured and the unit cost of the concrete mass. The concrete price is made up of several main components, including the supply of the concrete, the formwork and steel supply and fixing costs. On average, the cost of the steel would not be more than about 25% of the total cost of the concrete as placed. Considering also that it is rarely necessary to galvanize all steel in the structure and that the cost of the structural frame and skin of a building normally represents only about 25– 30% of the total building cost, the additional cost of galvanizing reduces to between 1.5 and 3.0% of the total building cost. However, by galvanizing only certain vulnerable or critical elements, e.g. surface panels, the additional cost of galvanizing reduces further still, perhaps to as little as 0.5 – 1.0%. These
Galvanized Steel in Concrete: An Overview
15
percentages, of course, relate only to total construction costs and, when taken against total project costs or final selling prices, the added cost of galvanizing becomes very small indeed, often less that 0.2%. This represents a very small fraction of the cost of repairs should unprotected reinforcement corrode. Similar costing analysis results have been reported elsewhere [13].
1.5. Corrosion Model for Galvanized Steel in Concrete To assist in understanding the behaviour of galvanized steel in concrete, a servicelife model has been proposed [43]. Based on the traditional Tuutti model for the corrosion of bare steel in concrete [44], an additional stage — protection — has been included between the initiation and propagation stages, as shown in Fig. 5. The protection stage is that during which the galvanized coating reacts and slowly dissolves, thereby providing ongoing protection to the steel. The initiation and protection stages of the model thus relate solely to the behaviour of the galvanized coating itself, that is, the behaviour of the zinc and zinc –iron alloys in concrete, while the propagation stage relates to the active corrosion of exposed steel that is not cathodically protected by adjacent zinc. The three stages of this model are discussed in the following sections.
Figure 5: Corrosion model for galvanized steel in concrete [23].
16
S.R. Yeomans
1.5.1. Initiation Stage As indicated above, the zinc coating is initially passivated by reaction of the pure zinc outer layer with the highly alkaline environment of wet concrete or cement. The evolution of hydrogen is a by-product of this reaction and, on average, about 10 mm of zinc is consumed in this process. Once these reactions subside, the zinc surface remains passive and the rate of further metal loss is exceedingly small until corrosion initiates. Since the corrosion rate of zinc remains very low as the pH drops, it is most unlikely that the depassivation of zinc in concrete can be ascribed to the reduction in pH of the cover concrete due to carbonation processes alone. That is, compared with conventional steel, where both carbonation and chloride ions can induce corrosion, the depassivation of zinc is more than likely due to the accumulation of a threshold level of chloride ions at the depth of the reinforcement. Many studies, both experimental and field, have attempted to determine the chloride threshold levels in concrete at which corrosion of the zinc occurs. Because of variations in the behaviour of different types of galvanized coatings when placed in concretes of widely varying quality, mix proportions, mineral constituents, pH and exposure conditions, it has not been possible to date to determine precise threshold values for the chloride-induced corrosion of zinc in concrete. In many instances, spot data only (e.g. local chloride content and remaining zinc coating thickness) are available, from which it is difficult to determine the threshold concentration. What is clear, though, is that zinc can withstand exposure to chloride-ion concentrations several times higher — at least 2.5 times [23] and perhaps as much as 4 – 5 times [22] — than would cause corrosion of uncoated steel reinforcement. Galvanizing thus delays the onset of chloride-induced corrosion because of the longer exposure time necessary to concentrate the chlorides to the higher threshold levels at the depth of the reinforcement. In good quality concrete that is well compacted and cured and in which the reinforcement is deeply embedded, galvanizing affords considerable long-term protection to the reinforcement, which is reflected in an increased initiation period to the onset of corrosion and clearly outperforms conventional steel reinforcement.
1.5.2. Protection Stage On depassivation of the zinc, corrosion commences and the galvanized layer begins to dissolve. The outer pure zinc (eta) layer of the coating dissolves first, followed more slowly by reaction of the inner zinc–iron (primarily zeta and delta) alloy layers. These alloy layers typically comprise two-thirds or more of the coating thickness (see Fig. 2) and it would appear that they can survive in concrete
Galvanized Steel in Concrete: An Overview
17
for extended periods of time. This is supported by evidence from field surveys of galvanized bar in bridge decks and marine structures (as previously noted) which often show no more than 10 – 30% loss of the original coating thickness through the life of the structure and that more zinc remains on the bar surface than the specified minimum original coating thickness. As previously discussed, it is important to note that, during the protection stage when the zinc is corroding, little or no disruption to the cover concrete occurs, in contrast to the situation usually encountered when steel in concrete corrodes. Thus, even though the zinc is corroding, the cover concrete generally remains intact with little evidence of cracking or delamination and no red rust staining. Finally, even when the coating in a local area has been completely dissolved and the underlying steel exposed, the zinc sacrificially protects the steel. This would also apply to sites along the bars where the galvanizing has been damaged, perhaps during fabrication where cut-ends of bars have not been repaired. The region over which cathodic protection of this form can operate depends on many factors, not the least of which is the conductivity of the concrete electrolyte. Recent work has indicated that zinc protects exposed steel to a distance of at least 8 mm in a 0.8 w/c concrete exposed to continuous salt fog [23].
1.5.3. Propagation Stage Once the barrier afforded by the galvanized coating and the sacrificial protection of exposed steel by adjacent zinc has been exhausted, corrosion of the steel substrate will commence. By this time, however, it is reasonable to expect that the concentration of chloride ions at the depth of the reinforcement would be far in excess of that required to initiate corrosion of black steel. Further, if carbonation of the cover concrete had occurred, although this would not have affected the galvanized bar, the pH of the pore solution would be sufficiently low to prevent the steel from naturally passivating. In any event, immediate and rapid corrosion of the steel substrate would occur, leading to cracking or delamination of the cover concrete, red rust staining and eventual spalling of the mass.
1.5.4. Comparison of Models In Fig. 6, a comparison is made between the model for corrosion of zinc in concrete and that for corrosion of conventional steel. From the discussion above, it is clear there is a delay in the onset of corrosion of the steel substrate of galvanized reinforcement and this effect is reflected in a longer service life of galvanized reinforcement in concrete compared with steel.
18
S.R. Yeomans
Figure 6: Comparison of models showing increased service life for galvanized steel reinforcement in concrete [23].
The total life of the galvanized coating in concrete is made up of the time taken for the zinc to depassivate, which is known to be longer than that for steel, plus the time for the dissolution of the alloy layers in the coating. In high-to-extreme chloride conditions where poor quality concrete has been used (w/c greater than about 0.6) with shallow depth of cover to the reinforcement (15 – 20 mm or less), the overall life of the zinc coating may be reduced to the extent that it only gives a modest increase in life compared with steel. In better quality concrete, however, which has been properly compacted and cured and where the reinforcement has an adequate depth of cover to suit the exposure conditions, the galvanizing of reinforcement does afford long-term protection against corrosion, including concrete exposed to severe, high-chloride conditions. Concerning the presentation of the model itself, it is clear that effects due to cracking of the concrete, which can intensify the corrosion activity and lead to a shortening of the service life, have not been addressed. Of course, such effects similarly influence the behaviour of black steel in concrete.
1.6. Applications of Galvanized Reinforcement Galvanized steel bar and other fittings, including bolts, ties, anchors, dowel bars and piping, have been widely used in a variety of reinforced concrete structures
Galvanized Steel in Concrete: An Overview
19
and elements. The rationale for this is based on the philosophy that the coating provides a safeguard against early or unexpected corrosion of the reinforcement. Should such damage occur, costly repair and remediation of the structure may be necessary in order to realize the full design life of the structure. This represents an ever-increasing economic burden and the redirection of scarce resources. Particular circumstances where the galvanizing of reinforcement is likely to be a cost-effective and sound engineering decision include: * * *
*
* * *
lightweight precast cladding elements and architectural building features; surface exposed beams and columns and exposed slabs; prefabricated building units such as kitchen and bathroom modules and tilt-up construction; immersed or buried elements subject to ground water effects and tidal fluctuations; coastal and marine structures; transport infrastructure including bridge decks, roads and crash barriers; and high-risk structures in aggressive environments.
Many examples exist around the world where galvanized reinforcement has been successfully used in a variety of types of reinforced concrete buildings, structures and general construction including: * * * * * * * * * * *
reinforced concrete bridge decks and pavements; cooling towers and chimneys; coal-storage bunkers; tunnel linings and water-storage tanks and facilities; docks, jetties and offshore platforms; marinas, floating pontoons and moorings; sea walls and coastal balustrades; paper mills, water and sewerage-treatment works; processing facilities and chemical plants; highway fittings and crash barriers; and also lamposts and power poles.
A number of short articles and general reviews have been published over the years dealing with the use of galvanized reinforcing steel in many of these applications [15,21,45]. Some general experiences with hot-dip galvanized reinforcement, mainly from a European perspective, have also been published [13]. In addition, a detailed listing of galvanized reinforced structures, extending to several thousand entries and including buildings, transport infrastructure and chemical and treatment plant, has been published by the American Galvanizers Association [46]. More detailed reviews of the use of galvanized reinforcement,
20
S.R. Yeomans
specifically in highway bridge construction and off-shore constructions for the oil and gas industry in The Netherlands, have been published by Robinson [47] and Van Veen [48], respectively. Some prominent examples, many of which are well-known buildings and major structures from around the world, are listed in Table 1. Other examples in general construction, buildings, bridges and highways and coastal and marine structures are shown in Figs. 7 –10. It is also to be noted that, in Chapter 7 of this book, a detailed review is given of the use of galvanized reinforcement in a range of concrete construction in Bermuda — perhaps the most widely known and prominent large-scale example. Finally, it is worth recording that, in the State of the Art Report on Coating Protection for Reinforcement, originally published in 1992 by the Comite EuroInternational du Beton [13], the benefits from the practical use of galvanized reinforcement were listed as follows: *
*
*
*
*
*
*
*
*
* *
* *
*
proper galvanizing procedures have no significant effect on the mechanical properties of the steel reinforcement; for best performance, galvanized reinforcement should be passivated by chromate treatment; zinc coating furnishes local cathodic protection to the steel as long as the coating has not been consumed; galvanized reinforcement provides protection to the steel during storage and construction prior to placing the concrete; corrosion of galvanized steel in concrete is less intense and less extensive for a substantial period of time than with black steel; galvanized steel in concrete tolerates higher chloride concentration than black steel before corrosion starts; galvanized reinforcement delays the onset of cracking and spalling of concrete is less likely to occur or is delayed; the concrete can be used in more aggressive environments. Thus, a standard design of concrete components can be retained for various exposure conditions by the use of galvanized steel in the most aggressive cases; lightweight and porous concretes can be used with the same cover as for normal concretes; greater compatibility is obtained with low alkali cement; poor workmanship resulting in variable concrete quality (poor compaction, high water/cement ratio) can easily be tolerated; accidentally reduced cover is less dangerous than with black steel; unexpected continuous contact between concrete and trapped water can be tolerated; repair of damaged structures can be delayed longer than with black steel;
Galvanized Steel in Concrete: An Overview
21
Table 1: Examples of prominent structures utilising galvanized reinforcing steel. Sydney Opera House: 35 mm thick galvanized panels for cladding of sails and seawall units NZ Parliament House, Wellington: clad with galvanized reinforced precast fascia panels Bank of Hawaii, Waikiki: thin decorative precast arches reinforced with galvanized bar National Theatre, London: over 1000 t of galvanized reinforcement in exposed parapet walls Crocker Building, San Francisco: galvanized reinforcement in structural elements Collegiate Buildings, University College, London: galvanized reinforcement and mesh New Hall, Cambridge University: galvanized mesh in triangular roof segments Staten Island Community College, New York: brilliant white galvanized reinforced precast panels New Parliament House, Canberra: 1800 galvanized precast cladding panels and window units Offices, Westminster Bridge, London: galvanized reinforced white facing panels Department of Housing and Urban Development, Washington, DC: all reinforcement , 2 in. from surface hot dip galvanized Financial Plaza of the Pacific, Honolulu: precast cladding panels Wrigley Field Sports Arena, Illinois: galvanized reinforced precast panels in seating decks
Hydro-Electricity Commission, Hobart: clad with 950 galvanized reinforced precast panels Telecom Exhibition Exchange, Melbourne: clad with galvanized reinforced precast panels Intercontinental Hotel, Sydney: 1549 precast windows and fascia units ANDOC North Sea Oil Rig: 2000 t galvanized reinforcement in the roof of sea-bed storage tank Eastbourne Congress Theatre, UK: cladding panels and window mullions University Sports Hall, Birmingham: 37 mm thick panels using galvanized reinforcement Library Tower, Sydney: galvanized reinforcement in external columns and panels High Court and National Gallery, Canberra: galvanized reinforcement in all critical areas Barclays Bank, City of London: galvanized precast window surrounds National Tennis Centre, Melbourne: precast stadium support beams University of Wisconsin: precast panels and in situ concrete in numerous buildings Levi Strauss Building: California: precast panels Georgetown University Law Centre, Precast panels (continued)
22
S.R. Yeomans
Table1: Continued. Frontier Chemical Company, USA: galvanized reinforcing mat for floor slabs Bridge deck and road construction in New York, New Jersey, Florida, Iowa, Michigan, Minnesota, Vermont, Pennsylvania, Connecticut, Massachusetts, Ontario and Quebec IBM Data Processing Division HQ, White Plains, NY: hot dip galvanized reinforcement in precast facade panels Coke quenching towers, Dunkirk, France: galvanized structural reinforcement Arkansas Civic Centre: galvanized reinforcement in slim external columns Offshore piers at Ominichi, Japan and Riva di Traiano, Rome, Italy: galvanized reinforced throughout
*
*
*
US Coast Guard Barracks, Elizabeth City, NC: galvanized bar in 237 precast panels John F Kennedy Parking Garage, Detroit: galvanized reinforcing steel to protect against subsurface rusting Football Hall of Fame Stadium, Canton, OH: galvanized reinforcing steel Dome of the Mosque, Rome, Italy: galvanized reinforcement Power station cooling water ducts, Spijk, The Netherlands: fully galvanized reinforced Toutry Viaduct, St Nazaire Bridge and Pont d’Ouche Viaduct, France: galvanized reinforcing bars
galvanized hardware is acceptable at the surface of the concrete, as it is for the joints between precast panels; the use of galvanized reinforcement ensures a clean appearance of the finished concrete with no trouble arising at cracks, either from spalling or rust staining; and galvanized reinforcement is cleaner and easier to work with and makes it possible to consider the use of thinner wires as welded fabrics.
The report goes on to say that “it is important to remember that even if these benefits are achieved, the use of galvanized reinforcement should not be considered as an alternative to the provisions of adequate cover of dense, impermeable concrete, unless special design criteria have to be met. Galvanizing of reinforcement is a complementary measure of corrosion protection — a kind of insurance against the inability of the concrete to isolate and protect the steel”.
Galvanized Steel in Concrete: An Overview
23
Figure 7: Galvanized reinforcement concrete — general construction. (See colour plate 1.)
24
S.R. Yeomans
Figure 8: Galvanized reinforcement concrete — buildings. (See colour plate 2.)
Galvanized Steel in Concrete: An Overview
25
Figure 9: Galvanized reinforcement concrete — bridges and highways. (See colour plate 3.)
26
S.R. Yeomans
Figure 10: Galvanized reinforcement concrete — coastal and marine. (See colour plate 4.)
1.7. Summary Over a very long period of time (in fact about 60 years), the galvanizing of steel reinforcement has been shown to provide a cost-effective and reliable means of corrosion protection to concrete in a variety of exposure conditions. Clearly, galvanizing is only one of a number of protection systems that can be used in reinforced concrete. However, the convenience of manufacture and supply of
Galvanized Steel in Concrete: An Overview
27
the product, the ease of handling, transportation and installation and the fact that no special design requirements are needed has meant that it has been accepted in many countries for a wide range of concrete construction. The last 15– 20 years in particular have seen extensive research and field investigations undertaken of the characteristics and behaviour of galvanized reinforcement. This considerable body of work has repeatedly highlighted the benefits of galvanizing in delaying the onset of corrosion in reinforced concrete and in reducing the risks of cracking and rust staining of the concrete mass. The higher chloride threshold for zinc compared with steel and the fact that zinc in concrete is virtually unaffected by carbonation provides galvanized reinforcement with an inherent corrosion resistance well beyond that of conventional steel bar. The very presence of the coating itself further extends the service life of galvanized bar because of the time delay before which dissolution of the coating occurs. As with all corrosion-protection systems, there is a cost associated with galvanizing. Although the cost of the reinforcement may increase by about 50% when galvanized, when considered against total building and construction costs and the enormous potential costs associated with untimely repair of damaged concrete, the premium that is paid to galvanize reinforcement is very small indeed. Even if one cycle of local patch repairs over a large concrete structure can be avoided, the cost of the galvanizing would have more than been met. Primarily, however, the reason for using any corrosion-protection system is to extend the service life of the structure. Experience with galvanizing has shown this can be readily achieved in many types of reinforced concrete structures and elements in mild, moderate and severe exposure conditions. Above all, however, it is important to remember that, when using galvanized reinforcement (as with any protection system for concrete), the concrete is properly designed and placed and is appropriate for the type of element and the exposure conditions. Unless specific design requirements apply, such as reduced cover or ultra-lightweight construction, the concrete should be designed and placed as though conventional steel reinforcement was to be used. In essence, the use of galvanizing should not be at the expense of this basic quality and integrity of the concrete. In this way, the galvanizing can be considered to provide protection against those circumstances that may lead to premature corrosion of conventional reinforcement and deterioration of the concrete mass.
References [1] Hausmann, D. A. (1967). Steel corrosion in concrete. Materials Protection, 6, 11, 19–23. [2] Page, C. L., & Treadaway, K. W. J. (1982). Aspects of the electrochemistry of steel in concrete. Nature, 297, 109–114.
28
S.R. Yeomans
[3] Gibson, F. W. (1987). Corrosion, Concretes and chlorides — steel corrosion concrete: causes and restraints, ACI SP-102. American Concrete Institute, Detroit, MI, USA. [4] Berke, N. S., Chaker, V., & Whiting, D. (1990). Corrosion rates of steel in concrete, ASTM STP 1065. American Society for Testing and Materials, Philadelphia, PA, USA. [5] Page, C. L., Bamforth, P. B., & Figg, J. W. (1996). Corrosion of reinforcement in concrete construction, Royal Society of Chemistry, Special Publication No. 183. Royal Society of Chemistry, London. [6] Broomfield, J. P. (1997). Corrosion of steel in concrete. E & FN Spon/Chapman & Hall, London. [7] Bentur, N., Diamond, S., & Berke, N. S. (1997). Steel corrosion in concrete. E & FN Spon/Chapman & Hall, London. [8] Vassie, P. R. (1980). TRRL Laboratory Report 953. Department of Transport/Department of Environment, UK, 36 pp. [9] Beresford, F. D. (1979). Concrete Institute of Australia Conference, Canberra. Concrete Institute of Australia, North Sydney, NSW, Australia, pp. 2–10. [10] Transportation Research Board. (1991). Highway deicing: comparing salt and calcium magnesium acetate, Special Report 235. National Research Council, Washington, DC, USA. [11] ILZRO. (1981). Galvanized reinforcement for concrete — II. International Lead Zinc Research Organization, NC, USA. [12] ILZIC. (1995). Protection of reinforcement in concrete: an update. Indian Lead Zinc Information Centre, New Delhi, India. [13] Comite´ Euro-International du Beton. (1992). Coating protection for reinforcement: state of the art report. CEB Bulletin d’Information No. 211, Chapters 2 and 5 (also published by Thomas Telford Services Ltd, 1995). [14] Yeomans, S. R. (1993). Considerations of the characteristics and use of coated steel reinforcement in concrete, NISTIR 5211. National Institute of Standards of Technology, Gaithersburg, MD, USA, 41 pp. [15] Yeomans, S. R. (1987). Galvanized steel reinforcement in concrete, First National Structural Engineering Conference. Institution of Engineers, Australia, ACT, Australia, pp. 662–667. [16] Andrade, M. C., & Macais, A. (1988). Galvanized reinforcements in concrete. In: A. Wilson, J. Nicholson, & H. Prosser (Eds), Surface coatings — 2. Elsevier, London, Chapter 3. [17] Swamy, R. N. (1990). In-situ behaviour of galvanized reinforcement. In: J.M. Baker, et al. (Eds), Durability of building materials and components, Proceedings of Fifth International Conference, Brighton, UK. Chapman & Hall, London, pp. 299–312. [18] Porter, F. (1991). Zinc handbook: properties, processing and use in design. Marcel Dekker, New York. [19] American Galvanizing Association. (2002). Hot-dip galvanizing for corrosion prevention: a guide to specifying and inspecting hot-dip galvanized reinforcing steel. AGA, Centennial, CO, USA, 20 pp. [20] Galvanizers Association of Australia. (1999). After-fabrication hot dip galvanizing. Galvanizers Association of Australia, Melbourne, VIC, Australia, 72 pp. [21] Yeomans, S. R. (2001). Applications of galvanized rebar in reinforced concrete structures, NACE Corrosion 2001, Paper 01638. National Association of Corrosion Engineers, Houston, TX, USA, 12 pp.
Galvanized Steel in Concrete: An Overview
29
[22] Tonini, D. E., & Dean, S. W. (1976). Chloride corrosion of steel in concrete, ASTM STP 629. American Society for Testing and Materials, Philadelphia, PA, USA. [23] Yeomans, S. R. (1994). Performance of black, galvanized, and epoxy-coated reinforcing steel in chloride-contaminated concrete. Corrosion, 50, 1, 72–81. [24] Yeomans, S. R. (1998). Corrosion of the zinc alloy coating in galvanized reinforced concrete, NACE Corrosion, Paper 98653. National Association of Corrosion Engineers, Houston, TX, USA, 10 pp. [25] Rotheli, B. E., Cox, G. L., & Littreal, W. B. (1932). Effect of pH on the corrosion products and corrosion rate of zinc in oxygenated aqueous solutions. Metals and Alloys, 3, 73–76. [26] Bautista, A., & Gonzalez, J. A. (1996). Analysis of the protective efficiency of galvanizing against corrosion of reinforcements embedded in chloride contaminated concrete. Cement and Concrete Research, 26, 2, 215–224. [27] Maahn, E., & Sorensen, B. (1986). The influence of microstructure on the corrosion properties of hot-dip galvanized reinforcement in concrete. Corrosion, 42, 4, 187–196. [28] Hoke, J. H., Pickering, H. W., & Rosengarth, K. (1981). Cracking of reinforced concrete, ILZRO Project ZE-271. International Lead Zinc Research Organization, NC, USA. [29] Covino, B. S., et al. (1997). Interfacial chemistry of zinc anodes for reinforced concrete structures, NACE Corrosion 97, Paper No 97233. National Association of Corrosion Engineers, Houston, TX, USA, 20 pp. [30] Fratesi, R., Moriconi, G., & Coppola, L. (1996). The influence of steel galvanization on rebar behaviour in concrete. In: C.L. Page, et al. (Eds), Corrosion of reinforcement in concrete construction, Royal Society of Chemistry, Special Publication No 183. Royal Society of Chemistry, London. [31] Hime, W. G., & Machin, M. (1993). Performance variances of galvanized steel in mortar and concrete. Corrosion, 49, 10, 858–860. [32] Cornet, I., & Bresler, B. (1981). Galvanized steel in concrete: literature review and assessment of performance, Galvanized reinforcement for concrete — II. International Lead Zinc Research Organization, NC, USA, pp. 1– 56. [33] Tonini, D. E., & Cook, A. R. (1981). The performance of galvanized reinforcement in high chloride environments - field study results, Galvanized reinforcement for concrete — II. International Lead Zinc Research Organization, NC, USA, pp. 57 –74. [34] Griffin, D. F. (1979). Effectiveness of zinc coating on reinforcing steel in concrete exposed to a marine environment, Technical Note N-1032. Naval Civil Engineering Laboratory, Port Hueneme, CA, USA. [35] Stejskal, B. G. (1992). Evaluation of the performance of galvanized steel reinforcement in concrete bridge decks, CTL Project 050324. Construction Technology Laboratories, Skokie, IL, USA. [36] Pfeifer, D. W., Landgren, J. R., & Zoob, A. (1987). Protective systems for new prestressed and substructure concrete, Report No FHWA-RD-86-193. Federal Highways Administration, Washington, DC, USA. [37] Clear, K. C. (1981). Time-to-corrosion of reinforcing steel in concrete slabs, volume 4: galvanized reinforcing steel, Report No. FHWA-RD-82-028. Federal Highways Administration, Washington, DC, USA. [38] Stark, D., & Perenchio, W. (1975). The performance of galvanized reinforcement in concrete bridge decks, ILZRO Project ZE-206. International Lead Zinc Research Organization, NC, USA.
30
S.R. Yeomans
[39] Stark, D. (1978). Galvanized reinforcement in concrete containing chlorides, ILZRO Project ZE-247. International Lead Zinc Research Organization, NC, USA. [40] Dugan, B. P., Stejskal, B. G., & Stoneman, A. M. (1994). Survey of bridges containing galvanized rebar. In: R. N. Swamy (Ed.), Corrosion and corrosion protection of steel in concrete (pp. 157–170). Sheffield Academic Press, UK. [41] Allan, N. D. (1991). Galvanized Reinforcement: the Bermuda experience, Galvanized Rebar Seminar. Structural Integrity Research Unit, University of Sheffield, Sheffield, UK. [42] Kinstler, T. J. Research Update on galvanized reinforcing steel. Private communication. Industrial Galvanizers America, Birmingham, AL, USA. [43] Yeomans, S. R. (1994). A conceptual model for the corrosion of galvanized steel reinforcement in concrete. In: R. N. Swamy (Ed.), Corrosion and corrosion protection of steel in concrete (pp. 1299– 1309). Sheffield Academic Press, UK. [44] Tuutti, K. (1982). Corrosion of steel in concrete. Swedish Cement and Concrete Research Institute, Stockholm. [45] Macgregor, B. R. (1987). Galvanized solution to rebar corrosion. Civil Engineering, 18–21. [46] American Galvanizing Association. Listing of galvanized reinforced concrete structures. AGA, Centennial, CO, USA. [47] Robinson, P. E. (1974). Using galvanized steel reinforcing bars for highway bridge decks. The Construction Specifier, May. [48] Van Veen, A. L. (1976). The galvanizing of rebar for off-shore constructions, Proceedings 11th International Galvanizing Conference, Madrid. Zinc Development Association/ Portcullis Press, UK, pp. 173–179.
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 2
Design for Durability with Galvanized Reinforcement R. Narayan Swamy University of Sheffield, Sheffield, UK
2.1. Introduction Reinforced concrete (RC) is, in many respects, a wonder material of the construction industry. The combination of concrete and steel has produced a remarkable composite material, the constituents of which are highly compatible and empathetic to each other. Concrete is strong in compression whilst the steel carries all the tensile forces but, jointly, they are able to withstand a wide range of static and dynamic forces and produce structural elements that are strong, stable and aesthetically pleasing. Because of their composite nature, the long-term performance characteristics of RC, its response to loads and the surrounding environment as well as its designed service life are highly dependent not only on the engineering and durability properties of the individual components but also on their ability to preserve and sustain the bond between them and the resulting composite action throughout their service life. Since concrete supplies the bulk and the mass for RC, and since it also provides a protective environment for the steel from agents that destroy its electrochemical stability, concrete often becomes, and is treated as, the dominant partner of RC. Nevertheless, the stability of RC in aggressive environments is influenced as much by the integrity of steel as of concrete, and both are of serious concern to engineers and designers. Concrete, the primary constituent of reinforced construction, is an international construction material in its own right. Of all construction materials, it has the best ecological profile for a given engineering property such as strength or elastic modulus. It is probably the most widely and extensively used building material in the world due to its relatively low cost, its versatility and adaptability and the easy availability of its constituents. It is so easily prepared and fabricated into so many shapes and structural systems in the realms of infrastructure, transport, work and habitation that the material is often identified with a nation’s
32
R. Narayan Swamy
stability, economic progress and, indeed, the quality of human life. The most outstanding asset of concrete is its inherent alkalinity, providing a passivating mechanism and a safe, non-corroding environment for the steel reinforcement embedded in it. Long experience and a good understanding of the material have proved that concrete is a reliable and durable construction material when it is exposed to normal or even moderately aggressive environments. In spite of these intrinsic technical and economic advantages of concrete, and in spite of the tremendous scientific advances that have been made in our understanding of its microstructure and engineering, deterioration of concrete has become a major global problem and there is widespread concern about its lack of durability [1]. In recent times, in many parts of the world, reinforcement corrosion has become the main cause of early and premature deterioration, and sometimes failure, of RC structures [2– 5]. One of the major factors contributing to this deterioration process is the environmental and climatic conditions to which a concrete structure is exposed. Hot/dry and hot/wet salt-laden environments, as well as temperate/cold climates with extensive use of deicing salts, probably provide the most aggressive forces that undermine the stability and durability of concrete structures. When the severity of exposure is compounded with poorquality concrete and/or defective design and construction practices, the process of deterioration becomes interactive, cumulative and very rapid, and a cancerous growth occurs that cannot be easily stopped. This dichotomous situation poses one of the greatest challenges to engineers and concrete material scientists; namely, how do we design and construct structures that will have a long and durable service life and which will require a minimum of repair and retrofitting? In other words, how do we develop a RC element that can be designed to give optimum performance for a given set of load conditions and usage in harsh salt-laden environments consistent with the requirements of cost, service life, sustainability and durability? The overall aim of this chapter is to show that it is inherent in the nature of concrete as a material, and in construction as a technology, that there is no single approach that will ensure long-term concrete material stability and structural integrity when RC structures are exposed to severely aggressive environments. The chapter advocates a global design strategy — a holistic approach — incorporating a judicious integration of material properties and structural design that will give a structure its prescribed and specified durable service life. Such a strategy will involve four distinct operations, namely: * * * *
the development of a highly impermeable concrete matrix; the protection of concrete; the protection of reinforcement; and design for structural integrity.
Design for Durability with Galvanized Reinforcement
33
In harsh environments with rapidly changing climatic and ambient conditions, all these distinct design stages would need to be incorporated and close attention will have to be paid to all these aspects if structures are to function efficiently and satisfactorily. The focus of this chapter is to show that, in this design strategy, protection of steel from the external environment and the effects of the ingress of air-borne and water-borne contaminants is the first and indisputable step in the design process. This chapter discusses three distinct aspects of this steel protection technology, each of which will be explored in the following sections. These are: (1) the need to protect steel, (2) the basic requirements of galvanized reinforcing bars, and (3) material and structural design implications of the use of galvanized bars. Adequate cover and good-quality concrete are essential ingredients for durability of all reinforcing bars in concrete, and this is also true for all coated bars including galvanized, epoxy-coated and stainless clad bars. This chapter presents extensive data to show that, under such conditions, galvanized bars can provide excellent corrosion resistance to RC structures and substantially extend their durable service life. The great advantage of a metallic coating such as galvanizing is that it not only provides a primary barrier protection but also a secondary cathodic protection when the zinc coating is damaged and the steel substrate is exposed. It gives protection to the reinforcement before it is embedded in concrete and it preserves for much longer the electrochemical stability of steel, counteracting the effects of the inevitable practical problems arising during construction such as variable cover and inconsistent quality of concrete. The increased period before the initiation of corrosion ensures that the durable service life of the structure as a whole is significantly extended.
2.2. Current Design Specifications Experience during the last four to five decades has shown that corrosion of reinforcement continues to be the major cause of deterioration of concrete structures [1– 5]. There is now strong evidence to establish and recognize that current design codes and specifications lead to structures that do not have adequate resistance to severe exposure conditions, such as those prevalent in the torrid, temperate and frigid zones, and to the effects of chloride attack. The inadequacies lie largely in current approaches to design, which are based on the deemed to satisfy premise, namely that, if the requirements of concrete grade, cement content, water/cement (w/c) ratio and concrete cover are met, the structure is, per se, considered durable. These assumptions have been proved to be totally
34
R. Narayan Swamy
inadequate to ensure durability, even when the design specifications are correctly implemented [6,7]. Exposure to ambient environment is the one single, predominant external factor that is beyond human control, which can create an alarming degree of degradation in a short time and critically determine the stability and serviceability of concrete structures [8]. Engineers tend to underestimate the effects of aggressive environments on the durability of concrete and RC and the time-dependent and interactive effects of load, exposure and climatic conditions [9–13]. Thus, current design specifications related to concrete mixes and cover depths grossly underestimate the material and structural implications of exposure to aggressive environments. They will lead to insufficient margins of safety against durable service life and the number of failures primarily due to corrosion is likely to continue to be unacceptably high.
2.3. The Ubiquitous Nature of Exposure As noted earlier, one of the major factors directly influencing the durability and service life of concrete structures is the nature of the exposure condition and environment in which the structure has to breathe, live and carry out its intended function. Engineers and national building/structural design codes totally underestimate their cumulative and synergistic effects, the ability of structures to maintain their integrity and their capability to carry loads when exposed to such conditions over a long period of time. In the following sections, some of these time-dependent and interactive effects are briefly highlighted.
2.3.1. Environmental Considerations Adverse geomorphological and climatic conditions, such as severe ground contamination, high ambient salinity, high temperature and high humidity, generate the most imperceptible and insidious forces that damage the longterm ability of concrete to give sustained protection to steel. Similarly, daily and seasonal fluctuations of temperature and relative humidity (RH) can create cyclic thermal and moisture movements. Daily temperature fluctuations in the torrid regions, for example, can be as high as 20 –308C in the hot season, whereas the RH may range between 30 and 100% over a 24-h period [14]. In some regions, the RH can be well above 40 –50% most of the year and in Qatar, for example, well above 85% throughout the year. A highly variable, hot and moisture-laden environment can create damaging wetting/drying cycles of exposure.
Design for Durability with Galvanized Reinforcement
35
The high ambient salinity in a hot/humid environment also creates a permanent aggressive atmosphere for the concrete. The high salinity of coastal waters can generate a high rate of sea-salt deposition of the order of 1.5 £ 1023 kg/m2/y. This, in turn, will create high chloride and sulphate contents in the atmospheric air. The chloride content, for example, can be as high as 65 £ 1023 mg/m3 whereas the sulphate content can be of the order of 35 £ 1023 mg/m3. A high ambient temperature will accelerate the rate of corrosion but the synergistic effects of the simultaneous presence of chlorides, sulphates and elevated temperature on the long-term properties of concrete and steel corrosion are not fully understood [15– 17]. The presence in the atmosphere of industrial pollutants such as SO2 and CO2 gasses, and also carbonaceous particles due to fuel burning, can accelerate the deterioration process of concrete and steel. Coming from the burning of coal, oil or gasoline, SO2 is probably one of the most corrosive pollutants of the atmosphere [18 –20]. A high ambient humidity will increase the adsorption of SO2 onto carbon steel by almost 100 times as the RH increases from about 60 to 95% [21,22]. The maximum RH in the Arabian Gulf environment, for example, which is often well above 85% throughout the year [14], provides just the sort of conditions to accelerate corrosion damage. Atmospheric corrosion can be a maximum in a hot/humid environment. A combination of high average temperature, high RH, air-entrained sea salt and industrial pollutants can create the ideal conditions for maximum atmospheric corrosion. Serious rusting can occur in the presence of small concentrations of impurities, even without visible precipitation of moisture, once the RH of the air rises above a critical value. With SO2, for example, such rusting can occur at RH values of 65– 75% whereas, in the presence of chlorides, rusting may occur at a RH as low as 40%. For steel, copper, zinc and nickel, the rate of atmospheric corrosion generally increases rapidly above a critical RH of 50– 70% [21,23,24] and, in many parts of the torrid zone, the RH during most of the year can be well above 40 –50%. The synergistic effects on concrete of high average temperature, high fluctuations of temperature and RH, air-entrained chlorides and sulphates and the presence of industrial pollutants are not known and are difficult to predict. However, there is considerable evidence to show that a combination of high temperature and high concentrations of chlorides and sulphates will increase the free chloride ions in the pore solution of concrete [15–17,25]. Finally, radiation can be very high, between 500 and 800 MWh/cm2 over a 12-month period, and this can drastically enhance surface temperatures and thermal gradients in concrete structural elements, leading to cracking and corrosion.
36
R. Narayan Swamy
2.3.2. Geomorphological Considerations Generally, the top 5 – 6 m of soil in coastal areas in the torrid zones will contain a high concentration of chemicals such as chlorides and sulphates. A hot, drying climate causes high evaporation of water from the upper layers, leaving the salts accumulated in the soil. The relatively small changes of tidal movements in coastal areas can also contribute to the accumulation of salts in the soil. As a result, chloride and sulphate concentrations remain very high in both the soil and ground water. Often, concentrations of organic matter are found in the soil at a depth range of 0 – 5 m. Nitrate concentration can also occur in the ground water. Chloride salts are generally highly soluble and chloride concentration therefore increases with the depth of the soil. Sulphate salts, on the other hand, are either negligibly soluble or are only slightly soluble — and the high evaporation rate will ensure that sulphate concentration becomes highest in the top layers of the soil. Where impermeable soil layers exist near the ground surface, high concentrations of both chloride and sulphates may simultaneously occur in the top layers of the soil [26].
2.3.3. Effects on Materials and Concrete The maximum wind speed in many coastal areas can be well above 40 km/h throughout the year. Wind-transported air-borne dust and salt will lead to contamination of stored concrete materials, of exposed steel bars and of concrete itself. In addition, in a moist-laden environment, salt can be deposited on the exposed surfaces and may work into the cracks in existing structures. Evaporation rates can often greatly exceed precipitation — and the former can vary from 10 mm to as high as 25 mm in the summer. High evaporation rates and high levels of solar radiation would increase the water demand and accelerate the workability loss of fresh concrete and the hydration of cement, all of which can cause premature setting, leading to cold joints. The high rates of evaporation of the concrete mix and curing water would result in very poor-quality cover concrete that can accelerate the transport of chlorides and sulphates into the concrete and eventually lead to steel corrosion. They can also make chemical attack extremely rapid.
2.4. The Need for Protection of Steel in Concrete The implication of the discussion above is simply this. The exposure condition — the environment in which a structure is located — is the worst enemy of concrete and steel, enhancing cracking and damage of the concrete and destabilization of
Design for Durability with Galvanized Reinforcement
37
the steel. These effects may lead to premature deterioration of the structure with resultant loss of stiffness, strength, stability and safety. Continuous exposure to high ambient temperature and humidity and large fluctuations of temperature and humidity will lead to high thermal gradients, high thermal stresses, thermal fatigue, cyclic wetting and drying and internal and external microcracking. When these conditions are compounded by a salt-contaminated moisture-laden environment, their cumulative effects will inevitably lead to microstructural degradation of concrete in terms of increased porosity, coarsening of the pore structure and enhanced permeability. The question is how does the nature of concrete respond to this repeated environmental attack and how does it affect its ability to protect the embedded steel from corrosion?
2.4.1. The Nature of Concrete Concrete is basically a heterogeneous, discontinuous composite material with an intrinsically porous matrix. By its very nature, concrete is a highly variable material. One of the most intriguing and puzzling characteristics of concrete is the paradox that, whilst being intrinsically protective to steel through its alkalinity, it is also the same material that permits and controls the ingress of water, oxygen, chlorides, sulphates and other agents that will eventually damage the material and progressively destroy the electrochemical stability of the embedded steel. Whatever the good qualities of concrete may be, and however sound and durable it is, when exposed to severely aggressive environments over a long period of time, the material suffers from two inherent weaknesses: *
*
First, the hydration process, and hence the development of strength and of a highly impermeable pore structure, is a time-dependent operation that requires a favourable wet environment for the chemical interactions to initiate and continue. A high resistance to environmental attack therefore demands that concrete is permitted to mature and be shielded from adverse exposure conditions, particularly at early ages, when the material is highly vulnerable to drying as well as to diffusion processes arising from external exposure. Second, the degradation process of concrete leading to corrosion of the steel, apart from also being time-dependent, is often not the result of a single factor, process or sole aggressive agent. In a complex composite system such as concrete, an aggressive environment becomes a major factor in initiating a progressively cumulative damage activity. Indeed, the systematic advancement of deterioration is a synergistic process, a complex combination of many individual mechanisms, the exact role, effect and contribution of each of which to the totality of damage is not clearly understood.
38
R. Narayan Swamy
In structures carrying loads and exposed to aggressive environments, both these major processes do not occur in isolation or in idealized surroundings, but are dominated by the agents of deterioration in the environment and the highly variable ambient conditions. Together, they create a range of unknown and interactive effects that initiate a process of deterioration that will eventually destroy the capability of the element to function with adequate margins of safety and without sacrificing structural integrity. It can be readily seen that, with RC structures exposed to hostile environments, it is debatable whether one can totally rely on the quality of concrete alone to protect the embedded steel from corrosion, or if the steel would also need corrosion protection of some kind or other. Apart from the two aspects of concrete behaviour discussed earlier, there are three factors that control the electrochemical stability of steel in concrete, namely, cracking, depth and quality of cover to the steel and the overall quality of the structural concrete. Of these three factors, the effects of cracking are the most difficult and imprecise to evaluate. Cracking is a random phenomenon and all structural concretes are in effect in a cracked state whatever the quality of the concrete. The cracks themselves may not be easily visible to the naked eye since the tensile strain capacity of concrete is only of the order of 150–200 microstrain. There is only limited information on the role of cracking, both internal and external, on the service life of concrete structures [12,27,28]. The depth of cover to steel can be effectively controlled within defined limits provided adequate procedures are followed during construction. The quality of the cover concrete and that of the core concrete in a structural element can also be improved [29]. The fact remains, however, that, whatever the quality of materials and fabrication, the nature of concrete construction is such that microcracks and micropores will always exist at the surface and within the body of the cover concrete. These will not only trap aggressive elements but also provide a path for their transport into the interior of the concrete. Further, at water-cementitious ratios of 0.5 and above, continuous capillary channels can never be blocked. Similarly, with shallow covers of 10 –20 mm, even the best corrosion protection will begin to deteriorate with time. All these factors dictate that an integrated design approach alone can ensure material stability and structural integrity of concrete constructions.
2.4.2. Deterioration of RC Perhaps the strongest argument in favour of protection of steel reinforcement is the worldwide experience of structural deterioration during the last four to five decades. The fact that there have been an unacceptably high number of structures
Design for Durability with Galvanized Reinforcement
39
suffering from premature corrosion deterioration, particularly in bridge structures, even when specific code requirements for durability in terms of concrete quality and concrete cover have been adhered to, points out that modern Portland cements are not totally resistant to chloride-ion penetration, even when the water –cement ratio is as low as 0.45, as shown in Fig. 1 [8,30,31]. The worldwide deterioration of, and damage to, concrete and concrete structures is best illustrated by the following data. In the UK, for example, the total construction industry output in 1995 was about £52 billion, of which some 50% was spent on maintenance, repair and rehabilitation. The amount spent on repair and rehabilitation has risen from about £6.9 billion in 1981 to something in excess of £26.0 billion about 20 years later. As a specific example, the Midlands link motorway around Birmingham cost £28 million to construct; between 1972 and 1989, £45 million was spent on repairs and it is now estimated that another £120 million will be required in the next 15 years [32]. The cost of corrosion to the UK economy is estimated at £15 billion per year. In Europe, structural damage repair every year is estimated to cost 1.4 billion euro. The more recent state of the nation report grades the quality of the UK’s infrastructure from B (fair) to D þ (poor) with an overall average grading of C [33].
Figure 1: Effect of w/c ratio on chloride penetration into Portland cement concrete after 50 cycles.
40
R. Narayan Swamy
In the US, the 1998 ASCE Report on America’s Infrastructure estimated a fiveyear total investment need of $US1.3 trillion just to put back the roads, bridges, dams, water supply and other infrastructure systems to good serviceable life. The average state of America’s infrastructure was given a grade D (poor) [4,34]. As a specific example, some 40% of more than 500,000 highway bridges are rated as structurally deficient or functionally obsolete. Some $US100 billion is the estimated requirement to eliminate the current backlog of bridge deficiencies and maintain repair levels. The above details show that the record of concrete as a material of everlasting durability has been greatly impaired. There are many reasons for this situation. There is partly the perceived image of concrete as a material of enduring quality that needs no maintenance and as a material that will not deteriorate. There is also the assumption that somehow the impermeability of concrete and the protection of the embedded steel will be automatically and adequately provided for by the cover thickness and the presumed quality of concrete. Experience has shown that neither can be assumed as a natural consequence of the process of concrete fabrication. However, the major reason for this much lower resistance of modern Portland cement concretes to chloride-ion penetration is the significant changes that have occurred during this period to the chemical composition of Portland cement. Firstly, there has been a substantial increase in the C3S/C2S ratio from about 1.2 to 3.0, resulting in higher strengths at early ages and thereby the structural design strengths being achieved with lower cement contents and higher w/c ratios. Secondly, these changes in chemical composition have produced a consequent increase in the heat of hydration and, more importantly, an increase in peak temperature occurring at earlier ages. It is estimated that the average increase in peak temperature is about 17%, and this peak temperature is reached in less than half the time [35]. Microcracks induced by thermal gradients and thermal fatigue are hard to heal and take a long time to close. Although there is probably no direct relationship between strength and durability, the effects of these changes in Portland cement composition and heat of hydration are cumulative and adverse. It is possible that many of the disadvantages of Portland cement concretes can be overcome by incorporating siliceous pozzolanic/cementitious mineral admixtures such as fly ash, slag, silica fume and metakaolin in concrete. There is now incontrovertible evidence that such cement replacement materials can substantially enhance the durability properties of the resulting concrete through pore refinement and chemical bonding of the constituents, as shown in Fig. 2 [31,35]. However, the development of an impermeable microstructure in such concretes is also a time-dependent process and is controlled primarily by the availability of a favourable environment in which the pozzolanic/cementitious reactions can continue to bring about the superior pore refinement. But, more importantly, concretes with mineral admixtures crack at about the same tensile
Design for Durability with Galvanized Reinforcement
41
Figure 2: Role of mineral admixtures on chloride penetration into concrete-cement replacement: 30% fly ash (FA), slag 65%, silica fume (SF) 10%.
strain capacity as Portland cement concrete and, once cracked, the structure is open to attack by environmental agents. So, although mineral admixtures can impart significant and substantial improvements to the durability properties of concrete, they do not and cannot, per se, guarantee total resistance to chloride penetration for concrete exposed to heavily salt-laden environments.
2.5. Design Strategy for Structural Integrity So, what is the way forward? How can engineers ensure through design the electrochemical stability of steel in concrete, and thereby enhance the corrosionfree durable service life of concrete structures? All the previous data emphasize that, in harsh and aggressive environments and in rapidly changing ambient conditions, engineers need to think in terms of a global design strategy — a holistic design approach. Such a strategy should consider two important design steps [36– 38]: *
*
first, enable concrete to attain a high degree of dense microstructure and impermeability, and second, use steel that will have a high degree of corrosion resistance in saltcontaminated concrete.
Protection of concrete from an unfriendly environment is important and imperative, but protection of steel reinforcement is much more critical since the electrochemical destabilisation of steel will undoubtedly lead to loss of metal
42
R. Narayan Swamy
reinforcement, eventually to loss of strength and structural stiffness and, ultimately, to lack of safety. Corroding steel is like an incurable cancer and will eventually destroy the stability and integrity of any structure. The severity of the exposure conditions will finally decide whether both the concrete and steel need to be protected but, where chloride contamination is mild and never severe, protection of steel together with a high-quality concrete may in the long run be more cost-effective and prudent than protection of concrete alone. Protection of steel reinforcement then becomes a major strategy in the design for durability of concrete structures. The best solution for excellent corrosion resistance in severe chloride environments is the use of solid stainless steel or probably non-metallic bars, but engineers should be aware of the increased cost of the former in relation to galvanized bars and the structural implications of the latter [39,40]. However, there are many exposure conditions that are severe but not highly chloride contaminated, where surface coatings on steel can be equally effective and provide a long corrosion-free life of steel reinforcement. Metallurgical coatings have, in this respect, some unique advantages, whilst epoxy coatings, like concrete surface coatings, must satisfy a number of stringent specifications if they are to be fully corrosion resistant and effective in practice [41–47]. The focus of this chapter is, however, on galvanized reinforcing bars and the following sections are therefore devoted to this particular type of reinforcing steel only with comparisons to other steel where appropriate. The discussion is confined, in general terms, to those aspects of galvanizing and galvanized reinforcement, a clear understanding of which is essential for their proper and successful incorporation in concrete, avoiding misuse and being aware of their beneficial features as well as their limitations.
2.6. Design Implications of Galvanized Reinforcement 2.6.1. Galvanizing and Galvanized Bars As described in detail in Chapter 4, the process of forming the zinc layer on steel reinforcing bars is carried out by immersion in molten zinc at about 4508C. Both the chemical composition of the steel and the size of the bars control the duration of immersion. In general terms, it takes about 20–120 sec of immersion time to achieve a zinc layer thickness of 80– 200 mm (about 600–1500 g/m2) on bars of 8 – 25 mm diameter. Depending on the severity of exposure conditions, a coating thickness of 80 –200 mm should generally provide adequate corrosion protection. Zinc layers greater than 200 mm thick are not recommended, primarily due to the cracking of the zinc layer when bending the coated bar, although there are now
Design for Durability with Galvanized Reinforcement
43
techniques to apply the coating after the bars have been bent to design specifications and the reinforcement cage fabricated. Further, coatings of greater thickness may result in an inadequate bond between concrete and the coated bar. The hot-dip galvanizing process produces an integral and compact coating that is metallurgically bonded to the steel. When the reaction between iron and zinc is complete, the coating becomes a continuous layer with free zinc and steel at the outer and inner edges, respectively, and a series of iron-zinc alloy layers in between [48]. The outer layer of pure zinc is relatively soft but the underlying alloy layers become progressively harder, sometimes harder than the base steel itself, providing an overall tough coating tightly adhered to the steel substrate. As pointed out earlier, the chemical composition of the steel has a significant influence on the iron– zinc reactions. The constituent that has the greatest influence on these reactions is silicon (the so-called Sandelin Effect) and, at a galvanizing temperature of around 4508C, the thickness and form of the zinc coating is very much influenced by the silicon content of the steel and the duration of immersion. Tests show that micro-alloyed steels, which have generally higher silicon content, are not the best steels for galvanizing [49]. The galvanized layer of such steels tends to be less stable and consistent and does not perform well on bending. Cold-worked bars, on the other hand, appear to be sensitive to the galvanizing process and there is some evidence that there could be greater risk of hydrogen embrittlement with such bars due to pickling operations used to remove scaling prior to zinc coating. Experience now shows that thermomechanically treated steel bars, with a nominal silicon content of about 0.20%, produce the best zinc layers which possess good cohesion and consistency and have a high degree of adhesive bonding during bending [49]. Such bars also possess optimum weldability and ductility properties so that thermomechanical treatment appears to offer the most economic method of producing high-quality hot-rolled reinforcing bars for RC construction.
2.6.2. Cracking of Galvanized Coating One of the major concerns of engineers in using galvanized bars in RC is the cracking of the zinc layer when bending and the consequent loss of adhesion of the coating to the substrate steel. Bend and re-bend tests carried out according to BS449-1988 on galvanized reinforcing bars produced by thermomechanical treatment show that all the bars from 8 to 40 mm diameter performed satisfactorily [50]. The adhesion characteristics of the galvanized coating were then assessed after the bend tests on an arbitrary scale of 1–5, with 1 representing extreme cracking of the coating with flakes dislodged without touching and 5 representing no cracking of the coating. It was found that cracking of the galvanized layer
44
R. Narayan Swamy
occurred on all bar diameters — the 40 mm diameter bars performing worst — and the adhesion rating becoming progressively better as the bar diameter became smaller. In general terms, the coating adhesion rating was: 40 mm bar, 1–2; 32 mm bar, 2 – 3; 16 mm bar, 2 – 3; and 8 mm bar, 3–4. Surprisingly, even the 8 mm bar showed slight to moderate cracking. The precise effect of these cracked coatings on their long-term durability in concrete exposed to seawater or deicing salts is not clear. Fracture at a bend of galvanized high-tensile bars during handling is not unknown [51]. In the particular case concerned, the bars were pre-bent before galvanizing and it was thought that bars bent to a tight radius of 2d (where d is the bar diameter) may have cracked before dipping in zinc. A minimum radius of 3d was found to avoid such failures. Cold-worked deformed hightensile bars have generally low ductility, with a percentage elongation of around 40– 70% of that of a corresponding hot-rolled bar. Such bars, and particularly those of large diameter, may crack during bending and care should be taken to avoid this. And, as pointed out earlier, if cold-worked deformed high-tensile bars are used for galvanizing, adequate precautions need to be taken to avoid the possible risk of hydrogen embrittlement arising from galvanizing operations. Many bend tests show that the degree and intensity of cracking and the width of the cracks in the zinc coating are influenced by the bend radii, the diameter of the bars, the angle of bend and the thickness of the coating. In general, the smaller the bend radius, the larger the cracks and the thicker the coating, the greater the intensity of cracking. Cracking in the coating will invariably occur at right angles to the longitudinal axis of the bar and, if the intensity of cracking causes local debonding between the coating and steel substrate, the durability of the coating may be compromised. In practice, therefore, it is safer to galvanize the bars after bending and this should be considered especially where stirrups are concerned. The corrosion susceptibility of bent bars with cracked coatings has been investigated by storing them at 100% RH and 208C for nearly two years [52]. The galvanized bars showed no corrosion products at the cracks produced in bending and there were indications that zinc oxide formed at the cracks had closed the cracks in the zinc layer and thus prevented local corrosion in the base steel. Companion straight uncoated bars, on the other hand, showed considerable local corrosion and loose rust. Experience shows that the zinc coating has good resistance to abrasion and impact during storage. Tests also show that transportation of the coated bars and concreting operations will also not damage the coating, even if subjected to walking loads prior to concreting. Further, unlike black steels, galvanized steels will not corrode when exposed to moist and warm air.
Design for Durability with Galvanized Reinforcement
45
2.6.3. Mechanical Properties From an engineering point of view, the optimum thickness of the zinc layer is that which maximizes both the zinc-steel bond as well as its corrosion protection. The zinc–iron layer is brittle and cracks are thus bound to occur within this layer, either during bending or when stressed to the yield point. The danger with thick zinc layers is that there may be the risk of debonding at the zinc–iron interface, particularly with larger diameter bars. Hot rolled bars are slightly superior in this respect as they have rough surfaces that can enhance the zinc–iron bond, although within limits. Galvanizing does not adversely affect the static tensile-strength properties of the parent bar such as ultimate strength, yield or proof stress and percentage elongation, provided the process is carried out to specifications. The percentage elongation, however, may be affected if the galvanizing process is defective. There is some evidence that galvanized bars tend to be more brittle and tend to fracture if subjected to re-bending bends already formed and fracture at bends under impact conditions [51]. Pure zinc has low fatigue strength and so the fatigue strength of zinc-coated bars is affected more than their static properties. The limited toughness of galvanized steel may therefore be of some concern for structures designed to withstand earthquake forces and the possibility of failures of highly stressed bars under such conditions should be taken into consideration in design. There is still considerable uncertainty on the effect of galvanizing on the brittleness of bars of different composition and with different degrees of work hardening, despite satisfactory performance of galvanized bars in service. There is some argument that it is preferable to bend cold-worked bars after galvanizing, even at the risk of damaging the zinc coating [51]. There are only limited results available on the fatigue strength of galvanized bars. Fatigue tests reported from Germany [53] show that fatigue cracks start in the zinc layer, and that multiple cracking occurs side by side, which eventually penetrates the zinc– iron alloy layer and finally continues into the steel. A reduction in fatigue strength of about 15% — from about 290 to 250 MPa — was observed. Tests reported from Finland, on the other hand, show that the fatigue strength of certain structural steels may be reduced by as much as 25% as a result of hot-dip galvanizing and that the reduction is independent of the silicon content of steel or the thickness of the zinc layer [52]. In spite of these test results, fatigue tests on galvanized bars extracted from RC beams after being subjected to repeated bending stress loading in a cracked state and partially immersed in chloride solution over a period of about 18 months show that their fatigue strengths were similar to those of uncoated bars without exposure to the corrosion process [52]. These test results now confirm that the fatigue
46
R. Narayan Swamy
strengths of galvanized reinforcement in RC beams remain unaffected by exposure to corrosive environments.
2.6.4. The Galvanized Bar –Concrete Bond Since the bond and bond-slip relationship for galvanized bars in concrete are covered in detail in Chapter 8, this topic is only briefly dealt with here to supplement the information in that chapter and to relate these data to design. The bond between steel and concrete is fundamental to the performance of RC members and the bond strength and bond characteristics of the galvanized bar are therefore of great importance to structural engineers and designers. Most bond tests are carried out on axially loaded members but there is no direct relationship between flexural cracking in a RC beam and direct tension cracking as they are controlled by quite different mechanisms. So bond tests on prisms and cubes are qualitative but, nevertheless, such test results are highly informative, pertinent and appropriate to clarify the bond performance of bars. There is considerable contradiction and widely divergent results reported in published literature on the bond strength of galvanized bars in concrete compared with that of corresponding uncoated bars. Some tests show lower bond strengths for galvanized bars compared with black uncoated rusty bars, whereas others have reported lower bond strengths initially with negligible differences between the two with increase in the age of the concrete [51,54–56]. The loss in bond is often attributed to hydrogen evolution, which will impair the bond between the bar and concrete. The hydrogen evolution results from the complicated chemical reactions occurring between the zinc coating and the cement matrix. The degree of reactivity is influenced by the chemical composition of the cement and the nature of the zinc coating and it is therefore not surprising that pullout tests have shown contradictory results. In general, three factors influence the bond strength of galvanized bars: *
*
*
a reduction in the height of the ribs of deformed bars due to galvanizing. The reduction is generally small, of the order of 10% for bars up to 16–20 mm diameter and probably negligible for thicker bars; an increase in bar roughness by about 10–20%, i.e. from about 25–40 mm to about 30 – 50 mm for galvanized bars; and the evolution of hydrogen at the cement matrix–bar interface.
Of these three parameters, it is the evolution of hydrogen at the bar–cement paste interface that appears to be most critical to the bond behaviour of galvanized bars in concrete. The most important interaction between zinc and cement paste
Design for Durability with Galvanized Reinforcement
47
occurs here and begins immediately after the concreting operations when the zinc reacts with Ca(OH)2 formed as a result of cement hydration. This corrosion reaction is controlled by diffusion processes and results in the evolution of hydrogen and the transformation of zinc into calcium hydroxyzincate, as shown by the following chemical reaction: 2Zn þ CaðOHÞ2 þ 6H2 O ! Ca½ZnðOHÞ3 2 2H2 O þ 2H2 Thus, hydrogen gas is trapped at the cement matrix–galvanized bar interface and this is considered to be responsible for the degradation of bond between steel and concrete. The zinc corrosion reaction is generally active for only a few days — about 6– 10 days — and about 10 mm of the zinc layer is lost by the corrosion processes [57]. Subsequent loss of the zinc layer due to continued corrosion is no more than about 2 mm per year although, in carbonated concrete, this may increase to about 5 mm per year. The corrosion reaction in fresh concrete can be controlled by the presence of the hexavalent chromate ion (CrO22 4 ). A chromate content of 100 – 200 ppm, equivalent to 50– 100 ppm of Cr6þ, is adequate to restrain the corrosion reaction. A possible solution to prevent the problems associated with the evolution of hydrogen is then to use potassium dichromate as additive to the concrete mixing water or to adopt chromate passivation of the bars during galvanizing and prior to their use in concrete. However, there are concerns about health risks involved with the use of chromates and these practices are therefore not much in favour. An alternative is to rely on the presence of naturally occurring chromates in Portland cements to provide this passivation of galvanized reinforcement but, in practice, most Portland cements contain less than half of the required threshold level to provide complete control of hydrogen evolution. There is, however, considerable doubt and scepticism about the extent of the formation of hydrogen and its role on the loss of bond strength between galvanized bars and concrete. Tests reported by Bird [58] show that the hydrogen evolution occurs not at the pure zinc face but at the zinc–iron alloy layer. This implies that the hydrogen evolution may be much more short-lived than previously believed [57]. Bearing in mind that fresh concrete remains in a fluid state for only a few hours, and the fact that the hydration process is, at the very early stages, slow and lethargic, the zinc– cement matrix corrosion reaction appears to be more complex than indicated by the equation above. Since the galvanic cell between the galvanized surfaces and iron is rapidly polarized, it seems more plausible that the hydrogen evolution is much more short-lived than previously thought and it has been suggested that this hydrogen evolution ceases within 1 h [58]. Knowing that iron has a relatively low hydrogen over potential, it appears that it is the zinc– iron alloy or localized imperfections in the zinc coating that are primarily responsible for hydrogen evolution when the fresh concrete comes into
48
R. Narayan Swamy
contact with them. These observations imply that it is not the thickness of the zinc coating that is critical and nor can specifying a minimum thickness of the coating be presumed to eradicate the problem [58]. In practice, therefore, defective galvanizing processes which permit zinc– iron alloys to appear at the surface of the coating, damaged coatings, unrepaired cut-ends of galvanized bars and the use of ungalvanized steel binding wire can all be critical in allowing the evolution of hydrogen at the zinc– bar interface. More recent bond tests appear to confirm that the difference in bond strength between ribbed black bars and galvanized bars is not statistically significant, and that any early-age reductions in bond strength for galvanized bars completely disappear with the age of concrete [59–61]. Overall, these results confirm that galvanized bars can give just as good all-round bond-strength performance as uncoated bars. It is clear that there are complex chemical interactions between zinc/zinc–iron alloys and the cement matrix that lead to hydrogen evolution and the formation of zinc oxide. The formation of zinc oxide and other zinc compounds cause retardation of setting and lead to the slow development of good, strong bonding at the bar –concrete interface at early ages. With time, this early retardation of setting is overcome and a strong bond between the zinc coating and concrete is restored. It is thus clear that the early-age chemical interactions and hydrogen evolution have no visible adverse effects on the bond characteristics of galvanized bars at ages beyond 7 days, and chromate treatment is not necessary to counteract the effects of hydrogen formation [45 – 47,59– 63].
2.7. Durability of Concrete with Galvanized Bars There are a number of factors, some internal and some external, that create conditions favourable for corrosion of steel in concrete to initiate and propagate. These sources of concrete deterioration can be summarized as follows: * * * * * * *
carbonation; chloride and sulphate ion intrusion; atmospheric pollution; environmental conditioning; chlorides and sulphates in mix constituents; freezing and thawing; and expansive reactions like alkali-aggregate reactions.
In many instances, two or more of these sources of deterioration act together and lead to corrosion of steel. There are, however, two major processes through which corrosion proceeds in a RC structure, both through the permeability of concrete. Carbonation
Design for Durability with Galvanized Reinforcement
49
and atmospheric pollutants reduce the pH value of the concrete and thus remove the chemical passivity of the cement matrix surrounding the steel, leading in time to its corrosion. Chloride contamination can also remove this passivity and initiate a form of corrosion that is now recognized to be much more severe and detrimental to the long-term safety, stability and integrity of the concrete structure. Chloride ions can arise from a variety of sources — from the concrete mix constituents, from penetration from marine environment or deicing salts, directly or indirectly through surface water, air- and water-borne chlorides and traffic spray. Although corrosion of steel may not, per se, immediately affect the ultimate load capacity of a RC structural member, once the process starts it is almost impossible to stop (unless, of course, appropriate and always costly remedial measures are undertaken). The immediate effect of steel corrosion is staining, splitting and spalling of the cover concrete due to the enlarged volume of corrosion products, which can lead to loss of serviceability and integrity of the structural member. Carbonation and chloride-ion penetration are thus the two major factors influencing the durability of RC structures [47].
2.7.1. Resistance to Carbonation Apart from causing additional shrinkage, the major effect of carbonation of concrete is to reduce the pH of the pore water in the cement matrix from about 12.6– 13.5 to a value of about 9. In extreme cases, where all the calcium hydroxide is depleted, the value of pH may drop to as low as 8.3. Loss of alkalinity due to carbonation can occur not only due to the penetration of carbon dioxide in the atmosphere but also from sulphur dioxide and other acidic gases emitted from automobile exhaust gases and other sources [64]. In urban areas of major cities, the pH of rainwater is known to be reduced to very low values, sometimes as low as 3.0. When the carbonation front reaches the steel bar –concrete interface, it destroys the thin, protective and passive oxide film formed on the steel surface during the hydration process, and corrosion of steel takes place in the presence of moisture and oxygen. Further, all RC members are cracked at service loads and these cracks provide additional paths for the ingress of CO2 to the steel surface, causing local depassivation. Thus, even though the carbonation front as a whole may not have reached the steel surface, steel corrosion can occur locally at cracks, causing staining, further cracking and spalling of the cover concrete. The depassivation of unprotected carbon steel occurs at pH values below about 11.5. In the presence of chlorides, however, such depassivation can occur at higher pH values. Zinc, on the other hand, is an amphoteric metal and so reacts with both strong acids and bases. The reaction is very severe below pH 6 and above pH 13, but the rate of attack is very slow and the zinc remains passivated in the pH range
50
R. Narayan Swamy
8 – 12.5 due to the complex chemical interactions between zinc and the fresh concrete discussed earlier. Zinc-coated reinforcement can therefore remain passivated to pH values as low as 9.5 and thus offers significantly greater protection for a longer time than black steel against the effects of carbonation in concrete. Further, the corrosion products of zinc in concrete occupy a much lower volume of 5.36 cc per mole of metal consumed compared with that of iron of about 7.8 cc per mole of parent metal consumed, resulting in much lower swelling pressures and reduced cracking of the cover concrete [48]. This reduced destruction of the cover concrete can give RC longer and better electrochemical stability when galvanized steel is used as reinforcement. There are only limited test data reported on the pH of concrete reinforced with galvanized bars and exposed to chloride attack. In the tests reported by the author, test specimens under load were exposed to two regimes — a corrosive tidal zone and an accelerated laboratory cyclic wet and dry regime to simulate the high and low tides [41]. The test specimens contained a 19-mm deformed bar centrally located in prisms with 20, 40 and 70 mm concrete cover. A medium strength concrete with a cement content of 262 kg/m3 and water–cement ratio of 0.55 with about 25 MPa 28-day compressive strength was used in the tests. The test specimens were held in a specially designed reaction frame such that the bars were kept under a constant stress of about 200 MPa. The test specimens were thus cracked with crack widths varying from 0.11 to 0.25 mm for concrete covers of 20– 70 mm. The reinforcing bars were galvanized with a zinc layer thickness of about 70 mm and chromate treated [41]. The pH measurements, after 3 years of natural marine exposure in the tidal zone, showed that, generally, the pH value was more than 12 at a depth of 10 mm from the concrete surface. For all cover depths, the pH of the concrete in the neighbourhood of the reinforcing steel remained in the range 12.6–12.8. At this pH level, the corrosion rate of zinc is a minimum and this fact benefits the performance characteristics of galvanized bars in concrete. On the other hand, these pH values are also sufficiently close to those where small increases in pH begin to give large increases in corrosion rate. However, they are still substantially below the value of 13.35 ^ 0.10, which is generally considered to be the critical threshold for the onset of active corrosion of galvanized steel in alkaline solutions [65]. In any case, pH values of 12.6 –12.8 are unlikely to be detrimental in real practice since pH values in exposed concretes will only decrease with time, not increase. Further, it is also important to note that galvanized steels remain passive in carbonated concrete and the corrosion rate is generally of the same order of magnitude as that in non-carbonated concrete. On the other hand, the corrosion rate of black steel is more than a factor of 10 times greater in carbonated concrete [66]. These data also point to the distinct advantages to be gained by incorporating mineral admixtures in concrete [35].
Design for Durability with Galvanized Reinforcement
51
2.7.2. Resistance to Chloride Attack Chloride-pitting corrosion of steel has been found to be the major form of steel corrosion affecting the service life of many concrete structures, particularly since it can proceed rapidly even in the absence of carbonation. The penetration of chloride ions can locally depassivate steel and promote active metal dissolution. Chlorides react with the calcium aluminates and calcium aluminoferrites in the concrete to form insoluble chloroaluminates and calcium chloroferrites, but the reaction is never complete and some active soluble chloride always remains in solution in the aqueous phase of the concrete [47]. Even in the uncarbonated state, the level of free chloride in the aqueous phase can be very high. The effect of severe carbonation, however, is to break down the hydrates of cement phases and additional chlorides may then be released from the chloroaluminates. Thus, carbonated concrete may generate more free chloride than uncarbonated concrete. A combination of loss of alkalinity due to carbon dioxide and other acidic gases from the environment and the intrusion of chloride ions can thus pose the most destructive threat to the long-term stability of concrete structures [47,67,68]. Since both carbonation and chloride penetration progress due to the porosity and permeability of concrete, control of these two properties through mineral admixtures should enhance the impermeability and resistance to such attack of concrete, as shown in Figs. 1 and 2 [31,35]. However, since control of porosity and permeability also depends on many factors, such as concrete workability, quality control during construction, workmanship and curing, durable and repair-free performance in salt-laden environments can only be assured by a combination of high-quality concrete with a refined pore structure and protected steel [46,47,68]. Although experience shows that galvanized reinforcement can offer significant advantages over black steel in terms of substantial reduction or even total elimination of rust staining and better tolerance of all types of construction weaknesses, the more important concern is the extent to which galvanized bars can resist a chloride environment. In this respect, there is some contradictory data in the literature regarding the effectiveness of galvanized steel in chloridecontaminated concrete. Whilst exposure tests of small-scale laboratory specimens seem to show that galvanized steel may be corroded in heavily chloridecontaminated concrete [69], detailed examination of several bridge decks exposed to chloride salts well in excess of the threshold value needed to induce corrosion of untreated steel, and of structures exposed to severe salt-water environments, have shown no evidence of corrosion or impaired performance of the concrete and little or no structural impairment due to lack of bond [70–72]. Many laboratory corrosion studies show that the conditions of exposure have a significant influence on the results obtained. It is now known that cyclic wet and dry exposure regimes are far more severe on durability properties as they produce
52
R. Narayan Swamy
corrosion rates that are many times higher than those subjected to continuous exposure regimes [41,47,73]. Uncoated black steel, for example, exposed to cyclic salt-water wetting and drying (3-day immersion in 3.5% sodium chloride solution at 408C followed by oven drying for 4 days at 608C) produced an average rate of corrosion of 272 mm per year compared with 14 mm per year in a continuous salt fog exposure regime at 408C and 100% RH at the same chloride concentration [73]. Tests carried out under these exposure regimes in equivalent concrete show the following [73]: *
*
*
Uncoated black steel is highly susceptible to corrosion, even at relatively low chloride levels. Early cracking and disruption of the concrete cover lead to deep pitting and substantial loss of metal and reduction in steel area. Galvanized bars, on the other hand, can withstand exposure to chloride concentrations several times higher than that which causes corrosion of unprotected black steel. These and other studies [74] thus show that the chloride threshold for corrosion of galvanized reinforcement is several times that which can be tolerated by black steel in equivalent concrete and exposure conditions. Galvanized coating can provide protection to the base steel of about 4–5 times the period for the onset of corrosion of black steel in a similar concrete and exposure regime. Tests with cracked concrete and damaged coating confirm such protective ability for the galvanized steel [75].
Many of the studies reported in the literature have, however, been carried out on uncracked test specimens. Although there is no direct relationship between crack width and corrosion [76], many engineers would instinctively know that crack distribution and cover to steel will have a significant influence on the overall corrosion state of a RC member. Since the cover to steel influences the crack width, cracking and cover should have a profound influence on corrosion of RC structures. However, the role of crack width cannot be totally ignored since it not only helps to form local corrosion cells but also controls the depth of pitting at a given location [41,46,47]. In one of the comprehensive corrosion studies reported [41,46,47], concrete test specimens containing a centrally located 19-mm diameter hot-rolled deformed bar were exposed to corrosive environments whilst under load in a special reaction frame designed to ensure a constant stress in the steel of about 200 MPa. Three cover depths of 20, 40 and 70 mm were used and, at the steel stress level of 200 MPa, the surface crack widths on the concrete specimens were: 0.11– 0.14 mm at 20 mm cover, 0.16– 0.18 mm at 40 mm cover, and 0.22– 0.25 mm at 70 mm cover.
Design for Durability with Galvanized Reinforcement
53
The uncoated bars were used in the as-received condition after cleaning; the galvanized bars had a zinc coating thickness of about 70 mm and were chromate treated. Chromating does not affect the electrochemical performance of galvanized bars [75] so these results are just as valid for galvanized bars without chromating. Two exposure regimes were used — a natural marine exposure and an accelerated cyclic wet and dry laboratory test regime. In the former, the test specimens were exposed in a corrosive tidal zone; the seawater at this site had a pH of 8.1 and contained 18,100 ppm chloride ions. In the accelerated corrosion test, the cyclic wet and dry test regime simulated high and low tides and consisted of 6 h immersion in seawater at 608C and 6 h drying in air at ambient temperature. The seawater had an average chloride content of 18,000 ppm, sulphate content of 2,520 ppm and an average pH of 7.85. It is estimated that 1-year’s accelerated exposure test is equivalent to about 10 –15 years of natural exposure performance. The concrete used in these tests was of medium strength with a cement content of 262 kg/m3 and a water – cement ratio of 0.55, giving a 28-day compressive strength of about 25 MPa. Some of the major results arising from this study are discussed below.
2.7.2.1. Chloride Penetration Table 1 summarizes the chloride penetration results after the exposure tests. These and other data [46] show many interesting aspects of chloride infiltration into concrete. Apart from the crack distribution and concrete cover, the size of the section also has a distinct influence on chloride penetration into concrete. At cover depths of 10–20 mm, there will always be high concentration of chloride ions and this will Table 1: Chloride ion penetration into concrete after exposure tests. Distance from surface (mm)
Tidal exposure: 3 years 10 20 40 Accelerated exposure: 1 year 10 20 40
Chloride concentration (ppm) at cover in mm 20 mm
40 mm
70 mm
8500 10,000 –
7500 7000 7500
6000 5000 4000
8200 9000 –
8000 6000 , 4000
5000 3500 2000
54
R. Narayan Swamy
Table 2: Frequency of red rust formation — accelerated test. Type of bar
Cover (mm)
Frequency of rust (%) 1 year
3 years
Uncoated
20 40 70
80 25 8
100 98 75
Galvanized
20 40 70
15 7 5
72 8 6
increase with time in a marine environment and where external sources of chlorides are regularly present over a long period of time. At such cover depths, even the best corrosion protection will eventually begin to deteriorate. Ferritic stainless steels with low carbon and medium chromium content have not survived in such environments; even the austenitic, chromium–nickel containing steels (AISI 300 series) have shown signs of pitting corrosion in concrete with high levels of chloride [69]. In addition to high-quality concrete and steel protection, uniform and adequate depths of cover are thus very critical in salt-laden aggressive exposure conditions.
2.7.2.2. Rust Formation Tables 2 and 3 summarize the visually observed and measured red rust covering the uncoated and galvanized bars in the accelerated and natural exposure regimes, respectively. These data clearly reveal the superior performance of the galvanized bars in highly corrosive environments. The data also show that a 20-mm cover is inadequate to protect the galvanized bar under these conditions. Table 3: Frequency of red rust formation — natural exposure test. Type of bar
Cover (mm)
Frequency of rust (%) 1 year
Uncoated
20 40 70
95 10 2
Galvanized
20 40 70
Negligible Negligible Negligible
3 years 46 38 8 1.5 1.0 Negligible
Design for Durability with Galvanized Reinforcement
55
2.7.2.3. Pitting Depth The tests showed that the distribution of pitting generally corresponded to the distribution of cracks, as shown in Figs. 3 and 4. It is clear that even a 70-mm cover to steel will not give protection to the uncoated bar under very aggressive exposure conditions. With the galvanized bar, the maximum measured pitting depths were 2.02, 1.08 and less than 1.0 mm at cover depths of 20, 40 and 70 mm, respectively. At 40-mm cover, many pits were about 0.5 mm deep, whereas at 70 mm cover almost all the pits were less than 0.5 mm deep. From Tables 2 and 3 and Figs. 3 and 4, about 10% red rust and pit depths of 0.5 mm and less at cover depths of 40 and 70 mm may appear to be unacceptable degrees of corrosion in a concrete structure. It must, however, be borne in mind that the concrete used in these tests was designed to be of poor quality, with cement content of 262 kg/m3 and a water– cement ratio of 0.55. Further, the accelerated corrosion tests were designed to be highly favourable to the propagation of corrosion. It is estimated that 1 year’s accelerated exposure is equivalent to about 10 –15 years of natural exposure performance. Thus, the formation of 10% red rust and 0.5 mm deep pits should be related to a life of 25– 30 years in real life compared with rust formation of 10–40% at cover depths of 40 – 70 mm in 3 years for the uncoated bars — all in poor-quality concrete.
2.7.3. Long-Term Field Tests Mu¨ller has reported long-term field tests of RC beams with and without galvanized bars that were stored uncovered in the open air for 10 years [77]. The concrete in the beams had a cement content of 270 kg/m3 and the cement had a chromate concentration of 6.6 ppm Crþ6. The water –cement ratio for the concrete was 0.6. The zinc coating on the galvanized bars was 120 mm. Prior to field exposure, the beams were subjected to a carbonation process in air with 2% CO2 for a year, after which the carbonation depth reached a depth greater than the cover of 10 –20 mm. Some beams had chlorides of 0.5 and 2.0% by mass of cement. It was found that the zinc reaction in the fresh concrete caused a loss of coating thickness of 12– 20 mm and there was evidence of hydrogen evolution at the galvanized bar– concrete interface. In the tests, non-galvanized bars showed initiation of corrosion after 2 years of field storage. After 10 years of exposure, non-galvanized bars showed about 50% corroded surface, whereas the galvanized bars showed only negligible corrosion of about 5%. The loss of zinc after 10 years’ exposure varied from 30 to 50 mm. It was also found that uncoated bars in contact with the coated bars were protected by the zinc coating — thus, the uncoated bars showed reduced corrosion whilst the galvanized bars showed greater corrosion.
56
R. Narayan Swamy
Figure 3: Concrete cracking and depth and distribution of pitting in uncoated bars after 24 months of accelerated corrosion test.
Design for Durability with Galvanized Reinforcement
57
Figure 4: Variation of depth and distribution of pitting with concrete cracking in galvanized bars after 24 months of accelerated corrosion test.
The static and fatigue strength of the bars after 10 years of field exposure was only reduced by about 5%. The uncoated bars in beams with 2% chlorides showed 40% of corrosion whereas the corrosion covered only about 15% of the surface of galvanized bars. After 6 years of exposure, the total loss of zinc was 55 mm, while uncoated bars had corrosion pits of 0.7 mm. The loss of zinc coating of 55 mm in carbonated concrete compares with a loss of 90 mm in carbonated concrete reported by Treadaway et al. [69]. The results of these studies confirm the findings reported earlier in the overall better performance of galvanized bars in carbonated and chloride-contaminated concrete. Further information in this area is reviewed by Yeomans in Chapter 6.
58
R. Narayan Swamy
2.7.4. Engineering Implications From the results presented above, both cracking and depth of cover are seem to be the predominant factors influencing the initiation and progress of corrosion. These two factors are, however, inter-related in a contradictory manner — the larger the cover depth, the fewer the number of cracks but the greater their width. Thus, for the size of test specimens used here [47], at 70-mm cover, the number of cracks was about two or three whereas, at 20-mm cover depth, the number of cracks varied from 8 to 10. However, the maximum crack width at 70-mm cover was about twice that at 20-mm cover. In an aggressive environment, the width of the crack in a RC structure is critical as the amount of water, air and other corrosive elements intruding into the concrete is related to the width of the crack. On the other hand, with larger cover depths, the number of critical crack widths is less and limited. Examination of the test specimens at various stages [47] also revealed that crack widths of 0.10 –0.15 mm can be critical so far as intrusion of aggressive elements and initiation of corrosion are concerned. In addition, rust was also found to occur at locations without cracks, even in test specimens with a large cover. Another significant result observed from the tests was that, although the penetration of chloride ion decreased with distance from the concrete surface, there was some indication of increased chloride-ion concentration near the bar surface for cover depths up to 40 mm [46]. So crack width cannot be totally excluded from corrosion activity and a balance between crack spacing and crack width needs to be achieved to minimize corrosion, and in a chloride-laden environment, a reinforcement cover of 50– 70 mm appears to achieve this optimum. Apart from cover and cracking, the quality of concrete is also a major and significant factor influencing the rate and extent of steel corrosion [35,41,76,78]. At water – cement ratios of 0.55, even 28-day water curing will be inadequate to block the continuous capillary pore system of the gel with hydration products [79]. The main cause of concrete permeability is the presence of inter-connected capillary pores and, at w/c ratio of 0.5 and above, such a capillary pore system is bound to exist, bearing in mind the practical realities of site curing. Thus, cracking, cover and concrete quality are the major interactive and inter-related parameters that need to be carefully considered before translating laboratory test data into field practice [46]. A critical evaluation of all the test data presented above confirms that the protection afforded by galvanizing of steel against the two major threats of carbonation and chloride attack causing corrosion damage is a multi-faceted process. The superior benefits of galvanizing steel over uncoated black steel are: *
Zinc is only slightly attacked in solutions of pH 7–11. Thus zinc-coated bars can remain passivated to lower pH values than black steel and the pH of concrete in all exposure conditions can only decrease, not increase.
Design for Durability with Galvanized Reinforcement
*
*
*
*
59
Zinc-coated bars possess much higher tolerance against chloride levels compared with uncoated steel. There is no long-term alkali-induced corrosion of zinc in concrete although, under certain conditions, white rust may appear on the zinc surfaces. Zinc-coated bars provide galvanic or sacrificial protection to the steel substrate during the dissolution of the zinc-alloy layers of the coating. Zinc-coated bars give cathodic protection when the steel substrate is exposed and the coating is completely lost locally.
Thus, although galvanizing cannot give, and should not be expected to give, complete and permanent protection against continued chloride attack, galvanized bars can give a cost-effective enhanced durable service life through a substantial increase in initiation time and in protection time against corrosion. The benefits of galvanized reinforcement on design service life can be schematically represented by Fig. 5, which is an extension of the model proposed by Yeomans [80]. One of the major factors that contribute to the significant delay of the onset of corrosion of the base steel is the fact that galvanizing provides a metallurgically alloyed coating of consistent product quality that is highly resistant to damage during transportation, storage, site handling and concreting operations. It is this
Figure 5: Schematic representation of design service life of concrete structures with uncoated and galvanized bars.
60
R. Narayan Swamy
ability to remain intact and protective without damage or corrosion that enables the coating to act as an effective barrier to external attack and keep the cover concrete undamaged from staining, cracking and spalling. In many practical cases of RC construction, total protection of steel is not needed and galvanizing offers a sure and cost-effective solution to many problems, giving rise to lack of durability of reinforced and pre-stressed concrete structures. It should, however, be borne in mind that the use of galvanized bars should not be an excuse or a substitute for lack of workmanship and quality control, attention to detail, highquality concrete, adequate depth of cover and sound engineering design. It should also never be forgotten that protection of steel is only one aspect, but an important one too, in the chain of events that lead to deterioration and destruction of concrete structures.
2.8. In Situ Performance of Galvanized Reinforcement In spite of the inherent advantages of galvanizing as a means of corrosion protection of steel in concrete, there is considerable confusion in the minds of engineers regarding the long-term stability and durability of galvanized bars in concrete exposed to aggressive salt-laden environments. Part of this lack of confidence arises from conflicting and contradictory laboratory test data and part from the lack of observation of basic structural detailing principles when galvanized reinforcement is used in practice. In addition, there are various laboratory electrochemical studies reported in literature on the chemical reactions and corrosion of zinc in solutions supposedly representing the chemical environment in concrete [45,65]. Many of these studies are highly significant as they clarify the nature of the basic chemical reactions of pure zinc, but aqueous hydroxide solutions or pure acids and alkalis are unrealistic representations of the real nature of concrete. Concrete remains in the plastic state for only a few hours. Once it has hardened, the amount of free moisture within is progressively and drastically reduced by cement hydration and drying. Ionic diffusivity in concrete will thus be fundamentally different to that of zinc immersed in liquid solutions. Further, both carbonation and chloride attack are time-dependent activities involving both chemical interactions and physical processes, and concrete contains many compounds whose interaction with zinc is not clearly established. Extreme care should therefore be exercised in extrapolating and translating into practice laboratory corrosion results obtained from short-term tests or test in solutions simulating the liquid phase of concrete or mortar/concrete prism tests with covers restricted to 10 –20 mm [45,65]. Resistance to chloride attack, for example, depends on many factors, such as cover to steel, quality of concrete,
Design for Durability with Galvanized Reinforcement
61
nature and severity of exposure conditions and the stress state of the test specimens. Conclusions derived from tests in isolation without reference to these significant and influential parameters can be misleading and very unrealistic. Corrosion of galvanized steel in exceptionally severe corrosive conditions has been reported [81] but such instances are few and, generally, many types of metals are likely to corrode in such situations. Lack of careful attention to detail and quality control on site invariably leads to local corrosion spots in the same way as lack of workmanship and good concreting practice result in concrete deterioration. In aggressive environments, local corrosion of galvanized bars has been reported — these have arisen from operations such as welding after galvanizing and being left unrepaired, or galvanized bars bent or cut on site without repair to damage [82]. In contrast, there are a large number of cases of successful field performance of galvanized steel in aggressive exposure conditions reported in the literature [45,51,70,71,82– 85]. The sound condition of galvanized steel after 54 years of exposure in a marine environment has been reported by Halstead and Ridge [83]. Allan [82] has reported the good condition of galvanized reinforcement found in the Old Bus Garage in Hamilton, Bermuda when demolished after 45 years of service and exposure to the extremely aggressive marine environment there. The concrete-reinforcement interface was found to be dense and showed no signs of any zinc– concrete reaction. These observations from Bermuda are also covered in detail in Chapter 7. The islands of Bermuda have one of the worst marine-exposure conditions known in the world. Unprotected reinforced and prestressed steel used in concrete structures with 70-mm cover have often shown signs of corrosion within 3 years of construction [82]. Sound engineering design and careful attention to detailing with high-quality workmanship have, on the other hand, shown that galvanized reinforcement can give durable service life, even in such extremely aggressive salt-laden environments [82,85]. These results emphasize the need for sound holistic design [86] and confirm the unmistakable ability of galvanizing to resist chloride attack in severe marine environments, including the splash zone, and satisfy the design-life criteria. The most comprehensive field study on the performance of galvanized steel in concrete structures was reported by Stark and Perenchio [70] and Stark [71]. Figs. 6 and 7 show the measured pH values from two bridges in the United States [70]. Fig. 6 refers to dual three-span bridges on I35 carrying traffic over Long Dick Creek near Ames, Iowa. The concrete in the bridges had a cement content of 420 kg/m3, a water – cement ratio of 0.40– 0.41 and 5.2–6.2% air entrainment. Fig. 7 refers to a composite steel –concrete bridge in Montpelier, Vermont. The concrete in the deck had cement contents of 360–390 kg/m3, a water –cement ratio of 0.44 and 6% air entrainment. The results show pH values of 11.2–12.4 in the Iowa bridge after 7 years of exposure and 12.4 –12.7 in the Vermont Bridge
62
R. Narayan Swamy
Figure 6: Chloride concentration and pH values in galvanized bridge deck (135, lowa) after 7 years exposure to deicing salts.
3 years after construction. These data confirm that pH itself is never critical to the stability of galvanized bars in concrete, although many laboratory tests in saturated solutions show otherwise. Bearing in mind that pH values in concrete structures will only decrease with time, not increase, and the fact that galvanized steel remains passive in carbonated concrete [45,46], the superior performance of galvanized bars in real environments is not surprising.
Figure 7: Chloride concentration and pH values in bridge deck with galvanized top mat (Montpelier, Vermont) after 3 years.
Design for Durability with Galvanized Reinforcement
63
The field performance of galvanized steel in seven concrete structures varying in age from 7 to 23 years and exposed to severe marine environment in the islands of Bermuda has been reported [71]. The average corrosion layer measured in these structures varied from 2 to 8 mm with one value of about 13 mm. Over 90% of the zinc coating was considered to be remaining on the bars. The depth of corrosion of the zinc coating was thus extremely low. These results confirm other data presented earlier that galvanized steel is able to tolerate chloride concentrations well above the 275– 325 ppm (0.66– 0.78 kg/m3 of concrete) considered necessary to corrode unprotected steel.
2.9. Concluding Remarks Experience during the last 40– 50 years has shown that current design methods and design specifications, even if fully implemented, do not provide for adequate resistance of RC structures against carbonation and chloride-induced corrosion. Premature deterioration of concrete structures has become widespread and worldwide and is still unacceptably high. In chloride-contaminated environments, the time-dependent and interactive effects of load, exposure and climatic changes initiate cumulative deterioration mechanisms, which become an overall synergistic process, and a complex combination of many individual mechanisms, the exact role, effect and contribution of each of which to the totality of damage is not fully known. Corrosion of steel reinforcement thus continues to represent the single major cause of deterioration of RC structures. From engineering, economic and societal points of view, such a state of affairs cannot simply be tolerated. Further, there are now strong, unquestionable and undeniable environmental and sustainability issues that demand that the materials we use and the structures we design and build are eco-friendly, and have the specified durable design service life in the environments in which they have to breathe, live and carry loads. Galvanized steel has excellent inherent qualities to provide high corrosion resistance in carbonated concrete and concrete contaminated by chloride ions. There are, however, some laboratory and field studies that report pessimistic evaluations of the performance of galvanized steel whilst extensive assessment of their in situ performance contradicts such conclusions. The focus of this chapter has been to present a critical evaluation of all such laboratory and field studies and the field experience of the performance of galvanized steel in salt-laden aggressive environments. It has been shown that care should be exercised in translating into practice the results of electrochemical studies of pure zinc or galvanized steel in solutions simulating the chemical environment of concrete as well as the results of tests on mortar and concrete prisms with small cover depths. Cracking, cover
64
R. Narayan Swamy
and the quality of concrete are the three major interactive and inter-related parameters which influence and control the corrosion of steel reinforcement in concrete, but cover to steel is probably the most critical single factor that will preserve the electrochemical stability of reinforcing steel in concrete. Many tests confirm that galvanized steel, unlike unprotected steel, has similar corrosion rates in carbonated concrete as in non-carbonated concrete. There is also evidence that the calcium hydroxyzincate provides a dense and passivating layer on the zinc surface. It is clear that the role of hydrogen evolution in early-age concrete, and the implications of its effects on the initial retardation of setting and low bond strength, have been exaggerated. There is conclusive evidence that the bond between concrete and galvanized bars is not a problem once the concrete has matured. Further, the behaviour under load at both service and ultimate loads of beams with galvanized bars is no different to that of beams with uncoated steel. There is thus no bond weakness in a RC structural element with galvanized bars. Indeed, all the test evidence confirms that there is no need for chromate treatment either to improve bond or to counteract the effects of hydrogen evolution. In the present state of our knowledge of the mechanical properties of galvanized steel and the structural behaviour of concrete reinforced with galvanized steel, and of our understanding of the chemical interactions at the zinc–cement matrix interface, galvanized steel can be used in RC structures with confidence and reliability to enhance their field performance and durable service life in aggressive exposure conditions. Indeed, the benefits that can be derived by their use will far outweigh the marginal increased costs incurred by the process of galvanizing the steel reinforcement. It should, however, be clearly understood that the use of galvanized bars should not be an excuse or a substitute for poor-quality concrete or lack of workmanship. The concrete cover to steel and the quality of concrete finally determine the rate of corrosion deterioration of steel in concrete. As a result, attention to detail, adequate depth of cover and sound engineering design with high-quality concrete are essential to ensure long and durable service life. The state of the infrastructure in a country is a reflection of its economic progress and stability and, indeed, of the quality of its peoples lives. Over the decades, time, environment, neglect and human conflict have reduced the infrastructure in many countries to a condition of disrepair and sometimes nonexistence. This situation is currently heavily aggravated by the damage and destruction inflicted by storms, floods, landslides and mudslides and other climatic changes brought about by global warming. Galvanizing steel reinforcement can reduce the present extent of structural damage by more than 50% and, thereby,
Design for Durability with Galvanized Reinforcement
65
contribute to a better utilization of our material and energy resources and simultaneously enhance the quality of life of all peoples of the world.
References [1] Mehta, P. K. (1994). Concrete technology at the cross-roads — problems and opportunities. In: P.K. Mehta (Ed.), (pp. 1 –30). ACI publication SP-1144. [2] Rasheeduzzafar, Dakhil, F. H., & Al-Ghatani, A. S. (1985). Corrosion of reinforcement in concrete structures in the Middle East. Concrete International, 7, 9, 48–55. [3] Ingvarsson, H. & Westerberg, B. (1986). Operation and maintenance of bridges and other load-bearing structures, Publication No. 42. Swedish Transport Research Board, Stockholm, Sweden. [4] National Materials Advisory Board. (1987). Concrete durability — a multibillion dollar opportunity, Publication No. NMAB-437. National Academy of Sciences, Washington, DC, 94 p. [5] Wallbank, E. J. (1989). The performance of concrete bridges. Department of Transport, HMSO, London, 96 p. [6] Mehta, P. K. (1991). Durability of concrete — fifty years of progress. In: V. M. Malhotra (Ed.), (pp. 1–31). ACI Publication SP-126. [7] Gjorv, O. E. (1992). Durability of concrete structures in the ocean environment. Proceedings FIP Symposium on Concrete Sea Structures, September, London, 141–145. [8] Swamy, R. N., Hamada, H., & Laiw, J. C. (1994). A critical evaluation of chloride penetration into concrete in the marine environment. In: R.N. Swamy (Ed.), Corrosion and corrosion protection of steel in concrete (pp. 404–419). Sheffield Academic Press. [9] Swamy, R. N. (1974). Behaviour of high alumina cement concrete under sustained loading. Proceedings, Institution of Civil Engineers, 57 (Dec), 651–671. [10] Yamada, Y., Oshiro, T., Tanikawa, S., & Swamy, R. N. (1997). Field evaluation of an acrylic rubber protective coating system for RC structures. In: V. M. Malhotra (Ed.), Durability of concrete, vol. 1, (pp. 23 –40). ACI Publication SP-170. [11] Hobbs, B., Swamy, R. N., & Roberts, M. (1996). Corrosion performance of steel plated RC beams after long-term natural exposure. In: R. N. Swamy, & R. Gaul (Eds), Repair and strengthening of concrete members with adhesive bonded plates, (pp. 101–126). ACI Publication SP-165. [12] Swamy, R. N. (1990). Alkali-silica reaction and concrete structures. Structural Engineering Review, 2, 89 –103. [13] Swamy, R. N. (1996). Assessment and rehabilitation of AAR — affected structures. In: A. Shayan (Ed.), Proceedings, 10th International Conference on Alkali — Aggregate Reaction in Concrete, pp. 68–83. [14] Rahman, A. W. (1997). Engineering and durability properties of concrete in the qatar environment, PhD Thesis. University of Sheffield. [15] Al-Amoudi, O. S. B., & Maslehuddin, M. (1993). The effect of chloride and sulphate ions on reinforcement corrosion. Cement and Concrete Research, 23, 1, 139–146. [16] Al-Amoudi, O. S. B., Rasheeduzzafar, Maslehuddin, M., & Abduljauwad, S. N. (1994). Influences of sulfate ions on chloride-induced corrosion in Portland and blended cement concretes. Cement, Concrete and Aggregates, 16, 1, 3–11.
66
R. Narayan Swamy
[17] Yonezawa, T., Ashworth, V., & Proctor, R. P. M. (1898). The mechanism of fixing Cl2 by cement hydrates resulting in the transformation of NaCl to NaOH. In: K. Okada, S. Nishibayashi, & M. Kawamura (Eds), Proceedings, 8th International Conference on Alkali-Aggregate Reactions, Tokyo, pp. 153–160. [18] Barton, K. (1976). Protection against atmospheric corrosion: theories and methods. Wiley, London, vol. 3. [19] Jones, D. A. (1996). Principles and prevention of corrosion. Prentice Hall, Upper Saddle River. [20] Likens, G. E., Wright, R. F., Galloway, J. N., & Butler, T. J. (1979). Acid Rain. Scientific American, 241, 39 –47. [21] Uhlig, H. H., & Revie, R. W. (1985). Corrosion and corrosion control. Wiley, New York. [22] Johansson, L. G., & Vannerberg, N. G. (1981). The atmospheric corrosion of unprotected carbon steel — a comparison between field study and laboratory test. Werkstoff Und Korrosien, 32, 265–268. [23] Fyfe, D. (1976). Metal/environment reactions. In: L. L. Shreir (Ed.), Corrosion. NewnesButterworths, London. [24] Brown, P. W., & Masters, L. W. (1982). Factors affecting the corrosion of metals in the atmosphere. In: W.H. Ailor (Ed.), Atmospheric corrosion. Wiley, New York. [25] Maslehuddin, M., Rasheeduzzafar, Al-Amoudi, O. B. S., & Al-Mana, A. I. (1994). Concrete durability in a very aggressive environment. In: P. K. Mehta (Ed.), Concrete technology past, present, and future, (pp. 191–211). ACI Publication SP-144. [26] Awny, R. H., Eid, W. K., & Al-Mutairi, N. M. (1996). Chemical concentrations in soil and ground water in Kuwait. Kuwait Journal of Science and Engineering, 23, 2, 233– 248. [27] Swamy, R. N. (1994). Alkali-aggregate reaction — the bogeyman of concrete. In: P. K. Mehta (Ed.), Concrete technology — past, present and future, (pp. 105–139). ACI Publication SP-144. [28] Swamy, R. N. (1996). High performance and durability through design. In: P. Zia (Ed.), High performance concrete, (pp. 209–230). ACI Publication SP-159. [29] Suryavanshi, A. K., & Swamy, R. N. (1997). An evaluation of controlled permeability formwork for long-term durability of structural concrete elements. Cement and Concrete Research, 27, 7, 1047 –1060. [30] Tanikawa, S., & Swamy, R. N. (1994). Unprotected and protected concrete on-site chloride penetration with time in an aggressive environment. In: R. N. Swamy (Ed.), Corrosion and corrosion protection of steel in concrete, (pp. 1069 – 1080). Sheffield Academic Press. [31] Swamy, R. N. & Laiw, J. C. (1995). Effectiveness of supplementary cementing materials in controlling chloride penetration into concrete. In: V. M. Malhotra (Ed.), vol. 2, (pp. 657–674). ACI Publication SP-153. [32] Read, J. A. (1989). FBECR — The need for correct specification and quality control. Concrete, 23, 8, 23–27. [33] The State of the Nation (2000). New Civil Engineer, 25 May 2000, 15–17. [34] ASCE News (1998). American Society of Civil Engineers, Washington, USA, 24 Sept. 1998. [35] Swamy, R. N. (1997). Design for durability and strength through the use of fly ash and slag in concrete. In: V. M. Malhotra (Ed.), Advances in concrete technology, (pp. 1–72). ACI Publication SP-171.
Design for Durability with Galvanized Reinforcement
67
[36] Swamy, R. N. (1994). Design — the key to concrete material durability and structural integrity. International Conference on RC Materials in Hot Climates, United Arab Emirates University, vol. 1, (pp. 3–36). [37] Swamy, R. N. (1995). A global design/management strategy to enhance corrosion-free durable service life of concrete constructions. Journal of Chinese Corrosion Engineering, 9, 4, 205– 214. [38] Swamy, R. N. (1999). Design of durable service life — linchpin for sustainability in concrete construction. High performance concrete — performance and quality of concrete structures, ACI Publication SP-186 (pp. 765–788). [39] McDonald, D. B., Sherman, M. R., Pfeifer, D. W., & Virmani, Y. P. (1995). Stainless steel reinforcing as corrosion protection. Concrete International, 17, 5, 65–70. [40] Swamy, R. N., & Aburawi, M. (1997). Structural implications of using GFRP bars as concrete reinforcement. Non-metallic (FRP) Reinforcement for Concrete Structures, 2, 503–510. [41] Satake, J., Kamakura, M., Shirakawa, K., Mikami, N., & Swamy, R. N. (1983). Long term corrosion resistance of epoxy coated reinforcing bars. In: A.P. Crane (Ed.), Corrosion of Reinforcement in Concrete Construction (pp. 359–377). Ellis Horwood Ltd, Chichester. [42] Swamy, R. N. & Koyama, S. (1987). Epoxy coating of reinforcing steel for corrosion protection. 4th International Conference on Durability of Building Materials and Components, Singapore, vol. 2, pp. 647–653. [43] Swamy, R. N. & Koyama, S. (1988). Epoxy Coated Bars — A panacea for steel corrosion in concrete. UK Corrosion 88, vol. 3, pp. 197–209. [44] Swamy, R. N., Koyama, S., Arai, T. & Mikami, N. (1988). Durability of steel reinforcement in marine environment. ACI Publication SP-109 (pp. 147– 161). [45] Swamy, R. N. (1991). Insitu behaviour of galvanized reinforcement. In: J. M. Barker, P. J. Nixon, A. J. Majumdar, & H. Davies (Eds), Durability of building materials and components (pp. 299–312). E&FN Spon, London. [46] Swamy, R. N. (1990). Resistance to chlorides of galvanized rebars. In: C. L. Page, K. W. J. Treadaway, & P. B. Bamforth (Eds), Corrosion of reinforcement in concrete (pp. 586–600). Elsevier Applied Science, London. [47] Swamy, R. N. (1992). Durability of rebars in concrete. In: J. Holm, & M. Geiker (Eds), Durability of Concrete, (pp. 67 –98). ACI Publication SP-131. [48] Yeomans, S. R. (1995). Galvanized steel reinforcement — a perspective view. In: G. Singh (Ed.), Real world concrete, (pp. 57 –70). RN Swamy Symposium, CANMET/ACI. [49] Plowman, K. (1991). Modern methods of production of steel reinforcement bars and some effects on galvanizing properties, Seminar on Galvanized Bars, University of Sheffield, April 1991. [50] Sandberg (1990). The assessment of galvanized Tempcore reinforcing bar according to BS4449-1988, Test Report, Messers Sandberg, Consulting Engineers, London, March 1990. [51] Flint, A. R. (1991). Galvanized rebar, examples of use in buildings, Seminar on Galvanized Bars, University of Sheffield, April 1991. [52] Sarja, A., Jokela, J. & Metso, J. (1984). Zinc-coated concrete reinforcement, Research Report 306, Technical Research Centre of Finland, Esposo, 92 p. [53] Mang, R., & Mu¨ller, R. H. (1982). Untersuchungen zur anwendbarkeit feuerverzinkter bewvehrung im stahlbeton-bau. Stahl und Eisen, 18, 82, 889–894.
68
R. Narayan Swamy
[54] Robinson, K. E. (1956). The bond strength of galvanized reinforcement, Technical Report TRA/220. Cement and Concrete Association, London, 6 p. [55] Hofsoy, A., & Gukild, I. (1969). Bond studies on hot dip galvanized reinforcement in concrete. ACI Journal, 66, 174–184. [56] Mang, R. & Mu¨ller, H. H. (1980). Einflus einer feuerverzinkung auf die mechanischen eigenschaften von unbehandeltem betonrippenstahl, Lehrstuhl fu¨r Massivbau, Tech. Univ. Munchen, Report No. 1009, p 43. [57] Nu¨rnberger, U. (1990). Korrosion und korrosionsschutz der bewehrung im massivbau, Deutscher Ausschus fu¨r Stahlbeton, Heft 105, Berlin, 116 p. [58] Bird, C. E. (1962). Bond of galvanized steel reinforcement in concrete. Nature, 194, 4830, 798 p. [59] Swamy, R. N., Unpublished data. [60] Kayali, O., & Yeomans, S. R. (2000). Bond of ribbed galvanized reinforcing steel in concrete. Cement and Concrete Composites, 22, 6, 459–467. [61] Kayali, O., & Yeomans, S. R. (1995). Bond and slip of coated reinforcement in concrete. Construction and Building Materials, 9, 4, 219–226. [62] Koch, R., & Wohlfahrt, R. (1988). Effect of admixtures in concrete on the bond behaviour of galvanized reinforcing bars. Betonwerk þ Fertigteil 2 Technik, 54, 3, 64–70. [63] Yeomans, S. R. & Ellis, D. R. (1992). Further studies of the bond strength and slip characteristics of galvanized and epoxy coated steel reinforcement in concrete, Progress Report No. 5, ILZRO Project ZE-341, Dec 1992. International Lead Zinc Research Organization, NC, USA. [64] Chandra, S. (1990). Influences of pollution on mortar and concrete, Document D6: 1990. Swedish Council for Building Research, 83 p. [65] Macias, A., & Andrade, C. (1983). Corrosion rate of galvanized steel immersed in saturated solution of Ca(OH)2 in the pH range 12 –13.8. British Corrosion Journal, 18, 82–87. [66] Maahn, E., & Sorenson, B. (1986). The influence of microstructure on the corrosion properties of hot-dip galvanized reinforcement in concrete. Corrosion, 42, 187–196. [67] Suryavanshi, A. K., & Swamy, R. N. (1997). Well-crystallized and amorphous calcite in concrete slabs exposed to long-term atmospheric carbonation. Advances in Cement Research, 9, 35, 115– 125. [68] Swamy, R. N., & Suryavanshi, A. K. (1998). Durability of blended cement concrete structural elements of higher water-binder ratio against chloride and carbonation attack. Arabian Journal for Science and Engineering, 23, 1B, 17 –32. [69] Treadaway, K. W. T., Cox, R. N., & Brown, B. L. (1989). Durability of corrosion resisting steels in concrete. Proceedings, Institution of Civil Engineers (UK), Part 1, 86, 305–331. [70] Stark, D. & Perenchio, W. (1975). The performance of galvanized reinforcement in concrete bridge decks, Final Report Project No. 2E-206, Oct. 1975. International Lead Zinc Research Organization, NC, USA. [71] Stark, D. (1978). Galvanized reinforcement in concrete containing chlorides, Final Report Project No. 2E-247, April 1978. International Lead Zinc Research Organization, NC, USA. [72] Cornet, I. & Bresler, B. (1981). Galvanized steel in concrete: literature review and assessment of performance, Galvanized Reinforcement in Concrete — II, May 1981. International Lead Zinc Research Organization, NC, USA. [73] Yeomans, S. R. (1994). Performance of black, galvanized and epoxy-coated reinforcing steels in chloride-contaminated concrete. Corrosion, 50, 1, 72 –86.
Design for Durability with Galvanized Reinforcement
69
[74] Tonini, D. E., & Dean, S. W. (1976). Chloride corrosion of steel in concrete, ASTM STP629. ASTM, Philadelphia, PA, p 181. [75] Fratesi, R., Moriconi, G., & Coppola, I. (1996). The influence of steel galvanization on rebar behaviour in concrete. In: C. L. Page, P. B. Bamforth, & J. W. Figg (Eds), Corrosion of reinforcement in concrete construction (pp. 630–641). Royal Society of Chemistry. [76] Beeby, A. W. (1983). Cracking, cover, and corrosion of reinforcement. Concrete International — Design and Construction, 5, 2, 35– 40. [77] Mu¨ller, H. H. (1993). Behaviour of galvanized rebars in concrete. In: S. Nagataki, T. Nireki, & F. Tomosawa (Eds), Durability of building materials and components, vol. 1, (pp. 7– 156). E&FN Spon, London. [78] Mehta, P. K., & Gerwick, B. C. Jr (1982). Cracking-corrosion interaction in concrete exposed to marine environment. Concrete International — Design and Construction, 4, 4, 45 –51. [79] Powers, T. C., Copeland, L. E., & Mann, H. M. (1959). Capillary continuity or discontinuity in cement pastes. Journal Portland Cement Association, 1, 38– 48. [80] Yeomans, S. R. (1994). A conceptual model for the corrosion of galvanized steel reinforcement in concrete. In: R. N. Swamy (Ed.), Corrosion and Corrosion Protection of Steel in Concrete. vol. II, (pp. 1299–1309). Sheffield Academic Press. [81] Schikorr, G. (1941). Some failure phenomena occurring with aluminium, iron and zinc embedded in brickwork. Wissenschaftliche Abhandlungen der Deutschen Material pru¨fungsanstalten, 2, 51 p. [82] Allan, N. D. (1991). Galvanized reinforcement — the Bermuda experience, Seminar on Galvanized Reinforcement, University of Sheffield, April 1991. [83] Halstead, P. E., & Ridge, H. G. (1957). Corrosion of metals in contact with concrete. Chemistry and Industry, 1409 p. [84] Everett, L. H., & Treadaway, K. W. J. (1970). The use of galvanized steel reinforcement in building. Building Research Station Current Papers, CP 3/70, 10 p. [85] Allan, N. D., & Churchman, A. E. (1999). The New Watford Bridge — Bermuda. In: R. N. Swamy (Ed.), Infrastructure regeneration and rehabilitation — improving the quality of life through better construction — a vision for the next millennium (pp. 487–496). Sheffield Academic Press. [86] Swamy, R. N. (2001). Holistic design — key to sustainability in concrete construction. Proceedings, Institution of Civil Engineers (UK), Structures and Buildings, 146, 4, 371–379.
This Page Intentionally Left Blank
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 3
Corrosion of Metals in Concrete Alberto A. Sagu¨e´s University of South Florida, USA
3.1. The Steel Environment in Concrete This chapter highlights key issues of the corrosion of metals in concrete, with emphasis on chloride-induced deterioration, quantitative approaches to forecast corrosion performance, and strategies to improve durability in new construction. Hardened concrete is a material with a pore-size distribution range that extends down to the dimension of molecules. As a result of capillary depression of the vapor pressure of water, significant amounts of liquid water can exist in small concrete pores even when the external relative humidity falls well below 100%. The reinforcing steel in contact with that pore water can therefore be subject to metallic corrosion if the water chemistry can support the necessary anodic and cathodic reactions. In normal Portland cement concrete made with uncontaminated water and non-aggressive admixtures or aggregates, the pore water solution contains mostly Kþ, Naþ, Ca2þ, and OH2 ions (typically 12.5 , pH , 13.5) and dissolved O2 from atmospheric exposure. That environment places the reinforcing steel in conditions propitious to the formation of a passive layer and consequently small corrosion rates (typically much less than 1 mm/y) that pose little threat of deterioration of the surrounding concrete over long periods of time [1]. The stability of the passive layer is compromised if the pore water solution becomes contaminated by chloride ions or if the pH decreases substantially. These changes often result from exposure to the concrete service environment if it includes seawater or highway deicing salts, or from the slow carbonation attack of the concrete by atmospheric CO2. If enough oxygen is available to support a significant reduction reaction rate, metal dissolution can be expected to occur at a commensurate rate in the region of the steel surface where passivity breakdown has taken place. As chloride-induced corrosion is widespread and causes severe structural damage throughout the world, this chapter focuses on that mode of reinforcement
72
A.A. Sagu¨e´s
corrosion. Many of the findings and forecasting as well as control methods described in the following apply also, with appropriate modification, to the problem of carbonation-induced corrosion.
3.2. The Initiation – Propagation Model for Corrosion of Steel The corrosion of steel in concrete exposed to chloride-ion contamination can be viewed as developing through a two-stage mechanism developed by Tuutti [2] and illustrated in Fig. 1. In the first or initiation stage the chloride ions penetrate through the concrete that is present between the environment and the reinforcing bars, causing the concentration of chloride ions at the steel surface to increase with time but not yet reaching the critical level to depassivate the steel surface. The initiation stage ends when the critical chloride level (the concentration threshold) is reached and active corrosion of the steel begins. The next stage is the propagation stage, in which the steel cross-section is reduced and corrosion products accumulate at the steel surface. This causes increasing internal concrete stresses and subsequent cracking of the concrete cover, with attendant loss of mechanical steel– concrete bond. Reinforcement weakening from reduction in the steel cross-section and possible embrittlement of the steel may also occur. The overall deterioration of the structure continues during the propagation stage, which is declared to have ended when a given tolerable deterioration limit is exceeded. A similar two-stage description of the corrosion process can be used for carbonation-induced corrosion. In that case, during the initiation stage the concrete carbonation front penetrates through the concrete until the pH of
Figure 1: The initiation – propagation model for corrosion of steel in concrete.
Corrosion of Metals in Concrete
73
the pore solution at the reinforcement surface becomes low enough not to support passivity. The ensuing propagation stage proceeds as described above. Key factors affecting each of the deterioration stages are discussed in the following sections.
3.2.1. Initiation Stage: The Chloride Concentration Threshold Upon breakdown of passivity, the pore water in contact with the active corrosion region of the steel surface becomes enriched by metal ions. The enrichment in cations tends to attract additional chloride ions and reduce the local pH by mechanisms common with other localized corrosion processes. Thus, at least in its initial stages, chloride-induced corrosion of steel in concrete has the characteristics typically observed in pitting corrosion [3]. The chloride concentration threshold for steel in concrete has been found to be a function of the concentration of OH2 ions in simulated concrete-pore solutions [4–6] and of the potential of the steel surface with respect to the surrounding pore solution [7]. However, there is not a reasonably exact relationship between those variables that can satisfactorily describe all the results reported in the literature. Broadly speaking, it may be said that for plain steel bars the concentration threshold [Cl2]T is on the order of [OH2] (both as concentrations in the pore water) when the open circuit potential of the steel E is approximately 2150 mV in the Cu/CuSO4 scale (representative of passive steel in atmospherically exposed reinforced concrete). The value of [Cl2]T has been reported to increase as the potential is made more negative at a rate of one concentration decade for every ,100 to ,500 mV change in E [7 –10]. Direct measurement of the concentration of Cl2 ions in the pore solution is difficult, and concentrations are commonly obtained instead as a fraction of concrete or cement weight. Conversion to concentrations in the actual pore solution is challenging since only a fraction of the total chloride ion content is present in the pore solution; the remainder is in the form of chloride salts physically/chemically bound within the hydrated paste structure [11–14]. Direct indications of the actual pore solution pH are likewise hard to obtain [12,15,16], so evaluations based on pore water chemistry are generally limited to the prediction of relative changes and to comparative studies. Moreover, reliance on pore water composition only as a determining factor in the concentration threshold has been questioned by some authors [17]. Empirical determinations of the concentration threshold for steel in non-water-saturated, atmospherically exposed concrete give values on the order of 0.4% of the cement content of the concrete expressed in mass per unit volume of concrete (cement factor), sometimes adjusted for the presence of modifiers such as pozzolanic admixtures [18]. The influence of
74
A.A. Sagu¨e´s
the cement factor probably reflects the greater OH2 content of the pore water as more cement per unit volume of concrete is used.
3.2.2. Initiation Stage: Combined Factors Design for long-term durability from a corrosion standpoint aims to extend the length of either, or both, the corrosion initiation and the corrosion propagation stages. However, corrosion rates after chloride-induced depassivation tend to be high [16] and so much of the effort has been directed to extend as much as possible the length of the initiation stage. Key factors that determine the length of the initiation stage are discussed below. The penetration of chloride ions through appreciable depths in concrete can be considered on first approximation to be a diffusional process such that J ¼ 2DLC
ð1Þ
where J and C are the flux and concentration of chloride ions, respectively, and D is the diffusion coefficient. For the time being, Eq. (1) will be assumed to apply when C is the total chloride concentration expressed as mass per unit volume of concrete. Furthermore, if D is assumed to be independent of both space and time, then the non-steady state behavior of the concentration obeys Fick’s second law: MC=Mt ¼ DL2 C
ð2Þ
For reinforced concrete structures exposed to certain services regime (for example, some marine environments and deicing salt highway applications), the concentration of chloride ions at the concrete surface may approximate a limiting value CS after a relatively short fraction of the life of the structure [18,19]. Under those conditions, and if the initial chloride content of the concrete is negligible, the solution to Eq. (2) for a one-dimensional problem along x has the form x ð3Þ Cðx; tÞ ¼ CS 1 2 erf pffiffiffiffi 2 Dt where erf is the error function [20] and Cð0; tÞ ¼ CS : For a given concentration threshold CT and reinforcement cover thickness xr ; the length ti of the initiation stage is then given by ti ¼ ðx2r =4DÞðerfinvð1 2 CT =CS ÞÞ22
ð4Þ
where erfinv is the inverse error function. The behavior in Eq. (4) is summarized graphically in Fig. 2. For the cases in which Eq. (4) applies, for a given ratio of CT =CS ; the value of ti increases with the square of the concrete cover and inversely with the diffusion coefficient.
Corrosion of Metals in Concrete
75
Figure 2: Graphical representation of the relationship in Eq. (4).
The dependence with CT =CS is more complicated, but from the local slope of the curve it can be seen that for cases of frequent interest (CT < 1 kg/m3, CS < 10 kg/ m3 so that CT =CS , 0:1) the value of ti is roughly proportional to CTn and inversely proportional to CS2n ; with n < ð1=2Þ – ð1=3Þ: Thus, within the context of Eq. (4), the parameters most influential on ti are, in order of decreasing importance, xr ; D and CT =CS : More sophisticated models of chloride penetration exist, which include variation of D and CS with time, as well as the effect of initial chloride contamination and chloride binding, but the relative importance of the variables indicated above is comparable in those treatments as well [1,21].
3.2.3. Corrosion Propagation Stage The length of the corrosion propagation stage is the time from the beginning of active corrosion to the moment at which a tolerable deterioration limit is reached. In some service conditions, that moment may be considered to correspond to the appearance of the first corrosion-induced crack on the concrete surface. A related concept is the critical corrosion penetration (Xcrit ), the space-averaged depth of reinforcement corrosion that causes enough corrosion expansion to crack the concrete cover over the bar. If the corrosion of the reinforcement proceeds at a space-averaged rate, CR, that is also constant with time, the length tp of the propagation stage is given by tp ¼ Xcrit =CR
ð5Þ
76
A.A. Sagu¨e´s
The value of Xcrit (mm) has been determined empirically by a number of investigations [22]. When the corrosion rate and the cover thickness are relatively uniform over the length of the reinforcement and concrete is not water-saturated, Xcrit has been found to be approximately proportional to the ratio of concrete cover thickness to the reinforcement diameter (F): Xcrit < 0:01 mmðx=FÞ
ð6Þ
Combining Eqs. (5) and (6) yields tp < 0:01 mmðx=FÞ=CR
ð7Þ
Corrosion rates of active steel in concrete in aggressive environments are of the order of 10 mm/y [22]. If the rate were to stay at that value, Eq. (7) indicates that, for a typical configuration with x=F ¼ 4; the length of the corrosion propagation stage would be of the order of only a few (about 4) years. Average values of tp of about 3.5 years have been actually recorded in surveys of highway bridge decks built with plain reinforcement and exposed to deicing salt service. More sophisticated schemes to evaluate tp are available [23]. Nevertheless, as seen above, the propagation stage can be quite short compared with some common durability goals (e.g. 75 years). As a result, much effort to achieve long-term durability of reinforced concrete with plain reinforcement has been directed at postponing as long as possible the onset of active corrosion.
3.3. Integrated Corrosion Forecasting For a single structural element with uniform exposure conditions, the corrosionlimited service life may be expressed in terms of Fig. 1 as the length of time tc ¼ ti þ tp since, at the end of that period, external manifestations of damage (e.g. surface cracking) would appear. The value of ti and tp may be estimated as shown in the previous sections, introducing more sophisticated models as may be necessary. The effect of various exposure and material variables on corrosionlimited durability can then be examined in terms of how those variables affect ti and tp ; as shown in the following sections that discuss the relative merit of corrosion control approaches for concrete reinforcement. Before addressing specific corrosion-control methods, however, corrosion forecasting for extended systems (e.g. bridges and buildings) will be briefly addressed. For those entities, a damage function can be defined as the amount of damage in a structure as a function of service time. For a single structural element as mentioned above, the element could be declared damaged when corrosioninduced surface cracking becomes present, and not damaged before then. The damage function in such a case would have the shape of a step function with
Corrosion of Metals in Concrete
77
Figure 3: Example of corrosion forecasting model output (total damaged area in pair of bridges as function of service time) for a modeling scenario assuming CT ¼ 1:6 kg/m3 and distributed values of D and CS : TS1 to TS3 correspond to elevation regimes for tidal, lower splash and upper splash zones, respectively. TSA corresponds to the total damage, which is the sum of all three regions [28].
the transition at t ¼ ti þ tp : A large structure may be considered as being composed of many individual elements, each with different exposure and configuration parameters. The damage function in that case reflects the sum of numerous individual step functions, each experiencing a transition at a different time and thus resulting in a characteristic S-curve where the extent of damage increases gradually with time toward a terminal cumulative limit. An example of corrosion forecast of a distributed system is the work performed for the substructure of a set of bridges in a marine environment [24]. In that case, the surface of the partially immersed piles in the substructure was divided into a large number Nt of individual elements of equal area Ae : Estimates of D; the diffusion coefficient, were obtained from chloride analysis of multiple extracted concrete cores. From the results, a statistical distribution Pd ðDÞ was abstracted such that the number dN of elements having chloride diffusivity values between D
78
A.A. Sagu¨e´s
and D þ dD was approximated by dN ¼ Nt Pd ðDÞdD
ð8Þ
Additional measurement from the extracted cores, and multiple direct measurements of concrete cover thickness, yielded distribution functions PCS ðCS Þ and Pxr ðxr Þ: By appropriate combination of the three distribution functions, assumption of a fixed value of tp ; and application of Eq. (4), an integrated damage function was obtained for all the elements in the substructure that were within a given elevation range above the high tide level. The procedure was repeated for three elevation ranges that reflected different global exposure conditions. An integrated damage function forecast was then constructed for the entire pair of bridges, as shown in Fig. 3. In this case, the projected damage function is presented as total area (m2) of elements each with area Ae ¼ 0:1 m2 as a function of service time and the individual projections for each of the three elevation ranges is also included. Because of uncertainty inherent in the abstracted and assumed parameters, projections of this type tend to be used in decision making as a comparative, rather than an absolute, performance evaluation assessment. Statistical approaches of this type are encountering increasing application in a variety of structural systems [24 –26]. Advanced models introducing other effects, such as dependence of chloride threshold on reinforcement potentials, are being introduced as well [27,28].
3.4. Strategies to Improve Durability The following describes approaches to improve durability of reinforced concrete subject to reinforcement corrosion. The scope is limited to those measures that can be taken at the design and construction stage and that do not involve external added systems. Thus, methods such as cathodic prevention implemented at the moment of placement in service, cathodic protection of corroding structures, or other repair and rehabilitation procedures are not discussed. Those technologies are described in detail elsewhere in the literature [1].
3.4.1. Concrete Quality and Cover Thickness Based on the factors discussed above, design for durability has concentrated first on specifying a concrete cover as thick as possible commensurate with the mechanical design constraints of the structure. In structural members such as bridge footers and support columns that operate mostly under compression, this is relatively easy to achieve and concrete covers of 100 mm or more are
Corrosion of Metals in Concrete
79
not uncommon. Similar concrete thickness can likewise be achieved without excessive cracking in prestressed elements that are not subject to impact forces. However, structural elements subject to flexural loads or prestressed members such as driven piles that can experience mechanical damage from impact are susceptible to extensive cover cracking if excessive covers are specified. In those cases, the other factors responsible for extending ti become more critical. The diffusion coefficient D; considered on first approximation as a constant property of the concrete, may be viewed as a magnitude that decreases as the concrete quality increases. Apparent values of D; obtained by fitting Eq. (3) to chloride concentration profiles measured in field-extracted cores [18,19,28] have shown that D can vary widely, for example, from 1029 to 1027 cm2/sec depending on such factors as the w=c ratio and cement factor of the concrete, as well as pozzolanic additions, casting quality and the type of aggregates used. Control of these factors has been successfully used to consistently achieve values of D approaching 1029 cm2/sec in structures where long-term durability is a key design goal [18,29].
3.4.2. Concrete Surface Treatments and Chloride Barriers After manipulating concrete cover and concrete quality, durability extension design strategies concentrate on controlling CS and CT : Concrete surface treatments and impermeable membranes can hinder the penetration of salt water into the concrete and thus reduce CS : These surface modification methods are used successfully in specialized applications [30] but their applicability is subject to the physical arrangement of the system and can also be seriously limited by the magnitude of the initial cost and the need for periodic reapplication.
3.4.3. Corrosion Inhibitors Considerable effort has been directed at increasing CT by strategies that include the use of corrosion inhibitors, alternative reinforcement materials, and modification of the surface of otherwise conventional reinforcement material. A corrosion inhibitor is a substance added in small quantities to the concrete to improve the corrosion resistance of the reinforcing steel. The most commonly used corrosion inhibitor in concrete at present is calcium nitrite, added as an admixture at the time of mixing concrete. Typical rates of addition are 22 l of 30% calcium nitrite solution per m3 of concrete, which are intended to raise CT to a value of
80
A.A. Sagu¨e´s
the order of 10 kg/m3 of concrete [31]. Other more recently introduced inhibitors are based on organic substances [32,33]. Corrosion inhibitors are experiencing increasing use for reinforced concrete in aggressive environments, but key questions on the suitability for long-term durability design remain. In particular, the ability of the inhibitor to remain in place and be effective over the projected lifetime of the structure needs to be established. Of the inhibitors currently in use, only calcium nitrite has a documented record of use in actual field applications that clearly exceeds 10 years. Inhibitors can also represent a significant additional cost that has to be weighed against projected performance over periods of times for which no actual performance data have yet been established.
3.4.4. Alternative Reinforcement Materials 3.4.4.1. Carbon Steel Replacement Reinforcing steel bars made of stainless steels and composite materials have been used in a restricted number of applications. The concentration threshold of stainless steels is significantly higher than that of carbon steel [7,34] and active corrosion initiation can consequently be postponed. The superiority of stainless steels over conventional carbon steel has been well established in field tests in the marine environment [35,36]. Ferritic/martensitic alloys, and especially austenitic stainless steel, have shown extremely low rates of corrosion when compared with plain carbon steel in concrete contaminated to levels well exceeding 10 kg/m3. However, stainless steels have unit weight costs typically one order of magnitude higher than that of plain reinforcement steel and thus the use of stainless steel bar is severely constrained by economic considerations. Non-metallic composite materials for concrete reinforcement have been the subject of much interest as a means to avoid metallic corrosion problems. Fiberreinforced plastics, mostly glass-fiber based, have been proposed as steel replacement. However, while the glass fibers and the resin matrix are resistant to the action of chloride ions, the alkaline concrete environment is highly aggressive to the glass fibers, especially under wet conditions such as those prevalent in marine applications [37]. The ability of the resin matrix itself to withstand long-term exposure to the concrete pore solution at high relative humidity is also questionable. Other fiber materials such as aramids have been introduced for corrosion control but early results on behavior in aggressive wet environments have showed instances of deterioration [38]. Carbon fiber reinforced plastics were introduced recently and their corrosion resistance is under evaluation. Because carbon fiber reinforcement is electrically conductive, the possibility of galvanic deterioration exists. A recent investigation showed that
Corrosion of Metals in Concrete
81
significant macrocell action could develop if the composite material is in contact with steel components [39].
3.4.4.2. Coated Reinforcing Steel Reinforcement coatings are intended to improve corrosion performance while keeping an inexpensive plain steel bar core. The coatings may be divided into nonmetallic and metallic types. By far the most commonly used non-metallic type is epoxy coating. The coating is about 200 mm thick and is formed by quickly curing at high temperature (,2008C) epoxy powder deposited on the freshly blasted and clean reinforcement surface. Epoxy-coated reinforcement (ECR) has been used in the US and Canada since the mid-1970s in over 100,000 structures, notably highway bridge decks exposed to deicing salt. Use outside the US and Canada is very limited. The asproduced ECR costs about 30% more than plain steel bar. Additional costs may result from recommended procedures for transportation, handling and tying, inspection and patching, the need for longer lap joints and concreting precautions to avoid damage to the coating. While it is debatable whether the steel underneath a sound coating is in the passive state, the initiation of high corrosion rates in ECR under those conditions should be much delayed since the metal is separated from the concrete by the coating. If corrosion is initiated (for example at a coating break) the coating should hinder the cathodic reaction by limiting the transport of oxygen to the rest of the steel surface, thus reducing or eliminating the formation of corrosion macrocells. These effects should therefore result in significantly lower overall corrosion rates than in plain steel reinforcement with consequent extension of the length of the propagation stage of corrosion. These expectations have been confirmed in laboratory tests [40]. The reported performance of ECR has also been satisfactory for many of the State transportation agencies using this material [41,42]. However, inadequate ECR performance was reported for the substructure of several large Florida bridges in aggressive marine environments, leading to cease specification of ECR by Florida DOT in 1992 [26,43]. Since then, ECR use has come under scrutiny by transportation agencies in other states [44]. The poor performance in Florida was manifested by severe undercoating corrosion, which was preceded by marked loss of adhesion between the coating and the underlying metal. Breaks in the coating were found to be important in aggravating the corrosion process. Laboratory experiments showed that significant corrosion macrocells developed under severe marine substructure conditions, even if only 2% of the bar surface (the allowable level until the late 1980s) was affected by coating breaks [26]. Present-day specifications for ECR call for much stricter allowable coating breaks (equivalent to ,0.1% of the surface) and special coating plant quality assurance procedures have been implemented. In addition, coating
82
A.A. Sagu¨e´s
formulations and reinforcement substrate surface treatments have been recently introduced to enhance resistance of the coating to cathodic disbondment with the expectation that improvements in field performance will result. Coatings with inorganic silicate polymers have been used experimentally [45]. These coatings are thought to protect by both increasing the chloride concentration threshold and reducing the rate of corrosion during the propagation stage by hindering the oxygen reduction reaction. Their use in combination with an outer layer of epoxy coating has been proposed but not tested. Surface treatments such as cement paste application prior to concreting have also been used, but in a relatively small scale. The usefulness of these coatings has not been tested by extensive field application. Metallic coatings using a material more corrosion resistant than plain steel offers the potential of highly improved corrosion performance with low cost impact. Recently, interest has developed in the use of reinforcement clad with a thick (0.5– 1 mm) outer layer of stainless steel manufactured at a relatively low cost compared with that of solid stainless steel [46]. This material is presently being used in several demonstration projects. Other thin metallic coatings, such as copper-coated reinforcement, have received some attention in FHWA sponsored laboratory tests conducted by McDonald et al. [40]. Galvanized reinforcement, on the other hand, the subject of this book, represents by far the most widely used application of a metallic reinforcement coating to date. As galvanized reinforcement is addressed extensively in other chapters, only brief remarks are presented here concerning its durability parameters. In the first instance, the chloride concentration threshold of galvanized reinforcement (CTG ) has been reported to be double or even higher than the value (CTP ) for plain steel bars [47,48]. The implications of such threshold increases may be examined by assuming simplified conditions leading to the behavior summarized in Fig. 2, for example a typical exposure regime where CTP , 0:1CS ; and cases where CTG ¼ 2CTP or ¼ 3CTP : The projected resulting increase in ti over that for plain steel is of the order of , 50 and ,100%, respectively. If relatively mild exposure conditions were to apply, such as in moderate deicing salting regimes on bridge decks, one may encounter CT , 0:2CS and the corresponding projected increases in ti rise significantly to ,100 and , 500%, respectively. Information available on the values of CTG under various exposure circumstances is still quite limited for galvanized steel in concrete. As a result, the actual change in the length (i.e. time) of the initiation stage for corrosion that may be achieved upon replacement with galvanized reinforcement can only be estimated at this time over a range of scenarios, as exemplified above. Continuing research in this area is consequently important. Likewise, galvanized
Corrosion of Metals in Concrete
83
reinforcement has been reported to experience lower corrosion rates during the propagation stage than those of plain steel, resulting in an increase in tp as well. However, quantitative information to this effect (as discussed in other chapters) is also limited and mostly derived from laboratory studies that may not be fully relevant to actual field conditions. As in the case of chloride threshold information, additional data are essential to quantitatively assess the benefits of the use of galvanized reinforcement during the propagation stage. While carbonation-induced corrosion was not the focus of this chapter, improved resistance of galvanized reinforcement to that mode of attack has been documented as well in laboratory investigations such as those by Gonza´lez and Andrade [49]. Durability improvements for this mode could also be quantified in terms of a modified durability model; however, as in the case of the chlorideinduced corrosion model, input parameters are yet to be determined precisely. The overall galvanized reinforcement corrosion experience, summarized in the rest of this book, may be a promising starting point for future rational analyses based on the initiation– propagation model to glean effective values of the key durability parameters to supplement those to be derived from future research.
References [1] Bentur, A., Diamond, S., & Berke, N. (1997). Steel corrosion in concrete. E&FN Spon, New York. [2] Tuutti, K. (1982). Corrosion of steel in concrete. Swedish Cement and Concrete Research Institute, Stockholm, 18 pp. [3] Kaesche, H. (1985). Metallic corrosion. NACE International, Houston, TX. [4] Hausmann, D. A. (1967). Materials Protection, 6, 11, 19–23. [5] Gouda, V. K. (1970). British Corrosion Journal, 5, 9, 198 p. [6] Li, L., & Sagu¨e´s, A. A. (2001). Corrosion, 57, 19 p. [7] Pedeferri, P. (1996). Construction and Building Materials, 10, 5, 391 p. [8] Alonso, C., Andrade, C., & Castellote, M. (1997). The influence of the electrical potential in the chloride threshold for rebar depassivation. In: H. Justnes (Ed.), Proceedings of the 10th International Congress on the Chemistry of Cement, Gothenburg, Sweden, vol. 4 (Paper No. 4iv082). [9] Alonso, C., Castellote, M., & Andrade, C. (2002). Dependence of chloride threshold with the electrical potential of reinforcements. Electrochimica Acta, 47, 3469 p. [10] Li, L., & Sagu¨e´s, A. A. (2002). Effect of chloride concentration on the pitting and repassivation potentials of reinforcing steel in alkaline solutions. Corrosion, 58, 305 p. [11] Andrade, C., & Page, C. L. (1986). British Corrosion Journal, 21, 1, 49 p. [12] Diamond, S. (1986). Cement, Concrete and Aggregates, 8, 2, 97 p. [13] Rasheeduzzafar, Hussain, S. E., & Al-Saadoun, S. S. (1992). Effect of tricalcium aluminate content of cement on chloride binding corrosion of reinforcing steel in concrete. ACI Materials Journal, 89, 1, 3–12.
84
A.A. Sagu¨e´s
[14] Tritthart, J. (1989). Chloride binding in cement II. The influence of the hydroxide concentration in the pore solution of hardened cement paste on chloride binding. Cement and Concrete Research, 19, 5, 683– 691. [15] Sagu¨e´s, A. A., Moreno, E. I., & Andrade, C. (1997). Evolution of pH during in situ leaching in small concrete cavities. Cement and Concrete Research, 27, 11, 1747 –1759. [16] Andrade, C., & Alonso, M. C. (1994). Values of corrosion rate of steel in concrete to predict service life of concrete structures. In: G. Cragnolino, & N. Sridhar (Eds), Application of accelerated corrosion tests to service life prediction of materials, ASTM STP 1194. ASTM, Philadelphia, PA, 282 p. [17] Glass, G. K., & Buenfeld, N. R. (1997). Corrosion Science, 39, 5, 1001 p. [18] Bamforth, P. B. (1999). The derivation of input data for modelling chloride ingress from eight-year UK coastal exposure trials. Magazine of Concrete Research, 51, 2, 87 –96. [19] Cady, P. D., & Weyers, R. E. (1984). Deterioration rates of concrete bridge decks. Journal of Transportation Engineering, 110, 1, 34 –44. [20] Handbook of Physics and Chemistry. CRC Press, Boca Raton, FL, 54th ed., 1973 –1974, p. D-132. [21] Thomas, M. D. A., & Bentz, E. C. (2000). Life-365, computer program for predicting the service life and life-cycle costs of reinforced concrete exposed to chlorides. Concrete Corrosion Inhibitors Association, Potomac, MD. [22] Torres-Acosta, A. A., & Sagu¨e´s, A. A. (2000). In: V. M. Malhotra (Ed.), Concrete cover cracking with localized corrosion of the reinforcing steel, ACI special publication 192. American Concrete Institute, Farmington Hills, MI, 591 p. [23] Weyers, R. E. (1998). Service life model for concrete structures in chloride laden environments. ACI Materials Journal, 95, 4, 445–453. [24] Kirkpatrick, T., Weyers, R. E., Anderson-Cook, C., & Sprinkel, M. (2002). Cement and Concrete Research, 32, 8 p. [25] Hartt, W., Lee, S. K., & Costa, J. (1998). Condition assessment and deterioration rate projection for chloride contaminated concrete structures. In: W. P. Silva-Araya, O. T. de Rincon, & L. Pumarada O’Neill (Eds), Repair and rehabilitation of reinforced concrete structures: the state of the art. ASCE, Reston, VA, 82 p. [26] Sagu¨e´s, A. A., Powers, R. G., & Kessler, R. (2001). Corrosion performance of epoxycoated rebar in Florida Keys bridges, Corrosion 2001, Paper 01642. NACE International, Houston, TX. [27] Sagu¨e´s, A. A., & Kranc, S. C. (1998). Model for a quantitative corrosion damage function for reinforced concrete marine substructure. In: O. Troconis, & C. Andrade (Eds), Rehabilitation of corrosion damaged infrastructure, Proceedings of the Third NACE Latin–American Region Corrosion Congress, P Castro. NACE International, Houston, TX, p. 268, ISBN 970-92095-0-7. [28] Sagu¨e´s, A. A., Scannel, W., & Soh, F. W. (1998). Development of a deterioration model to project future concrete reinforcement corrosion in a dual marine bridge, Proceedings of International Conference on Corrosion and Rehabilitation of Reinforced Concrete Structures, Orlando, FL, Dec 1998. Federal Highway Administration, USA, CD ROM publication no. FHWA-SA-99-014. [29] Berke, N. S., & Hicks, M. C. (1992). Estimating the life cycle of reinforced concrete decks and marine piles using laboratory diffusion and corrosion data. In: V. Chaker (Ed.), Corrosion forms and control for infrastructure, ASTM STP 1137. ASTM, Philadelphia, PA, 207 p.
Corrosion of Metals in Concrete
85
[30] Flu¨ckiger, D., Elsener, B., Studer, W., & Bohni, H. (1994). Effects of organic coatings on water and chloride transport in reinforced concrete. In: R. N. Swamy (Ed.), Corrosion and corrosion protection of steel in concrete. vol. II, (1017 p). Sheffield Academic Press, Sheffield, UK. [31] Berke, N., & Sundberg, K. (1989). The effects of calcium nitrite and microsilica admixtures on corrosion resistance of steel in concrete. American Concrete Institute, SP 122-15, Detroit, MI, p. 265. [32] Nmai, C. K., Farington, S. A., & Bobrowski, G. S. (1992). Organic-based corrosioninhibiting admixture for reinforced concrete. Concrete International, 14, 4, 44 p. [33] Maeder, U. (1994). A new class of corrosion inhibitors. In: R. N. Swamy (Ed.), Corrosion and corrosion protection of steel in concrete. vol. II, (851 p). Sheffield Academic Press, Sheffield, UK. [34] Nurnberger, U. (1996). Corrosion behavior of welded stainless steel reinforced steel in concrete. In: C. L. Page, P. B. Bamforth, & J. W. Figg (Eds), Corrosion of reinforcement in concrete construction, The Royal Society of Chemistry, Special Publication 183, 623 p. [35] Cox, R. N., & Oldfield, J. W. (1996). The long-term performance of austenitic stainless steel in chloride contaminated concrete. In: C. L. Page, P. B. Bamforth, & J. W. Figg (Eds), Corrosion of reinforcement in concrete construction, The Royal Society of Chemistry, Special Publication 183, 662 p. [36] Treadaway, K., Cox, R., & Brown, B. (1989). Durability of corrosion resisting steels in concrete. Proceedings of the Institution of Civil Engineers, Part 1, 86. [37] Sen, R., Mariscal, D., & Shahawy, M. (1993). Durability of fiberglass pretensioned beams. ACI Structural Journal, 90, 5, 525–533. [38] Sen, R., Shahawy, M., Rosas, J., & Sukumar, S. (1998). Durability of aramid pretensioned elements in a marine environment. ACI Structural Journal, 95, 5, 578– 587. [39] Torres-Acosta, A. A., Sagu¨e´s, A. A., & Sen, R. (1998). Galvanic interactions between carbon fiber reinforced plastic composites and steel in chloride contaminated concrete, Corrosion’98, paper no. 648. NACE International, Houston, TX. [40] McDonald, D. B., Pfeifer, D. W., & Sherman, M. R. (1998). Corrosion evaluation of epoxy-coated, metallic-clad and solid metallic reinforcing bars in concrete, Report no. FHWA-RD-98-153. Federal Highway Administration, Washington, DC. [41] Sohanghpurwala, A., & Scannell, W. T. (1998). Verification of effectiveness of epoxycoated rebars, final report. Pennsylvania Department of Transportation, Harrisburg, PA, Project no. 94-05. [42] Smith, J. L., & Virmani, Y. P. (1996). Performance of epoxy coated rebars in bridge decks, Report no. FHWA-RD-96-092. Federal Highway Administration, Washington, DC. [43] Sagu¨e´s, A. A. (1994). Corrosion of epoxy-coated rebar in Florida bridges, final report. Florida Department of Transportation, Florida DOT Research Center, Tallahassee, WPI no. 0510603. [44] Weyers, R. E., Pul, W., & Sprinkel, M. M. (1998). Estimating the service life of epoxy coated reinforcing steel. ACI Materials Journal, 95, 5, 546– 557. [45] Sagu¨e´s, A. A., Boucher, B., & Chang, X. (1992). Corrosion performance in concrete of reinforcing steel with a protective inorganic coating, Corrosion’92, paper no. 197. NACE International, Houston, TX. [46] Cui, F., Sagu¨e´s, A. A., & Powers, R. G. (1991). Corrosion behavior of stainless steel clad rebar, Corrosion’91. NACE International, Houston, TX, Paper 01645.
86
A.A. Sagu¨e´s
[47] Yeomans, S. R. (1994). Performance of black, galvanized, and epoxy coated steel reinforcement in chloride contaminated concrete. Corrosion, 50, 1, 72 –81. [48] Cornet, I., & Bresler, B. (1981). Galvanized steel in concrete: literature review and assessment of performance, Galvanized reinforcement for concrete — II. International Lead Zinc Research Organization, Research Triangle Park, NC, USA. [49] Gonza´lez, J. A., & Andrade, C. (1982). Effect of carbonation, chlorides and relative ambient humidity on the corrosion of galvanized rebars embedded in concrete. British Corrosion Journal, 17, 1, 21–28.
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 4
Zinc Materials for Use in Concrete Thomas J. Langill a and Barry Dugan b a
American Galvanizers Association, USA
b
Zinc Corporation of America, USA
4.1. Introduction Zinc metal has a number of characteristics that make it well suited for use as a coating for protecting iron and steel products from corrosion. The excellent field performance of zinc coatings is due to its ability to form dense, adherent corrosion product films and a subsequent rate of corrosion considerably below that of ferrous materials, some 10 – 100 times slower, depending on the atmosphere. While a fresh zinc metal surface is quite reactive, the zinc metal forms a thin film of corrosion products on exposure to the atmosphere. This film of corrosion products transforms into a dense, transparent barrier layer that prevents strong attack on the zinc metal. Further, the barrier film is not water-soluble and erodes slowly over time [1,2]. In addition to creating a barrier between the steel and the environment, zinc also has the ability to galvanically protect steel. If the coating is damaged and the underlying steel exposed to the environment, zinc — being anodic to iron — will preferentially corrode and sacrificially protect the exposed steel against rusting. In this form of corrosion protection, broadly known as cathodic protection, the base metal becomes the cathode in the corrosion cell and the coating metal the anode. Fig. 1 shows the relative positions of zinc and steels in the galvanic series of metals. In a bimetallic couple, anodic metals (e.g. zinc, aluminum, cadmium) will sacrifice themselves to protect more cathodic metals (iron or steel in this case) against corrosion. This dual mode of corrosion protection afforded by zinc coatings on iron and steel, in that there is both barrier protection due to the presence of the coating and sacrificial protection in the event that the underlying steel is exposed, and the fact that the zinc coating is robust and can be easily transported and handled, has resulted in the widespread use of galvanizing for corrosion protection of a wide variety of iron and steel products exposed to many different environmental conditions.
88
T.J. Langill and B. Dugan
Figure 1: Arrangement of metals in the galvanic series. Zinc, being the more anodic metal, sacrificially protects iron.
Steel that is embedded into concrete can also be protected by a zinc coating. The zinc surface is more resistant to chloride ions than bare steel by a factor of up to 5 – 10 times and the corrosion rate of zinc, when the system goes into the active corrosion mode, is significantly lower than that of bare steel. Another feature is that the corrosion products of zinc migrate away from the coated bar and into the concrete matrix. As a result, they do not build up around the bar in the same way as do steel corrosion products, which, due to the swelling forces generated, can cause physical distress to the cover concrete. These features of the behavior of zinc in concrete will be discussed in greater detail in other chapters of this book. Zinc can be applied to the surface of steel by a number of commercial processes, including hot-dipping, electroplating, spraying and mechanical alloying [1–4]. Each process results in a coating having a range of thickness and performance characteristics that dictate choices in the selection of different types of zinc coating protection for iron and steel products. The typical structure and relative thickness of the coatings produced by several such processes are shown in Fig. 2. The anticipated service life of zinc coatings of varying thickness in a variety of atmospheric exposure conditions is shown in Fig. 3. Hot-dip galvanizing is the most common of the coating methods. In this process, steel parts are cleaned and dipped into a bath of molten zinc, either as individual
Zinc Materials for Use in Concrete
89
Figure 2: Microstructures of various zinc coatings on steels (after AGA).
pieces or as a continuous strip of sheet or wire product. The resultant coating comprises a series of zinc and zinc –iron alloy layers that are metallurgically bonded to the steel. The overall coating itself is quite thick, generally more than 100 mm, is tough and damage resistant, and is strongly adhered to the base steel. Further detail of the hot-dip galvanizing process is given in Section 4.2 of this chapter. Zinc plating, sometimes known as electrogalvanizing, is a versatile and effective method of applying a protective coating of metallic zinc to small
Figure 3: Life of protection versus thickness of zinc coating and type of atmosphere (after AGA).
90
T.J. Langill and B. Dugan
items, such as fasteners, or continuous sheet. The coating formed is relatively thin (ranging to about 100 g/m2) and therefore zinc plating should not be used for outdoor exposure without supplementary coatings being applied. Mechanical plating, or peen plating, is an electroless method used to coat steel with zinc. In this, a steel drum is used to tumble small steel parts with a mixture of zinc powder, chemicals and glass beads, and the zinc powder is peened onto the steel part. This process is used only on small items, particularly threaded components and close tolerance items. Zinc can also be sprayed onto fabricated items which usually cannot be (hot-dip) galvanized because of their size or because the coating must be applied on site. Known as zinc spraying or zinc metallizing, this process coats the steel with a layer of pure zinc that may be up to 250 mm (1500 g/m2) thick. Metallizing is also used as a means of repair or rejuvenation of steel members in need of in-service corrosion protection and in the application of zinc coatings for sacrificial anode-cathodic protection systems. Finally, zinc can also be applied via various types of paint systems. Zincrich (paint) coatings consist of zinc dust carried in organic or inorganic binders/vehicles and may be applied by brush or spray. They are barrier coatings that also provide cathodic protection to small exposed areas of steel. Suitable zinc-rich paints provide a useful repair system for touch-up and repair of damaged galvanized coatings. Pre-construction primers are another form of zinc-rich paint. They are relatively thin and weldable, and are used widely for ship building, storage tanks and similar steel plate construction intended for subsequent top coating. Of all the coating systems available to apply zinc to steel, the only coating that forms a metallurgical bond with the steel and has sufficient thickness to provide extended protection in a variety of environments is the hot-dip galvanizing process. As a result, hot-dip galvanizing is widely used throughout the world for the coating of steel products and components, fabricated items and structural sections, and also for steel reinforcing products including bars, wire and welded mesh. This process and the nature of the zinc-alloy coating will be discussed in detail in the following. It is worth noting that the term galvanizing, the chief characteristic of which is the formation of a metallurgical bond at the zinc–iron interface, is often loosely used to mean the coating of steel with zinc. This may include, of course, all of the coating systems referred to above and for this reason it is important to be precise when specifying galvanizing in order that the requisite coating thickness and coating morphology will be obtained. Thus, where substantial alloy-layer coatings are required on structural sections, for example, including reinforcing steel, hot-dip galvanizing should always be specified.
Zinc Materials for Use in Concrete
91
4.2. The Hot-Dip Galvanizing Process Unlike paint coatings that form an adhesive bond with the underlying steel, galvanized coatings develop a metallurgical bond through the formation of a series of iron– zinc alloy layers. However, in order for these intermetallic layers to form properly, the steel must be prepared and processed in a specific sequence. The galvanizing process consists of three basic steps, namely surface preparation, fluxing and galvanizing. Each of these steps, as discussed in the following, is important in obtaining quality-galvanized coatings. The layout of a typical hot-dip galvanizing line is shown in Fig. 4.
4.2.1. Surface Preparation It is essential that the material surface be clean and uncontaminated if a uniform, adherent coating is to result. Surface preparation is usually performed in sequence by caustic (i.e. alkaline) cleaning, water rinsing, acid pickling and water rinsing. The caustic cleaner is used to clean the material of organic contaminants such as dirt, paint markings, grease and oil, which are not readily removed by acid pickling. Scale and rust are normally removed by pickling in hot sulphuric acid (at 658C) or hydrochloric acid at room temperature. Sulphuric acid solutions typically contain 6– 12% by weight of acid, while hydrochloric acid solutions are about 15% by weight of acid. Water rinsing usually follows caustic cleaning and acid pickling. Surface preparation can also be accomplished using abrasive cleaning as an alternate to chemical cleaning. Abrasive cleaning is a mechanical process by which sand, metallic shot or grit is propelled against the material by air blasts or
Figure 4: Schematic of the hot-dip galvanizing process (after AGA).
92
T.J. Langill and B. Dugan
rapidly rotating wheels. This process physically dislodges rust and other materials from the surface of the steel and microscopically roughens the surface.
4.2.2. Fluxing The final cleaning of the steel is performed by a flux. The method of applying the flux to the steel depends upon whether the wet or dry galvanizing process is used. Dry galvanizing requires that the steel be dipped in an aqueous zinc ammonium chloride solution and then thoroughly dried. This pre-fluxing prevents oxides from forming on the material surface prior to the actual galvanizing step. In the wet galvanizing process, a layer of liquid zinc ammonium chloride is floated on top of the bath of molten zinc. The final cleaning occurs as the article being galvanized passes through this flux layer before entering the galvanizing bath.
4.2.3. Galvanizing The article to be coated is immersed in a bath (or kettle, as it is often known) of molten zinc maintained at a temperature of 435–4608C. Typical bath chemistry used in hot-dip galvanizing contains a minimum of 98% pure zinc along with a variety of trace elements or alloy additions. These additions, which could include lead (to 1.2%), aluminum (to 0.005%), tin (about 0.05%), nickel (to 0.1%), and bismuth (about 0.1%), can be made to the bath to enhance the appearance of the final product or to improve the drainage of the molten zinc as the material is withdrawn from the zinc bath. In the USA, the bath chemistry is specified by ASTM Specification B6 [5]. As the cleaned steel passes into the molten zinc, it is immediately wetted and a diffusion reaction between the zinc and the iron begins as it is heated to the temperature of the bath. This results in the formation of a series of iron–zinc alloys at the surface of the iron that commences quite rapidly but slows with time in the bath. The time of immersion in the galvanizing bath varies depending on the thickness and the chemical composition of the steel being coated. Thick steel sections take longer to galvanize than thin sections since the steel must be heated to the temperature of molten zinc before the galvanizing reactions can occur. However, the thickness of the coating does not substantially increase with longer immersion times. At the completion of galvanizing, the articles are slowly withdrawn from the bath and the excess zinc at the surface is removed by draining, wiping, vibrating or (for small items) centrifuging. The metallurgical reactions that result in the formation and structure of the alloy layers continue after the articles are withdrawn from the bath as long as their temperature remains near to the bath
Zinc Materials for Use in Concrete
93
temperature. Freshly galvanized articles are normally cooled in either water or in ambient air immediately after withdrawal from the bath. Water quenching stops the alloying reaction by removing the residual heat from the article. This prevents the continuation of the alloying reaction that could, if allowed to continue, convert the free zinc layer into an iron–zinc alloy. When cooled quickly in air, a thin layer of pure zinc remains on the surface, giving the freshly galvanized article a characteristic bright, shiny and smooth appearance.
4.3. Metallurgy of the Galvanized Coating and Alloy Structure 4.3.1. Coating Structure During the galvanizing process, a series of alloy layers form as a result of the metallurgical reaction between the molten zinc and the steel. The formation of these layers is influenced by a number of factors, including the chemistry of the steel, the immersion time in the zinc bath and the bath temperature, the surface roughness of the base steel, and the rate of withdrawal of the article from the zinc bath. The cross-section of a typical galvanized coating developed on steel with low silicon content (,0.03%) is shown in Fig. 5. The coating consists of a very thin gamma (g) layer next to the steel substrate, a blocky delta (d) layer and columnar growth of zeta (z) crystals. These iron– zinc intermetallic layers are covered by a layer of pure zinc eta (h) that remains on the surface as the article is withdrawn from the molten zinc bath. The various alloy layers contain different amounts of iron, with the highest iron content in the layers closest to the steel. The outer zinc
Figure 5: Cross-section of typical (bright) galvanized coating showing the various zinc – iron alloy layers (200 £ ).
94
T.J. Langill and B. Dugan
layer, which is dragged out with the product, is essentially the same composition as the zinc bath. This outer layer gives the galvanized product its distinctive shine and spangled appearance. Typical properties and characteristics of the alloy layers are given in Table 1. This data (and also Fig. 5) indicate that, while the outer pure zinc layer is very soft and so can be easily scratched, the underlying zeta and delta layers are somewhat harder than typical low carbon structural steels. This provides the coating with a significant abrasion resistance, thereby allowing the coating to be used in situations where abrasive wear is to be expected. It also allows the galvanized article to be handled, transported and fabricated in much the same way as ordinary steel. Galvanized coatings do not necessarily contain all of the alloy layers shown in Fig. 5. Depending on the steel chemistry and the processing conditions, the coating may contain only one or two of the layers. For example, the microstructure of the coating on silicon-containing steel shown in Fig. 6 consists almost entirely of enlarged zeta crystals, due to the reactive nature of the steel. Also, the zeta crystals have grown to the surface of the coating and have consumed the outer pure zinc layer. This coating would have a dull gray surface appearance. Similarly, when galvanized steels are heated (annealed) at temperatures above about 430–4508C, the growth of the zeta phase is accelerated, which can result in the complete disappearance of the eta layer at the surface. This effect is utilized in the manufacture of galvannealed products where improved painting, weldability and forming of steel sheet and strip products are obtained by converting the coating fully to an iron–zinc alloy. Although galvanized coatings may have a variety of microstructures, essentially no change occurs in the corrosion resistance of the coating. As previously noted,
Table 1: Characteristics of alloy layers in hot-dip galvanized coatings on steel [4]. Layer
Alloy
Eta (h)
Zinc
Zeta (z)
FeZn13
Composition Melting Crystal (%Fe) point (8C) structure
Hardness Characteristics (DPN)
0.03
419
Hexagonal
5.7– 6.3
530
Monoclinic 175 – 185
530 –670 670 –780
Hexagonal 240 – 300 Cubic
, 1530
Cubic
7– 11 Delta (d) FeZn7 Gamma Fe3Zn10 20– 27 (g) Base Iron 98– 99 steel
70 – 72
150 – 175
Soft and ductile Hard and brittle Ductile Thin, hard and brittle Ductile
Zinc Materials for Use in Concrete
95
Figure 6: Galvanized coating on silicon-containing steel showing excess growth of the zeta alloy layer through coating (200 £ ).
the extent of corrosion protection is a function of coating thickness, not coating structure. As such, the service life of bright, shiny coatings is similar to those with a dull gray appearance.
4.3.2. Coating Appearance The surface appearance and the coating thickness of the galvanized coating can be affected by a number of variables including: * * * * *
the alloy content (chemistry) of the steel; the roughness of the steel surface (e.g. whether blast cleaned); the temperature of the galvanizing bath; the immersion time of the article in the galvanizing bath; and control of the cooling rate after galvanizing (e.g. by water quenching or air cooling).
Of all these factors, the alloy content of the steel has the greatest influence on the structure and appearance of the coating and the two major alloying elements in the steel that cause problems during galvanizing are silicon and phosphorus. Both these elements act as catalysts during the galvanizing process, resulting in rapid growth of the iron–zinc alloy layers of the coating. Silicon is a strong deoxidizing element and is used in steel-making (to about 0.40% in fully killed steels) to promote sound, dense ingots and for continuously
96
T.J. Langill and B. Dugan
cast products. Silicon significantly strengthens and stiffens steel without an appreciable loss in ductility. When galvanizing such steels, however, very thick coatings are obtained since, as the work is withdrawn from the galvanizing bath, the alloy layers in the coating continue to grow through the outer zinc layer such that virtually no free zinc remains. The resulting galvanized coating is dull gray in color with a rough surface and may be quite brittle. The life of this coating is, however, proportional to its increased thickness and is unaffected by the appearance of the coating, provided the coating is sound and continuous. The effect of silicon on the galvanizing reaction is generally known as the Sandelin effect after the original researcher [6]. The effect of silicon can be seen in Fig. 7, which shows, as a function of immersion time in a zinc bath at 4508C, the change in coating weight (measured in oz/ft2) for a silicon-killed steel containing 0.26% Si and an unkilled steel with a residual silicon content of 0.015%. As indicated, the thickness of galvanized coatings developed on silicon-killed steels increases as a linear function of immersion time, whereas unkilled steels show a rapid initial development of the coating mass but which diminishes significantly with time. Aluminum is another alloying element commonly used in steel making as a deoxidizer. When galvanizing aluminum-killed steels, however, the rate of growth of the coating is similar to that of unkilled steels. Fig. 8 shows the dramatic effect that the silicon content of the steel can have on the development of the zinc coating, both in terms of the coating thickness and its metallurgical structure. These data show a preliminary peak in the coating mass at
Figure 7: The effect of silicon-killed steel on the galvanizing reaction.
Zinc Materials for Use in Concrete
97
Figure 8: Variation in coating weight and coating structure for different silicon-containing steels immersed at 4608C (after AGA).
a silicon content of about 0.065% Si, indicating that steels with silicon content at or near this peak reactivity are difficult to control during the galvanizing process, resulting in overly thick coatings. Fortunately, most silicon-containing steels that are fully killed have a silicon content of about 0.2%, which positions the steel at the valley of the Sandelin curve, thereby making the steel somewhat more controllable during galvanizing. The addition of nickel to the galvanizing bath in the range 0.06 –0.08% controls the more reactive steels and has minimal effect on the less reactive steels, such that coating thickness specification minima can be maintained. This variation of the hotdip process is called Technigalva. Experience has shown that the extra cost involved in the use of alloyed zinc for the bath (about 10% cost premium) can be offset by the slight reductions in coating thickness on conventional steels as well as significant savings and improvement in the coating quality on reactive steels. This is particularly the case over the range 0.05– 0.15% Si that coincides with the preliminary Sandelin peak. Above about 0.3 – 0.4% Si, the Technigalva process has less effect although the coatings that form on such steels are coherent, often slightly thinner than if nickel was not present and give good service performance [2]. Phosphorus, which is generally carried into steel from the raw materials used in iron making, also has an adverse effect on the formation of the galvanized coating. Phosphorous increases the strength and hardness of steel but, at the same time,
98
T.J. Langill and B. Dugan
markedly reduces its ductility and impact resistance. Accordingly, most structural and constructional steels normally specify a maximum phosphorus content of 0.04%, although actual levels are usually controlled to less than 0.025%. It is to be noted, however, that phosphorous additions of 0.10–0.12% are made to free machining steels to improve their cutting properties. Phosphorus, working alone or in conjunction with silicon, can cause rapid development of the galvanized coating. Normally, these steels are not a problem unless rephosphorized steel is used or if the phosphorus content exceeds the maximum specified level of 0.04%. Variations such as this in the thickness and appearance of galvanized coatings highlight one of the most significant problems faced by galvanizers when processing steels of uncertain origin and/or unknown chemistry. While bright coatings are the preferred outcome when galvanizing, if reactive steels are galvanized, the coating will usually be quite thick and gray in appearance. These variations become particularly obvious when different steels, some of which may be reactive, are mixed in the same galvanizing batch. Since the galvanizer cannot normally control the morphology of the finished coating on reactive steels, they cannot be held responsible for gross variations in the coating thickness or appearance that may result.
4.4. Fabrication of Galvanized Products The general fabrication of steel products to be galvanized (e.g. welding, cutting, punching, drilling, etc.) should ideally be done prior to galvanizing. Postfabrication galvanizing provides coating protection to all edges and joints and takes full advantage of the corrosion protection afforded by the zinc coating. It also avoids unnecessary damage to the coating and minimizes the exposure of unprotected edges. Although damage to the coating can be repaired by applying zinc-rich paints or zinc solders, the repair is never as good as the original coating, nor will it last as long. There are, of course, many situations in which galvanized products need to be fabricated, formed or assembled either in the workshop or in the field. In these situations, some damage to the coating must be expected but with due care the severity of this damage can be reduced. The contractor should thus be aware of the properties and the limitations of the coating. When processing reinforcing bar, it is generally most convenient and economical to galvanize straight lengths of reinforcing bar with all fabrication being done after galvanizing. During fabrication of galvanized bar, the tendency for cracking and flaking of the galvanized coating in the area of the bend increases with bar diameter and severity and rate of bend. Damage to the coating can be minimized by using large bend diameters and appropriately sized mandrels and formers. On the whole, the methods used for the handling, fabrication and
Zinc Materials for Use in Concrete
99
Figure 9: Galvanized steel products for use in concrete. (See colour plate 5.)
transportation of galvanized reinforcement are similar to those used for traditional steel reinforcement and no special requirements or techniques need be considered. As an alternative to fabricating straight bars after galvanizing, pre-fabricated bars bent to special configurations or complete cage sections (e.g. spiral column reinforcement or formed mesh) can be galvanized. This offers the distinct advantage that little or no damage to the coating will occur as may be the case with normal fabrication practices. Examples of a variety of steel products commonly used for reinforcement in concrete are shown in Fig. 9.
4.4.1. Bending Products that have been (batch) galvanized after-fabrication exhibit different bending characteristics than sheet-galvanized products. This is due primarily to differences in the coating thickness and the coating structure that is developed on
100
T.J. Langill and B. Dugan
each of the products. The coating thickness on sheet-galvanized products is much thinner than that on batch-galvanized products. Most galvanized sheet products have either entirely a pure zinc coating or a coating that is totally alloyed as in galvannealed coatings. Both types of coating have excellent bending properties; the pure zinc coating stretches during forming operations, while galvannealed coatings develop small cracks to relieve the bending stresses. In contrast, the coating structure on batch-galvanized products such as reinforcement is typically a combination of alloy layers and structures, as shown in Fig. 6. During bending, the outer pure zinc layer is stretched, while the alloy layers relieve the stresses generated by cracking. Flaking of the coating can occur if the bending is too severe. As a general rule, products that have an excessively thick coating (. 250 mm) should not be bent.
4.4.1.1. Bending Prior to Galvanizing Good fabricating practices state that when bending steel bar prior to galvanizing, a minimum 3d bend diameter should be used, where d is the bar diameter. Bending to this size minimizes the damaging effects of cold working of the microstructure of the steel that could result in strain age embrittlement during galvanizing. Ordinary, hot-rolled bar with a yield stress of about 250 MPa, or higher strength reinforcing bar (.400 MPa) strengthened by either quench and temper processing (e.g. Tempcore) or micro-alloying, which is bent prior to galvanizing, remains ductile after bending, thereby allowing limited straightening and re-bending. Bending prior to galvanizing will result in a superior product since damage to the coating due to cutting and forming operations is totally avoided. This also applies to the galvanizing of pre-fabricated and welded reinforcing elements such as column reinforcement and pre-cast panel reinforcement. 4.4.1.2. Bending after Galvanizing For practical reasons, bending is normally undertaken after galvanizing. The transportation and processing of bundles of straight bars is easier and more economical, and is preferred by most galvanizers. Special handling of bent pieces (stirrups, ties, hooks, etc.) that may involve unbundling, tagging and rebundling is not required. As a result, pre-bent pieces cannot be lost or misplaced during handling and storage and scheduling delays are reduced. Although the bendability of most galvanized bar is only marginally altered from that of uncoated bar, to minimize cracking of the galvanized coating, the following minimum bend diameters (for 908 bends) are generally recommended: * *
up to 16 mm bar diameter — 5d bend; greater than 16 mm bar diameter — 8d bend.
Zinc Materials for Use in Concrete
101
Any forming operation, including bending, may cause some cracking and flaking of the galvanized coating. This is particularly so if bend diameters less than the recommended minima are used, and if re-bending or straightening of bent bars is required. Any damage to the coating should be repaired using a suitable zincrich paint as recommended in most galvanizing standards. The use of heat for bending or re-bending galvanized reinforcing bar should be avoided due to the possibility of the zinc coating causing liquid-metal embrittlement. Similarly, should welding of galvanized bar be required, the galvanized coating should first be removed by pickling, grinding or grit blasting (refer Section 4.2.3).
4.4.2. Abrasion and Impact Resistance of Galvanized Coatings As previously noted, the zeta and delta alloy layers of the coating are harder than many structural grade steels (see Fig. 5). These alloy layers offer excellent resistance to abrasion and mechanical damage during heavy loading and severe service conditions. The relatively soft eta layer has good impact resistance and, although this may be removed by mechanical damage, exposure of the harder underlying layers generally resists further abrasion.
4.4.3. Corner and Edge Protection Since galvanizing is a total immersion process, all areas of the product are coated including those that are hidden or hard to reach such as recesses and internal surfaces. Galvanizing naturally produces coatings that are at least as thick at corners and edges, and sometimes thicker, as on other parts of the product (see Fig. 10). Due to the vectorial growth of the alloy, the coating does not thin out on edges and corners as do paint or spray applied coatings. Since coating damage is most likely to occur at these regions during handling, transportation and fabrication, the extra thickness affords additional protection where it is most needed.
4.4.4. Welding of Galvanized Steel Materials that have been galvanized, including reinforcement, may be satisfactorily welded by all common welding techniques. Although welding can be accomplished by welding through the galvanized coating, the preferred method is to remove the zinc coating in the region of the weld, generally by grinding or grit blasting, and directly weld the exposed base metal. In general, anything that can be
102
T.J. Langill and B. Dugan
Figure 10: Corner protection afforded by hot-dip galvanizing.
welded before galvanizing can be welded after galvanizing, although some minor changes to the welding technique need to be incorporated to ensure full weld penetration. These changes are primarily intended to allow the galvanized coating to burn off at the front of the weld pool. Welding is generally carried out at a slower travel speed than is usual for bare steel. In areas where the zinc coating has been removed to facilitate welding, and where the heat of welding has damaged the remaining coating, repairs should be undertaken by one of the methods described in Section 4.4.5. Before attempting the repair, however, all welding slag, spatter and other debris should be removed from the weld and adjacent surfaces. For normal flat (or butt) welds on galvanized steel, the welding current can remain the same as on bare steel. For fillet welds, however, the current may need to be increased by about 10 A. Butt welds may require a slightly wider gap since the penetration of the weld for galvanized steel is less than for uncoated steel. Travel speeds for the root pass should be reduced by 10–20% and the electrode drag angle should be increased. All these items are intended to increase the weld penetration and to stabilize the arc that can be disturbed by the evolving zinc vapor. When galvanized steel is welded, copious clouds of zinc oxide fume are produced. If inhaled in sufficient quantity, the fumes can result in so-called metal fume fever or zinc chills and, in severe cases, vomiting can occur. These flu-like symptoms are of short duration and typically pass within 24 h. Adequate ventilation or fume extraction should be used and the welder’s head should never
Zinc Materials for Use in Concrete
103
be in the fume cloud. If adequate ventilation is not possible, the welder should be fitted with a respirator.
4.4.5. Repair of Galvanized Coatings Galvanized coatings that have been mechanically damaged or when welded can be repaired using one of the following three methods [7]. A standard practice for repair of damaged hot-dip galvanized coatings using these methods is contained in ASTM A780 [8].
4.4.5.1. Zinc-Based Solders Common solders used for repair include zinc–tin –lead, zinc–cadmium and zinc– tin– copper alloys. The solder is applied in stick or powder form to a prepared and pre-heated surface. Surface preparation can be by wire brushing, light grinding or mild blast cleaning. The surface to be repaired must be free of grease and solid matter. A paste or liquid flux is applied as the surface is heated, generally with a gas torch, to a temperature of about 3008C. Caution must be taken while heating to prevent oxidizing the exposed steel or damaging the surrounding galvanized coating. The molten solder is spread with a knife or spatula, then wiped with a wet cloth to remove the flux residue. Solders are not suited to the touch-up of large areas and the resultant coating is inherently quite thin. 4.4.5.2. Zinc-Rich Paint The application of a zinc-rich paint is the most rapid and convenient means of repair. Zinc dust paints must contain between 65 and 69% zinc by weight or greater than 92% metallic zinc in the dry film. The paints are classified as either organic or inorganic, depending on the nature of the binder. Inorganic paints are particularly useful for repair and touch-up that may spread over undamaged galvanized areas. Paints are applied by brush or spray over a surface that has been prepared to a near-white finish. Thorough surface preparation is important for good adhesion. A total film thickness of 100 mm is usually specified for optimum performance and thickness measurements are taken to ensure that the required coating is applied. 4.4.5.3. Zinc Metallizing Sprayed zinc (or metallizing) should be applied to a surface that has been cleaned to a white metal finish. Zinc wire or zinc powder can be used to feed the metalspraying guns. The zinc used for spraying is normally 99.5% pure although zincaluminum alloys can also be used. The sprayed coating should be applied as soon as possible after surface preparation (certainly within 4 h) and before visible
104
T.J. Langill and B. Dugan
deterioration of the surface has occurred. Adhesion of the zinc spray to the base metal is by mechanical means and is dependent on the quality of the surface preparation and the extent of cleaning. The coating is normally applied to a thickness equivalent to that of the undamaged coating and measurements should be taken to ensure the required coating has been applied.
4.5. Zinc Inhibitors Inhibitors are chemicals that react with, or precipitate at, the metal surface, thereby significantly reducing the corrosion rate of the metal. Extensive research over many years has shown that many substances in solution can inhibit the corrosion of zinc [9]. Some inhibitors operate by the precipitation of almost insoluble compounds at the metal surface (adsorption inhibitors) and for zinc in water, alkali silicates, phosphates and borates fall into this category. Other inhibitors act to passivate the metal surface (anodic inhibitors) via the formation of tight and adherent protective layers as a result of some corrosion of the metal surface. In this case, corrosion may initially occur quite rapidly, but the rate of corrosion diminishes to very low levels (after a short time) once the passivating film has developed. For zinc in water, chromates, alkaline nitrates and phosphate polymers act as passivating agents. Inhibitors can be applied directly to freshly galvanized products to protect them during shipping or storage, or they may occur naturally in the environment where the galvanized product is placed. For protection of zinc coatings during shipping and storage, chromates and molybdates have been shown to be very effective and are the basis for many conversion coatings. These conversion coatings help retard the formation of wet storage stain on materials that are stacked or tightly nested prior to installation. Chromate salts, in particular, have been very widely used as a passivating chemical for zinc and galvanized steel. Since chromates are anodic inhibitors, their concentration should be kept above about 0.5% in solution in order to minimize the risk of pitting. Because of environmental concerns and issues relating to occupational health and safety, the continued use of chromates is likely to be severely restricted in the future. Many other materials and compounds can be effective in passivating zinc; those which are effective in either water or alkaline solutions are given in Table 2. Many inhibitors also occur naturally in water and help to passivate the zinc surface. Seawater, for example, contains sodium chloride that is aggressive but it also contains magnesium salts that act as a natural inhibitor for zinc. The lack of natural inhibitors can therefore give misleading results in laboratory tests, since such tests typically use pure chemicals. For this reason, caution should always be exercised when interpreting the results of accelerated corrosion tests.
Zinc Materials for Use in Concrete
105
Table 2: A selection of inhibitors for zinc in water or alkaline solutions [9]. Agar –Agar Amino chromate Benzotriazole Borax Chromate or dichromate Hexametaphosphate Lanolin Maleic hydrazide
Monochromate Phosphate Phosphate glasses Picrate Silicate Sodium benzoate Sodium dodecamolybdophosphate Zinc salts
4.6. Codes of Practice and Standards The regulation of the hot-dip galvanizing of steel reinforcing bars is handled in different ways around the world. Some countries treat steel reinforcing bars in the same way as any another steel products and so the hot-dip galvanizing of reinforcement falls under a general galvanizing standard. This approach presently applies in countries such as Australia, New Zealand, Canada, United Kingdom, South Africa and Sweden, who all use their national standard for hot-dip galvanizing to cater for the galvanizing of steel reinforcing bars and other products. A selection of such standards is given in Table 3. In all these general galvanizing standards, an average minimum thickness (or mass) of the coating is specified, depending on the type and thickness of the base material. For structural sections heavier than 5–6 mm thick, which would include reinforcement and most other reinforcing products, a minimum average coating thickness in the range 600– 610 g/m2 is specified, which equates to a coating thickness of 85 – 87 mm. Similar requirements are contained in ISO 1461 [10]. This international standard has either replaced or is read in conjunction with national standards in a number of countries, as also shown in Table 3. In ISO 1461, a minimum average coating thickness of 85 mm is specified, with a minimum local coating thickness of 70 mm, for steel sections greater than 6 mm thick. Several other countries, including the United States, France, Germany, Italy and India, have developed product-specific standards for the hot-dip galvanizing of steel reinforcing bars. These are also listed in Table 3. In the USA, ASTM A767 [11] deals specifically with zinc-coated (galvanized) steel reinforcement for concrete and specifies two classes of coating mass (classes I and II) based on the actual area of the bar. The minimum weight of the zinc coating for Class I coatings is 1070 g/m2 (equivalent to a coating thickness of 150 mm) and 610 g/m2 for Class
106
T.J. Langill and B. Dugan
Table 3: Standards for hot-dip galvanizing of reinforcing bar. Designation
Title of Standard
General Galvanizing Standards Australia/New Zealand AS/NZS 4680 Canada CAN/CSA G164
After-fabrication hot-dip galvanizing Hot-dip galvanizing of irregularly shaped articles South Africa SABS/ISO 1461 Hot-dip galvanized coatings on fabricated iron and steel articles Sweden SS-EN ISO 1461 Hot-dip galvanized coatings on fabricated iron and steel articles United Kingdom BS EN ISO1461 Hot-dip galvanized coatings on fabricated iron and steel articles International Standards ISO 1461 Hot-dip galvanized coatings Organization on fabricated iron and steel articles Reinforcing Steel Standards United States ASTM A 767 United Kingdom
BS ISO 14657
France
NF A35-025
Italy
UNI 10622
India
IS 12594
International Standards ISO/DIS 14657 Organization
Zinc-coated (galvanized) steel bars for concrete reinforcement Zinc-coated steel for the reinforcement of concrete Hot-dip galvanized bars and coils for reinforced concrete Zinc-coated (galvanized) steel bars and wire rods for concrete reinforcement Hot-dip coatings on structural steel bars for concrete reinforcement specifications Zinc-coated steel for the reinforcement of concrete
II coatings (86 mm). While a clear definition of different product types is not given in this standard, it is usual to assume that a Class I coating applies to the heavier structural bars, while Class II applies to the lighter structural bars and to architectural or non-load bearing bars. Guidance is also given on galvanizing before or after fabrication of the bars, the finish and adherence of the coating, inspection and rejection criteria, and (when specified) repair of damage to the coating as a result of bending. ASTM A767 also specifies that the galvanized coating shall be chromate treated so as to preclude any reaction between the bars and the fresh Portland cement paste. This is normally accomplished by quenching freshly galvanized bars in an
Zinc Materials for Use in Concrete
107
aqueous solution containing at least 0.2 wt% sodium dichromate or 0.2% chromic acid. The quenching solution should be at a temperature of at least 328C and the reinforcement immersed for at least 20 sec. If the bars are at ambient temperature, 0.5 –1.0% sulphuric acid should be added as an activator of the chromate solution. The standards also indicate that some cracking and flaking of the coating in the area of the bend shall not be the cause for rejection and that such damage is not subject to repair unless ordered in accordance with the supplementary requirements of the specification. Should repairs to the coating be required, a zinc-rich formulation is used, generally an organic zinc-rich paint, containing a high proportion of metallic zinc in the dry film. Although not specifically required, it is generally accepted that cut ends of galvanized bars should be repaired. The International Standards Organization (ISO) has recently moved to develop an international standard incorporating a general set of zinc-coating requirements for galvanizing steel reinforcing bars [12]. The draft standard, ISO/FDIS 14657 Zinc-coated steel for the reinforcement of concrete, was at the final approval stage in mid-2004 and publication might be expected towards the end of the year.
4.7. Field Handling Techniques In the following, a brief summary is given of techniques for the field handling of hot-dip galvanized reinforcing steel. This has been compiled by the American Galvanizers Association and, while not meant to be fully inclusive, is presented as an industry guide to best practice [13]. Other issues in field handling may arise from time to time and these should be assessed in the light of the general recommendations given here.
* *
* * *
*
Material receipt and inspection: visually inspect for damage; check for secure tie-downs on transport. Unloading and job site handling: no special handling or care necessary; lift bundles at multiple pick-up points; or use a spreader bar with additional nylon straps to prevent sag and bar-to-bar abrasion in longer bundles. Storage: block material and store on a slant to allow for water drainage and air flow.
108
*
*
* *
* *
*
* *
*
*
T.J. Langill and B. Dugan
Placement: no special care necessary. Bar supports and tie wire: bar supports, spacers and reinforcement supports should all be hot-dip galvanized; 16.5 gauge or heavier galvanized tie wire should be used; other acceptable materials for these parts are plastic or non-conductive coated steel. Splicing and coupling details: a bar-lock coupler is recommended, either galvanized or stainless; for welded splices, all welds must be touched up as recommended in ASTM A780; use appropriate protective masks and suitable ventilation when welding. Field cutting: field cutting should be avoided; repair of cut ends shall be done using touch-up procedures from ASTM A780. Final inspection and repair: touch-up of cut and burned ends should be done following procedures recommended in ASTM A780. Concrete pour: no special handling or care necessary.
References [1] American Galvanizers Association. (2000). Galvanizing for corrosion protection: a specifier’s guide. American Galvanizers Association, Englewood, CO, USA. [2] Porter, F. (1991). Zinc handbook: properties, processing and use in design. Marcel Dekker, New York, Chapters 2, 6, 10. [3] Galvanizers Association of Australia. (1999). After-fabrication hot-dip galvanizing. Galvanizers Association of Australia, Melbourne, Vic., Australia, 15th ed. [4] American Society for Metals. (1987). Handbook, volume 13: corrosion. ASM International, Materials Park, OH, USA, pp. 432–445. [5] ASTM B6. (1998). Standard specification for zinc. American Society for Testing and Materials, Philadelphia, PA, USA. [6] (a) Sandelin, R. W. (1940). Galvanizing characteristics of different types of a steel. Wire and Wire Products, 15, 11, 655–676; (b) Sandelin, R. W. (1940). Galvanizing characteristics of different types of a steel. Wire and Wire Products, 15, 12, 1940, 721– 749; (c) Sandelin, R. W. (1940). Galvanizing characteristics of different types of a steel. Wire and Wire Products, 16, 1, 1941, 28–35.
Zinc Materials for Use in Concrete
109
[7] American Galvanizers Association. (2001). The inspection of products to be hot-dip galvanized after fabrication. American Galvanizers Association, Englewood, CO, USA. [8] ASTM A780. (1993). Specification for repair of damaged and uncoated areas of hot-dip galvanized coatings. American Society for Testing and Materials, Philadelphia, PA, USA. [9] Porter, F. (1994). Corrosion resistance of zinc and zinc alloys. Marcel Dekker, New York, pp. 277–279. [10] ASTM A767M. (1997). Standard specification for zinc-coated (galvanized) steel bars for concrete reinforcement. American Society for Testing and Materials, Philadelphia, PA, USA. [11] ISO 1461. (1999). Hot-dip galvanized coatings on fabricated iron and steel articles. International Standards Organization, Geneva, Switzerland. [12] DIS 14657. (2001). Zinc-coated steel for the reinforcement of concrete. International Standards Organization, Geneva, Switzerland. [13] American Galvanizers Association. (2000). Field handling guide: hot-dip galvanizing versus fusion bonded epoxy. American Galvanizers Association, Englewood, CO, USA.
This Page Intentionally Left Blank
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 5
Electrochemical Aspects of Galvanized Reinforcement Corrosion Carmen Andrade and Cruz Alonso Institute of Construction Science, Spain
5.1. Introduction Steel in concrete is protected against corrosion by a passivation mechanism. The reason for this passivation is the high alkalinity of the concrete pore-water solution, the pH of which in hydrated concrete is greater than 12.5. During the first short period after mixing, the solution filling the pores of concrete is over-saturated in Ca(OH)2. Later, equilibrium is reached with other species such as NaOH, KOH and CaSO4·2H2O. The pH value of this aqueous solution varies from 12 to 14 as a function of the alkali content of the cement and the degree of hydration [1]. One particular feature of these alkaline solutions, which is relevant for the behavior of zinc, is that the concentration of Ca2þ ions decreases when the pH increases [2]. Fig. 1 shows the variation of Ca2þ concentration as a function of pH for cement pastes with water/cement (w/c) ratio of 0.5. It is possible to distinguish at the lower pH, typical of the early stages of hydration, that the Ca2þ concentrations fall above the lines of Ca2þ equilibrium, which indicate oversaturation in the solution. The presence of Ca2þ is needed for the passivation of zinc in alkaline solutions, as will be explained further. Under the highly alkaline conditions of concrete, a microscopic oxide layer is formed on the steel surface of the reinforcement, the so-called passive film. This passive film impedes the dissolution of the iron and so the corrosion of steel reinforcement is severely limited, even in the presence of moisture and humidity.
5.2. Galvanized Reinforcement as a Protection Method 5.2.1. Corrosion of Galvanized Steel in Alkaline Media In the case of zinc, the situation is different to steel since zinc is an amphoteric metal. This means that the zinc is stable over a wide range of pH, from
112
C. Andrade and C. Alonso
Figure 1: Variation of Ca2þ concentration as a function of OH2 for cement pastes made with w/c , 0.5 [2].
approximately 6– 12.5, but below and above these values the corrosion rate increases exponentially as shown in Fig. 2 [3]. A similar deduction comes from an examination of the Pourbaix diagram for zinc, as shown in Fig. 3 [4]. At the alkaline pH values of aqueous pore solutions, the zinc dissolves very quickly with the evolution of hydrogen since the initial corrosion potential lies below the potential for hydrogen evolution and the metal cannot be considered stable in that medium. However, the corrosion products that form of the surface of the zinc induce further passivation of the metal. The behavior of zinc in both weak and strong alkaline solutions has been widely studied and reported in the literature. An important contribution was made by Bird [5], who concluded that, below a certain OH2 content, the first anodic product is Zn(OH)2 and for pH . 12.9 the main anodic product is the soluble zincate ion (ZnO22 2 ). Another important contribution came from Grauer and Kaesche [6], who identified the formation of a thin layer of zinc oxide (ZnO) in 0.1 M NaOH solutions. Shams El Din et al. [7] studied the anodic oxidation of zinc in 0.001–1.0 M NaOH solutions. They also established two types of behavior; in solutions more dilute than 0.3 M, gradual precipitation of Zn(OH)2 occurs while, in concentrations greater than 0.3 M, the zinc dissolves as zincate ions according to the reaction 2 Zn þ 4OH2 $ ZnO22 2 þ 2H2 O þ 2e
Vorkapic et al. [8] suggested different mechanisms for the passivation of zinc in concentrated KOH solutions of concentrations 1, 3, 6 and 10 M. They found a high
Electrochemical Aspects of Galvanized Reinforcement Corrosion
113
Figure 2: Corrosion rate of pure zinc as a function of the pH [3].
initial corrosion rate associated with the hydrogen evolution and a further decrease with the formation of a film of ZnO. Zembura and Burzynska [9] studied the corrosion of zinc in alkaline deaerated solutions (pH in the range 11.6–13.3) þ where the control reaction is the diffusion of ZnO22 2 and ZnHO formed at pH 11. Macias and Andrade [10] studied the stability of galvanized reinforcements in NaOH and KOH solutions in a pH range from 11 to 14, both with and without the presence of Ca(OH)2. They found that, in a pH interval between 12 and 13.2 ^ 0.1, the galvanized coating corrodes at an acceptably low rate. At pH , 12, localized corrosion takes place while, at pH . 13.2, total dissolution of the coating occurs with no passivation. They also found a pH of 12.8 ^ 0.1 as the threshold for the onset of hydrogen evolution. Several other authors have studied the corrosion of zinc in calcium-containing alkaline media, either in solution or in cement paste or concrete, that have confirmed the previous findings [11–15]. Lieber and Gebauer [16] were the first to identify the corrosion product that caused the passivation of zinc in calcium-rich alkaline solutions as calcium hydroxyzincate (CaHZn). Rehm and La¨mmke [17] published an extensive paper in agreement with the findings of Lieber, although they suggested a primary
114
C. Andrade and C. Alonso
Figure 3: Potential-pH equilibrium diagram for the zinc –water system at 258C [4].
formation of Zn(OH)2. Finally, Liebau [18] proposed a reaction mechanism of the type Zn þ 2H2 O $ ZnðOHÞ2 þ H2 2ZnðOHÞ2 þ 2H2 O þ CaðOHÞ2 ! CaðZnðOHÞ3 Þ2 ·2H2 O The final product was identified as the compound CaHZn. Further to this, ZnO and 1-Zn(OH)2 were identified as being formed during the corrosion process in these media. As previously noted, Macias and Andrade [10,19] found that galvanized steel become passivated at a pH value , 13.2 ^ 0.1. They confirmed that the passivating film was calcium hydroxyzincate, but also observed that its morphology varied with the pH of the solution in which it formed [20]. The existence of the previously mentioned insoluble corrosion products ZnO and Zn(OH)2 was also confirmed, although no passivating properties were observed with the ZnO product. Once the passive film of calcium hydroxyzincate is formed, its stability is not altered if a further increase of the pH occurs, even if the pH increases to a value of 13.6 ^ 0.1 [20,21]. Regarding the morphology of the passivating layer, scanning electron microscopy analysis [19,22] of the corrosion products of galvanized steel in calciumrich alkaline solutions shows that, when the pH is around 12.6, the surface is
Electrochemical Aspects of Galvanized Reinforcement Corrosion
115
Figure 4: CaHZn crystals after 24 h in saturated Ca(OH)2 solution (pH ¼ 12.6) (160 £ ).
totally covered after 2 or 3 days, mainly by CaHZn crystals, as shown in Figs. 4 and 5. Its appearance is that of a very compact carpet of crystals. As the pH increases so does the CaHZn crystal size to the point that the crystals cannot completely cover the surface. As a consequence of this, small regions of the base metal are left exposed and thus without protection (Fig. 6). Under these conditions, complete passivation of the surface is not possible and dissolution continues at a high corrosion rate. At even higher pH values (above 13.5), the crystals of CaHZn become quite coarse and they grow as isolated crystals (Fig. 7). The reason for this is that at pH values above 13.2, the concentration of Ca2þ ions in solution is depleted. At these high pH values, because the formation of the passive layer of
Figure 5: CaHZn crystals after 28 days in saturated Ca(OH)2 solution (pH ¼ 12.6) (160 £ ).
116
C. Andrade and C. Alonso
Figure 6: CaHZn crystals after 24 h in saturated Ca(OH)2 þ 0.1 M KOH solution (pH ¼ 13.1) (180 £).
CaHZn crystals is impeded, the dissolution of the zinc is not retarded and as a result the galvanized coating may completely dissolve in a short period of time. Fig. 8 shows general trends in measurements of corrosion current ðIcorr Þ and corrosion potential ðEcorr Þ for the reaction of zinc in alkaline solutions [10,23,24]. From this figure, it is possible to deduce that a saturated solution of Ca(OH)2 of pH 12.6 ^ 0.1 quickly passivates the zinc accompanied by a rapid shift in Ecorr to more noble values below 21000 mV SCE, and a lowering of Icorr to about 0.1 mA/cm2. The presence of stronger alkalies, be they NaOH or KOH in solution with Ca(OH)2, will increase both the pH and the corrosion rate and lengthen the period until the passivation current ðIcorr Þ becomes less than 0.1 mA/cm2 [24].
Figure 7: CaHZn crystals after 10 days in saturated Ca(OH)2 þ 0.2 M KOH solution (pH ¼ 13.24) (20 £ ).
Electrochemical Aspects of Galvanized Reinforcement Corrosion
Figure 8: Ecorr and Icorr of galvanized steel in alkaline solutions showing the effect of further increase of the pH [24].
117
118
C. Andrade and C. Alonso
Figure 9: Influence of the pH on the corrosion product and type of attack on galvanized steel [25].
Taking into account the results of various researchers, the sequence of reactions which best seems to represent the corrosion process of zinc in strongly alkaline, calcium-containing solutions is 2 Zn þ 4OH2 ! ZnðOHÞ22 4 þ 2e
Zn þ 2OH2 ! ZnO þ H2 O þ 2e2 ZnO þ H2 O þ 2OH2 ! ZnðOHÞ22 4 2þ 2ZnðOHÞ22 þ 2H2 O ! CaðZnðOHÞ3 Þ2 ·2H2 O þ 2OH2 4 þ Ca
A summary of the expected behavior of galvanized steel in aqueous solutions in the range pH 11– 14 is given in Fig. 9 [25]. This shows the nature and ranges of formation of different corrosion products, the morphology of the attack, and whether hydrogen evolution occurs. The stability range of galvanized steel is identified in the pH range 11.5– 13.2.
5.2.2. Evolution of Hydrogen During Cement Setting As mentioned earlier, simultaneous with the processes of corrosion and passivation of galvanized steel in concrete, hydrogen evolution occurs along the bar surface. This effect can be deduced from the Pourbaix diagram for zinc (Fig. 3), which indicates that as the corrosion potential of zinc drops below line a of the diagram (i.e. when in contact with the alkaline pore concrete solution) the water is
Electrochemical Aspects of Galvanized Reinforcement Corrosion
119
hydrolized, resulting in the evolution of hydrogen following the reaction 2H2 O þ 2e2 ! 2OH2 þ H2 Soon after the first contact between the zinc coating and the alkaline media, the generation of hydrogen gas is quite vigorous. However, it steadily decreases with time due to the continued formation of the surface layer of CaHZn, which eventually becomes continuous on the surface with the hardening of the cement paste. Fig. 10 depicts an example of the time interval over which hydrogen evolution occurs [26]. The value of the corrosion potential ðEcorr Þ is taken as the indicative parameter. In the case of mortar without chlorides, the hydrogen evolution lasts 1 or 2 h only while, in the presence of CaCl2, although it induces a certain decrease in the pH value of the pore solution, the hydrogen evolution does not cease for many more hours.
Figure 10: Ecorr measurements for galvanized steel. Hydrogen evolution occurs for Ecorr , 21000 mV SCE [26].
120
C. Andrade and C. Alonso
The generation of hydrogen is common under these circumstances although the total duration of the reaction depends primarily on two factors: *
*
the chromium content of the cement, which varies considerably in different cement types and may completely suppress the hydrogen evolution [27]. The presence of chromium reduces the corrosion rate of galvanized bars and delays the passivation (Icorr , 0:2 mA/cm2); and the alkali content of the cement and therefore the pH of the pore solution.
The main consequence of the hydrogen gas evolution is the formation of pores at the bar/concrete interface due to the trapping of bubbles formed during the setting of the cement. This reduced bar/concrete contact area is thought to be responsible for some loss in bond of galvanized reinforcement when no chromium additions are made to the concrete or when separate chromate passivation of the bars is not performed [28]. However, this decrease in contact zone at the interface is only transitory, as the formation of the CaHZn crystals progressively fills these gaps. This issue will be considered further in Chapter 8.
5.2.3. Influence of Alkali Content on Passivation of Galvanized Steel The component of cement that most influences the behavior of zinc in alkaline media is the alkali content (Naþ, Kþ). These alkali ions are incorporated into the cement in different proportions [29,30] as a function of the raw materials used for cement manufacture and the sulphate content of the fuels used in manufacture. They are usually present in the form of alkaline sulphates. Different cements can produce different pore solutions due to the presence of alkali ions, which are the most soluble components and thus responsible for the final pH of the pore solution. Fig. 11 shows the corrosion rates of galvanized reinforcement embedded in mortars made with various cement types and held at 100% RH or partially immersed in water [31]. After 1 year, the corrosion rates for the different mixes differ by about one order of magnitude but the cement which induces the highest rate of attack is an Ordinary Portland Cement (OPC) concrete having the highest total alkali content. The corrosion rate is usually reduced as the alkali content of the cement is lowered. Corrosion rates higher than 0.2 mA/cm2 (which represent attack penetration of 2.2 mm/year) means that a galvanized coating about 60 mm thick will completely dissolve in less than 30 years. Fig. 12 shows the relationships between corrosion rates at particular ages (1, 28, 90 and 365 days) and the pH value of cement suspensions prepared with w/c ratios of about 1.0. In spite of the scatter of the values, the trend of the corrosion rates coincides more or less with the increasing pH values of the cements, as would be
Electrochemical Aspects of Galvanized Reinforcement Corrosion
121
Figure 11: Icorr of galvanized reinforcement embedded in mortar fabricated with different cement types and cured at 100% RH and partial immersion [31].
122
C. Andrade and C. Alonso
Figure 12: Relationship between the pH value of cement suspensions and Icorr [29].
Electrochemical Aspects of Galvanized Reinforcement Corrosion
123
expected for zinc in alkaline media [29]. As previously mentioned, however, there is a pH threshold for zinc in concrete pore solutions at a pH between 12.8 and 13.2 ^ 0.1. Above this pH limit, the corrosion rate developed is so high that it implies a risk of complete dissolution of the galvanized coating during the life of the structure [32]. This approximate relationship between the cement alkali content and the corrosion rate may explain the observed different behaviors and life of galvanized coatings in concrete. The type of cement in contact with the galvanizing is very important because it allows the formation of a compact passive layer of calcium hydroxyzincate. This effect may also explain part of the controversy and bad experiences with galvanized reinforcement in concrete when the type of cement used has not been taken into account. What is clear is that the layer of passivating corrosion products develops during the first hours after mixing, when the pH value of the pore concrete solution is lower than 12.8 ^ 0.1. This protective layer completely passivates the reinforcement. If the pH is between about 12.8 and 13.2, the passivating layer develops slowly and the galvanized coating may continue to dissolve until full passivation is reached. If the pH is greater than 13.2, the passive layer is not developed and the galvanized coating continuously dissolves until it disappears. Fortunately, as shown in Fig. 13, pH values greater than 13.2 do not develop in concrete pore solutions during the first hours after mixing if sulphate is used as a setting regulator or enough alkaline sulphates are present [2]. While sulphate ions are present in the pore solution, the pH value does not increase beyond 13.2. Only when the sulphates disappear from the solution, due to the formation of sulphoaluminates, does the pH rise to a maximum value which is a function of the total alkali content. This usually happens several hours or days after mixing, by which time the passivating CaHZn layer has all but completely formed and, as a result, the increase in pH is not harmful to the galvanized coating.
5.2.4. Influence of Galvanized Coating in the Passivation Process The structure of the galvanized coating depends on the composition of the bare steel, mainly the carbon and silicon content, and also on the temperature and composition of the hot dip bath as well as the time in the galvanizing bath. The more common galvanizing temperature is 4508C [33–35]. Full details of the galvanizing process and the structure of the alloy layer coating is considered in Chapter 4 and elsewhere. The typical galvanized coating is made up of several different layers:
124
C. Andrade and C. Alonso
2 Figure 13: Variation of SO22 4 concentration as a function of OH concentration [2].
Eta layer (h) Zeta layer (z) Delta layer (d) Gamma layer (g)
the external layer consisting of almost pure zinc. Fe/Zn alloy layer with ,7% Fe. Asymmetric monclinic crystals. Fe/Zn alloy layer with 7–11% Fe. Hexagonal structure formed in two layers with different compaction. Fe/Zn alloy layer with 20–27% Fe. Cubic structure which is very thin.
Electrochemical Aspects of Galvanized Reinforcement Corrosion
125
Figure 14: Changes of galvanized steel microstructure: (a) 4508C; (b) 4308C; (c) Al addition to the bath; (d) annealing (500 £ ).
Examples of different galvanized microstructures that can be obtained are shown in Fig. 14 [36]. It can be observed that the thickness of the pure zinc layer varies, being thicker for lower galvanizing temperatures and for Al additions to the galvanizing bath (Fig. 14a – c). On the other hand, annealing of
126
C. Andrade and C. Alonso
the galvanized coating induces the growth of the Fe–Zn alloy layers through to the outside of the coating with complete disappearance of the eta layer of pure zinc at the surface (Fig. 14d). The addition of Ni to the bath (, 0.16%) also changes the galvanized coating microstructure decreasing the thickness of the zinc alloy layers. The presence of these different microstructures has a significant effect on the stability of the galvanized coating in contact with alkaline solutions. This is primarily because it is the outer pure zinc layer (h) that provides the most effective passivation, while the underlying Fe – Zn layers are less stable particularly so in the presence of chlorides which selectively attack them, inducing progressive disintegration of the coating [37]. This effect will be considered further in this chapter. The nature of the attack on the galvanized coating during passivation is shown in Figs. 15 and 16. These indicate that the attack progresses by dissolution of the external pure zinc layer in conventional galvanized coatings while, for annealed galvanized coatings (Fig. 17), the Fe– Zn alloy layers disintegrate by selective attack in which the zinc is used to develop the calcium hydroxyzincate film. From observations of this type, it was deduced that galvanized coatings should have sufficient reserve of the pure zinc layer (thicker than around 10 mm) to enable the development of a perfect, and so passivating, film of calcium hydroxyzincate. The influence of the coating microstructure on the corrosion rate is shown in Figs. 18 and 19. Fig. 18 shows the different corrosion rates measured during passivation in Ca(OH)2 of the galvanized layer formed on bare steel of 400 or
Figure 15: Wells and tunnels in the h layer of galvanized coating after 365 days testing (200 £ ).
Electrochemical Aspects of Galvanized Reinforcement Corrosion
127
Figure 16: Dissolution of the h layer of galvanized coating (200 £ ).
500 MPa strength [38]. The difference in strength is due to the different tempers used during heat treatment of the reinforcement: for the 500 MPa steel, a fine martensitic structure develops while, in the 400 MPa steel, a coarser structure develops. This is the reason for the differences in the galvanized layers between the two steels to the extent that the total alloy layer is thinner for the 500 MPa steel than for the 400 MPa steel (90 and 135 mm, respectively). The external pure zinc layer is also thinner on the 500 MPa steel than on the 400 MPa steel (18 mm compared with 30 mm). The result is an overall higher corrosion rate for the galvanized 500 MPa steel. Fig. 19 represents the case of various types of galvanized coatings embedded in mortar specimens and kept in partial immersion. The annealed galvanized coating exhibits the higher corrosion rates while the galvanized mild steel shows the lowest [23].
Figure 17: Annealed galvanized reinforcement embedded in mortar made with OPC (160 £ ).
128
C. Andrade and C. Alonso
Figure 18: Corrosion potential and corrosion rate of different galvanized steels (400 and 500 MPa).
5.3. Behavior of Galvanized Steel in Aggressive Media When the aggressive species, either the carbonation front or chloride ions, reach the galvanized reinforcement, in order to initiate corrosion they have to disrupt the physical barrier of the calcium hydroxyzincate film.
5.3.1. Corrosion of Galvanized Steel in Carbonated Concrete The carbonation or neutralization of the cover concrete is one of the principal reasons for reinforcement corrosion. The pH of the aqueous phase changes from highly alkaline to values around neutrality (pH 7). By reference to Figs. 2 and 3, the diagrams of Roetheli and Pourbaix, respectively, at or near neutral pH the rate of corrosion of zinc is very low and it would be expected that the galvanized coating would perform quite well. As a result, it is generally noticed that galvanized steel does not corrode in carbonated concrete. Experience has shown [39 – 41] that carbonation does not significantly increase the corrosion rate of galvanized bars in concrete and in some cases it is even reduced. This effect is shown in Fig. 20 where the corrosion rate of
Electrochemical Aspects of Galvanized Reinforcement Corrosion
Figure 19: Ecorr and Icorr of various types of galvanized steel embedded in OPC mortar.
129
130
C. Andrade and C. Alonso
Figure 20: Evolution of Icorr of galvanized reinforcement embedded in carbonated and uncarbonated mortar, with and without chlorides [40].
galvanized reinforcement embedded in carbonated and uncarbonated mortars, along with sequential changes in relative humidity, are plotted. From this data it can be seen that, before the concrete is carbonated, the galvanized steel shows high corrosion rates due to the coating being consumed in forming the protective layer of CaHZn [40]. During carbonation, the galvanized coating may depassivate with a consequent rise in the corrosion rate, but later (after 6 days in Fig. 21) a sharp decrease in the corrosion rate occurs [42]. A new passivated layer is formed in this condition, probably due to the precipitation of zinc carbonates on the surface.
Electrochemical Aspects of Galvanized Reinforcement Corrosion
Figure 21: Changes in Ecorr and Icorr of galvanized steel during carbonation of alkaline solutions.
131
132
C. Andrade and C. Alonso
5.3.2. Corrosion of Galvanized Steel in the Presence of Chlorides Chlorides are the more aggressive ions for reinforced concrete and are the most frequent cause of reinforcement distress. The chlorides are present in the concrete from two sources: first, from the mixing as part of the raw materials (water, aggregates or as an admixture) and, second, if they penetrate from marine environments or from the use of deicing salts. In both cases, the attack produced on the reinforcement is localized, resulting in a reduction of the cross-section of the reinforcement. For bare steel, corrosion initiates when a threshold value of Cl2 in the pore solution is reached. This threshold concentration depends, among other factors, on the pH and increases as the pH increases. In contrast, there is not universal agreement on the resistance to chloride attack of galvanized reinforcement. What seems clear, however, is that zinc is also attacked by chlorides [43], although a higher chloride threshold is needed [44]. The overall behavior depends on the origin of the chlorides and the state of the galvanized surface. Fig. 22 represents a comparison of the range of potential and chloride ion concentrations in Ca(OH)2 saturated solution at which zinc shows stability of
Figure 22: Corrosion of zinc in the presence of chloride ions [44].
Electrochemical Aspects of Galvanized Reinforcement Corrosion
133
the passive state or pitting corrosion [44]; in effect, these are the potential and chloride ion concentration at which film breakdown and anodic dissolution occur. From Fig. 22, it can be deduced that zinc becomes susceptible to pitting attack in Ca(OH)2 solutions polluted with chlorides at chloride ion concentrations of approximately 0.45 M, while steel corrodes when the chloride ion concentration exceeds 0.08 M. Arliguie [11] studied the increase in chloride threshold for galvanized reinforcement and found that it is pH dependent following the expression: Er ¼ 2965 – 485 log½Cl2 However, the dissolution of the galvanized coating differs depending on the state of the galvanized surface. If the coating exhibits a pure zinc layer, it dissolves uniformly at first and only later does localized attack occur in the alloyed layers. Figs. 23– 25 show such behavior for galvanized steel exposed to several chloridecontaining environments. In Fig. 25, it can also be observed that zinc induces a level of cathodic protection as no attack on the exposed section of the base steel is evident.
5.3.2.1. Chloride Ions Added During Mixing The addition of chloride in the mixing water produces changes in the pH of the pore solution. Thus, while the CaCl2 produces a decrease of the pH, NaCl does not alter it significantly [45 – 48]. In fact, when calcium chloride is added to mortar,
Figure 23: Corrosion of galvanized steel (4508C) embedded in OPC mortar with 2% CaCl2 addition at 100% RH (160 £ ).
134
C. Andrade and C. Alonso
Figure 24: Generalized attack of the zinc layer of galvanized steel (400 MPa) after 40 days in 0.5 M NaCl (50 £ ).
the initial corrosion rates can be smaller than without admixtures, as shown in Fig. 26 [23], although alternative periods of activation and passivation occur before complete passivation is reached. In the case of NaCl, Fig. 27 indicates behavior which, for pure zinc bars, shows higher corrosion rates than with CaCl2, at least in the initial stages of corrosion. This effect has been attributed to the higher pH values produced by NaCl additions. Later, however, the passivity state seems more perfect with NaCl than with CaCl2. Furthermore, as previously mentioned, the galvanized microstructure also has an important influence, as shown in Fig. 28. The absence of the external zinc layer produces much higher corrosion rates while the alloy layers are less resistant to chloride attack than the pure zinc.
Figure 25: Localized attack of galvanized steel (400 MPa) alloy layers after 175 days in 0.5 M NaCl (50 £ ).
Electrochemical Aspects of Galvanized Reinforcement Corrosion
135
Figure 26: Ecorr and Icorr of galvanized steel without h layer embedded in cement mortar with 0, 1 and 2% CaCl2 additions.
136
C. Andrade and C. Alonso
Figure 27: Ecorr and Icorr of zinc in alkaline solutions and with CaCl2 addition.
Electrochemical Aspects of Galvanized Reinforcement Corrosion
137
Figure 28: Ecorr and Icorr of various types of galvanized steel embedded in OPC mortar with 2% CaCl2 addition.
138
C. Andrade and C. Alonso
In consequence, the most resistant galvanized coatings are those with a thicker external layer of pure zinc. When plain zinc bars are used, low corrosion rates are measured [36]. On the other hand, galvannealed coatings are catastrophically destroyed when they are used in chloride-containing concrete. In summary, then, the more corrosion-resistant part of a galvanized coating is the pure zinc outer layer and the more vulnerable is the underlying alloy layers.
5.3.2.2. Chloride Ions Penetrating from Outside The resistance of galvanized reinforcement against chloride penetration depends on the compactness of the CaHZn layer and on the microstructure of the remaining coating. By the time the chlorides reach the reinforcement, the CaHZn layer should have already been formed. Therefore, if it is compact and continuous, and the remaining coating has a thick enough pure zinc layer to resist pitting attack, the galvanized coating will resist chloride attack quite well, as shown in Figs. 29 and 30 [45,49].
Figure 29: Evolution of Icorr for galvanized steel (400 MPa) in the presence of chlorides.
Electrochemical Aspects of Galvanized Reinforcement Corrosion
139
Figure 30: Attack on galvanized steel (400 MPa) after 60 days in saturated Ca(OH)2 þ 0.5 M NaCl solution (50 £ ).
Figure 31: Evolution of Icorr for galvanized steel (500 MPa) in the presence of chlorides.
140
C. Andrade and C. Alonso
However, when, instead of a thick pure zinc layer (as for the 400 MPa steel), it is so thin (as for the 500 MPa steel) that it is consumed in forming the CaHZn layer, once the chlorides reach the reinforcement they find, below the CaHZn layer, only the zinc– iron alloy layers that allow a much higher rate of attack. This effect is shown in Fig. 31. If the level of chlorides continues to increase, the threshold, although higher than for bare steel (described in Fig. 22), can be reached and the corrosion of the galvanized layers develops. This delay in the onset of corrosion with respect to the bare steel is known as the extension of the service life of the reinforcement provided by the galvanizing [30,50–53]. In addition to the delay in the onset of corrosion, zinc corrosion products are not expansive in nature and, therefore, during the time the galvanized coating is dissolving, the life of the structure is also effectively being extended. In comparison, corrosion products that have a more
Figure 32: Icorr of plain and galvanized steel embedded in concrete made with and without 1 NO2 2 addition and immersion in seawater for 5 2 years.
Electrochemical Aspects of Galvanized Reinforcement Corrosion
141
expansive character are those of the base steel and are much more damaging to the cover concrete [54]. This extension is shown also in Fig. 32 for concrete specimens employing various protection methods, in this case galvanizing and the addition of inhibitors considered separately. The specimens had been immersed for up to 5 12 years in sea water. The data shows that the galvanized steel always delays the initiation of the activated state compared with the behavior of bare steel [55].
5.4. Conclusions From all the data and experience contained in the literature, it can be deduced that there are two major factors responsible for the long-term stability of galvanized coatings in contact with concrete: *
*
The alkali content of the cement, which induces different pH values in the pore solution and therefore affects the compactness of the passivating CaHZn layer. The microstructure of the galvanized coating in which the most resistant component is the pure zinc phase that needs to be thicker than 10 mm to provide enough reserve of zinc to form a continuous and compact CaHZn layer. The Fe – Zn alloy phases are very prone to attack, either by high alkali contents or by chloride ions, and therefore their presence should be minimized. The optimum galvanized coating would be of almost pure zinc with a total thickness higher than 50– 60 mm.
Concerning the resistance of reinforcement to the presence of aggressive media, the following points can be made: *
*
Galvanized steel presents negligible corrosion in carbonated concrete. In contact with solutions at about pH 8, the zinc is rather stable and usually remains passive. Although galvanized steel corrodes in presence of chlorides, a higher threshold concentration is needed to initiate corrosion. This behavior does, however, depend on the origin of the chlorides, whether added during mixing or penetrating from the outside, and on the previous state of the galvanized coating.
References [1] Longuet, P. (1976). La protection des armatures dans le be´ton arme´ e´labore´ avec des ciment de laitier. Silicates Industriels, 7/8, 321–328. [2] Moragu¨es, A., Macı´as, A., & Andrade, C. (1987). Equilibria of the chemical composition of the concrete pore solution. Part 1. Comparative study of synthetic and extrated solutions. Cement and Concrete Research, 17, 173–182.
142
C. Andrade and C. Alonso
[3] Roetheli, B. E., Cox, G. L., & Littreal, W. B. (1932). Effect of pH on the corrosion products and corrosion rate of zinc in oxygenated aqueous solutions. Metals and Alloys, 3, 73–76. [4] Pourbaix, M. (1973). Lectures on electrochemical corrosion, Chapter 4: electrochemical equilibria. Plenum Press, New York, pp. 143–145. [5] Bird, C. E. (1964). Influence of minor constituents in Portland cement on the behavior of galvanized steel in concrete. Corrosion Prevention and Control, 11, 7, 17 –21. [6] Grauer, R., & Kaesche, H. (1972). Elektronenmikroskopische und elektrochemisque untersuchunger u¨ber die passivienung von zink in 0.1 M natronlauge. Corrosion Science, 12, 617–624. [7] Shams El Din, A. M., Abd El Wahab, F. M., & Abd El Haleem, S. M. (1973). Effect of electrolyte concentration on the passivation of Zn in akaline solutions. Werkstoff und Korrosion, 24, 389–394. [8] Vorkapic, L. Z., Drazic, D. M., & Despic, A. R. (1974). Corrosion of pure and amalgamated zinc in concentrated alkali hydroxide solutions. Journal of Electrochemical Society, 121, 11, 1385– 1392. [9] Zembura, Z., & Burzynska, L. (1977). The corrosion of zinc in deaerated 0.1 M NaCl in the pH range from 1.6 to 13.3. Corrosion Science, 17, 879–891. [10] Macı´as, A., & Andrade, C. (1987). Corrosion of galvanized steel reinforcements in alkaline solutions. Part I. Electrochemical results. British Corrosion Journal, 22, 113– 118. [11] Duval, R., & Arliguie, G. (1974). Passivation du zinc dans l’hydroxyde de calcium eu e´gard au comportement de l’acier galvanise´ dans le beton. Memoires Scientifiques Revue de Metallurgie, 71, 11, 719–727. [12] Abd El Hallen, S. M. (1979). Initiation and inhibition of pitting corrosion in alkaline solutions under potentiostatic polarization conditions. Werkstoff und Korrosion, 30, 631– 637. [13] Everett, L. H., & Treadaway, K. W. J. (1970). The use of galvanized steel reinforcement in building, Current Papers, CP3/70. Building Research Station, UK, 10 pp. [14] Baught, L. M., & Higginson, A. (1985). Passivation of zinc in concentrated alkaline solution. I. Characteristics of active dissolution prior to passivation. Electrochemica Acta, 30, 9, 1163 –1172. [15] Unz, M. (1978). Performance of galvanized reinforcement in calcium hydroxide solution. ACI Journal, 75, 3, 91 –99. [16] Lieber, W., & Gebauer, J. (1969). Einbau von zink in calcium silicathydrate. Zementkalk-Gips, 4, 161–164. [17] Rehm, G., & La¨mmke, A. (1970). Untersuchungen u¨ber reaktionen des zinks unter einwirkung von Alkalien im Himblick auf das Verhalten Verzinkter Sta¨hle im Beton. Betonsteinzeitung, 6, 360–365. [18] Liebau, F., & Amel-Zadeh, A. (1972). The crystal structure of Ca(Zn2(OH)6) 2H2O — a retarder in the setting of Portland cement. Kistall und Technik, 7, 1–3, 221–227. [19] Macı´as, A., & Andrade, C. (1987). Corrosion of galvanized steel reinforcements in alkaline solutions. Part I. Electrochemical results. British Corrosion Journal, 22, 113– 129. [20] Blanco, M. T., Andrade, C., & Macı´as, A. (1984). SEM study of the corrosion products of galvanized reinforcements immersed in solutions on the pH range 12.6 to 13.6. British Corrosion Journal, 19, 1, 41–48.
Electrochemical Aspects of Galvanized Reinforcement Corrosion
143
[21] Macı´as, A., & Andrade, C. (1986). Stability of the calcium hydroxizincate protective layer developed on galvanized reinforcements after a further increase of the pH value. Materiales de Construccio´n, 36, 204, 19 –27. [22] Macı´as, A., & Andrade, C. (1987). Corrosion of galvanized steel reinforcements in alkaline solutions. Part II. SEM study and identification of the corrosion products. British Corrosion Journal, 22, 119 p. [23] Santos, P. (1986). Influencia de la estructura metalogra´fica y del tipo de cemento en la corrosion de armaduras galvanizadas. Doctoral Thesis. UAM. [24] Andrade, C., Va´zquez, A. J., & Gonza´lez, J. A. (1977). Comportamiento electroquı´mico del acero galvanizado en disolucio´n saturada de hidro´xido ca´lcico — evaluacio´n cuantitativa de su velocidad de corrosio´n. Metalurgia (CENIM), 13, 3, 142–145. [25] Andrade, C., & Macı´as, A. (1987). Galvanized reinforcements in concrete. In: A. D. Wilson, J. W. Nicholson, & H. J. Prosser (Eds), Surface Coatings 2 (pp. 137–179). Elsevier Applied Science, London. [26] Gonza´lez, J. A., Vargas, R., & Andrade, C. (1977). Revisio´n sobre el comportamiento de las armaduras galvanizadas en el hormigo´n. Hormigo´n y Acero, 124, 113–132. [27] Sarmaitis, R. R., & Rozovsky, V. G. (1984). The mechanism of protection of zinc against atmospheric corrosion by chromate coating, Ninth International Conference on Metallic Corrosion (ICMC), Toronto, pp. 390–395. [28] Va´zquez, A. J. (1974). Estudio Sobre la adherencia de las armaduras galvanizadas en la construccio´n de hormigo´n armado. Informes de la Construccio´n, 258. [29] Andrade, C., Molina, A., Huete, F., & Gonza´lez, J. A. (1983). Relation between the alkali content of cement and the corrosion rates of galvanized reinforcements. In: A. P. Crane (Ed.), Corrosion of Reinforcements in Concrete Construction (pp. 343– 355). Society of Chemical Industries, London, Chapter 20. [30] Moreno, E., & Sagu¨e´s, A. (1996). Performance of plain and galvanized reinforcing steel during the initiation stage of corrosion in concrete with pozzolanic additions, Corrosio´n 1996, Paper 326. NACE International, Houston, TX, 326. [31] Andrade, C., & Gonza´lez, J. A. (1984). Comportamiento del acero galvanizado en el hormigo´n armado. Rev. Ibe´r. Corros. y Prot, 15, 4, 35 –41. [32] Andrade, C., & Macı´as, A. (1981). Influencia del contenido en alcalis de los cementos sobre la corrosio´n de armaduras galvanizadas. Materiales de Construccio´n, 184, 83 –93. [33] Mackowiak, J., & Short, N. R. (1979). Metallurgy of galvanized coatings. International Metals Reviews, 1, 237, 1–19. [34] Devillers, L. P., & Niessen, P. (1974). Etude de la galvanisation des aciers au silicium. Rev. Metal, 71, 11, 731–733. [35] Va´zquez, A. J., & Sistiaga, J. M. (1978). Galvanizacio´n de aceros con bajo contenido en silicion; Reactividad y estructura del recubrimiento. Revue de Metallurgie, 14, 6, 346–354. [36] Molina, A., Andrade, C., & Blanco, M. T. (1983). Efecto de la etructura del galvanizado en la corrosio´n de armaduras en contacto con diferentes cementos. Revista Ibe´rica de Corrosio´n y Proteccie´n, 14, 135–140. [37] Sergi, G., Short, N. R., & Page, C. L. (1985). Corrosion of galvanized and galvanealed steel in solutions of pH 9.0 to 14.0. Corrosion, 41 11, 618–624. [38] Sinibaldi, G. (1995). Uso del nitrito di calcio come ihibitore di corrosione per armature zincate in calcestruzzo. Minor Doctoral Thesis. University of La Sapienza.
144
C. Andrade and C. Alonso
[39] Gonza´lez, J. A., & Andrade, C. (1982). Effect of carbonation, chlorides and relative ambient humidity on the corrosion of galvanized rebars embedded in concrete. British Corrosion Journal, 17, 1, 21–28. [40] Gonza´lez, J. A., Va´zquez, A. J., & Andrade, C. (1982). Les effets des cycles d’humidite sur la corrosion des armatures galvanisees dans les mortiers carbonates et non Carbonates. Materiaux et Constructions, 15, 88, 271–278. [41] Arliguie, G., & Grandet, J. (1987). Influence de la corrosion atmospherique des armatures d’acier galvanise´ sur leur comportement dans le beton, RILEM, First International Congress on Durability of Construction Materials, pp. 998–1004. [42] Santos, P., Macı´as, A., Alonso, C., & Andrade, C. (1986). Behaviour of galvanized bars in solutions simulating carbonated concrete pore liquid: study by DC and AC techniques, 27th Corrosion Society Symposium, Birmingham, UK. [43] Guo, R., Weinberg, F., & Tromans, D. (1995). Pitting corrosion of passivated zinc monocrystals. Corrosion Science, 356–366. [44] Isikawa, T., Cornet, I., & Bresler, B. (1969). Electrochemical study of the corrosion behaviour of galvanized steel in concrete, Fourth International Congress on Metallic Corrosion (ICMC), Netherlands, pp. 136–138. [45] Alonso, C., Andrade, C., Sinibaldi, G., & Proverbio, E. (1996). Corrosion of galvanized steel in presence of chlorides, Eurocorr 96, Niza, 96. [46] Macı´as, A., & Andrade, C. (1990). The behaviour of galvanized steel in chloride containing alkaline solutions: I — the I\influence of cation. Corrosion Science, 30, 4/5 393– 407. [47] Ramirez, E., Gonza´lez, J. A., & Bautista, A. (1996). The protective efficiency of galvanizing against corrosion of steel in mortar and in Ca(OH)2 saturated solutions containing chlorides. Cement and Concrete Research, 26, 10, 1525–1536. [48] Page, C. L., Sergi, G., & Short, N. R. (1989). Corrosion behaviour of zinc coated steel in silica fume concrete. In: V. Malhotra (Ed.), Third International Conference on Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Vol. 2, pp. 887–897. [49] Alonso, C., Sinibaldi, G., Andrade, C., & Cigna, R. (1997). Protection of reinforcements from chloride attack by the use of galvanized steel and calcium nitrite. In: V. Malhotra (Ed.), Fifth International Conference on Superplastizers and Other Chemical Admixtures in Concrete, pp. 123–142. [50] Hildebrand, H., & Schwenk, W. (1986). Effect of galvanizing on the corrosion of steel in concrete immersed in NaCl solution. Werkstoffe und korrosion, 37, 163–169. [51] Tonini, D. E., & Cook, A. R. (1978). The performance of galvanized reinforcement in high chloride Environments: field study results, Corrosion 1978, Paper 5. NACE International, Houston, TX. [52] Swamy, R. N. (1990). Resistance to chlorides of galvanized rebars. In: C. L. Page, K. W. J. Treadaway, & P. B. Bamforth (Eds), Corrosion of reinforcement in concrete (pp. 586–600). Society of Chemical Industries. [53] Yeomans, S. R. (1994). Performance of black, galvanized and epoxy-coated reinforcing steels in chloride contaminated concrete. Corrosion, 50, 1 72 –81. [54] Yeomans, S. R. (1998). Corrosion of the zinc alloy coating in galvanized reinforced concrete, Corrosion 1998, Paper 98653. NACE International, Houston, TX, 10 pp. [55] Andrade, C., Alonso, C., Gon˜i, S., & Gonza´lez, J. A. (1988). Corrosio´n de armaduras en hormigones en contacto con agua de mar: El efecto de la galvanizacio´n y el uso de nitritos como inhibidor, Seventh Congreso Internacional de Corrosio´n Marina e Incrustaciones, November 1988.
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 6
Laboratory and Field Performance of Galvanized Steel in Concrete Stephen R. Yeomans University of New South Wales, Canberra, Australia
6.1. Introduction Over the last 30– 40 years, a significant body of research has been undertaken to investigate the characteristics and behaviour of galvanized steel for use in reinforced concrete construction. This has included studies of galvanized components immersed in simulated concrete pore-water and other aqueous solutions, as well as exposure studies on galvanized bars embedded in cement pastes, mortars and concrete mixes. A number of large-scale field studies have also been undertaken with surveys of in situ galvanized reinforcement in existing structures, mainly bridge decks and marine structures. Overall, many hundreds of such investigations have been completed and the record of published reports, conference and seminar papers and scientific articles relating to this research is quite extensive. Much of this research has been curiosity-driven, coming at a time when a great deal of worldwide interest and research was being undertaken on a range of topics to do with concrete materials and concrete durability generally. The growing use of galvanized reinforcement at this time was just one area of interest in this body of work. In addition, however, a significant investment was made in contract research funded by various government agencies such as the Federal Highways Administration in the USA (FHWA) and other private sector groups such as the International Lead Zinc Research Organization (ILZRO) and Zinc Development Associations and Galvanizing Associations in several countries including the USA and UK, Canada, Japan, India and Australia. Over the years, a number of authoritative reviews of the literature relating to galvanized reinforcement in concrete have been published [1 –5]. These milestone papers are widely referred to in other articles; they provide excellent snap-shots of the history of the development of the science and technology of galvanized reinforcement and a chronology of the debate surrounding the use of the product,
146
S.R. Yeomans
which at times has been quite controversial. In addition, a number of technical brochures and guides to specifying galvanized reinforcement [6–9] and review articles [10– 12] have been published, as well as a number of national and international Standards (refer Chapter 4). The controversy, if it can be called such, developed in the mid-1970s in the USA. The FHWA was faced with the repair or replacement of thousands of concrete bridge decks damaged by the chloride-induced corrosion of reinforcement caused by the widespread use of de-icing salts. Galvanized reinforcement had been in use in bridges for many years by this time, and its performance had been the subject of a number of laboratory and field investigations. While a majority of the reports published had findings favourable to galvanizing, others were either unfavourable or showed no long-term benefit. A summary of this research compiled by the FHWA (in FHWA Notice 5140.10, July 9, 1976) acknowledged that much of the work reported was confused, that there was little consistency in the types of concrete and exposure conditions used, and that the results of the field investigations were equally inconclusive. By issuance of this Notice, however, galvanized reinforcement for use in bridge decks was classified as an “experimental system (limited to three decks per State) until definite favourable conclusions could be drawn from existing field and laboratory experimental installations”. This ruling prompted a surge in research on galvanized reinforcement, particularly in high chloride-exposure situations, and this is reflected in the extent of published work after 1976. In 1983, the FHWA announced (Notice N5140.21) that “considerable research has been done on the effectiveness of the various protective systems and sufficient information is now available to allow States more flexibility in the selection of a cost-effective system”. The following year, the FHWA amended its policy relating to concrete bridge decks to allow States the “flexibility to select a protective system and reconstruction method based on local conditions and experience (Federal Register/Vol. 49/No. 6/January 1984). To this time, protective systems acceptable to the FHWA had included epoxy-coated reinforcement, dense concrete, latex-modified concrete, membranes and cathodic protection. The January 1984 ruling deleted this listing in its entirety, thereby allowing States to choose the most cost-effective system, including galvanizing, as dictated by local conditions. In the sections to follow, a review is presented of the diverse range of laboratory and field-based research that has focused on galvanized reinforcement. As previously noted, there are many hundreds of such references and not all can, or need, be included here. For those that have been referenced, a level of information is provided to give a sense of the work as well as a summary of the pertinent results and conclusions. Reference should be made to the primary sources for full details of the work undertaken.
Laboratory and Field Performance of Galvanized Steel in Concrete
147
Finally, in compiling this book, it is inevitable that there will be some overlap between chapters; for example, some work included in the overview in Chapter 1 is also referenced in other chapters in this book. A detailed review of the electrochemical behaviour of galvanized steel in aqueous solutions and simulated pore-water conditions is presented in Chapter 5. Similarly, a review of experience with galvanized reinforced concrete in the Islands of Bermuda, perhaps the largest and best-known single case study, is given in Chapter 7, and issues relating to bond and slip of galvanized bar in concrete are dealt with in Chapter 8. This body of work is not repeated here. Rather, this chapter will focus on key laboratory-based research on galvanized reinforcement embedded in mortar and concrete specimens and field investigations of longer-term performance of galvanized reinforcement. To give a sense of the era, this chapter is separated into a number of time periods.
6.2. Some Early Research An early review of the corrosion of metals in buildings was published by Halstead [13]. This survey was in part a digest of reports on the corrosion of a number of metals in concrete, including zinc, published over the preceding 40 years, as well as a summary of proposed methods for preventing corrosion of steel in contact with concrete. In introducing this broad topic, Halstead noted that at this time very little information was available on the corrosion of metals in concrete with binders other than normal Portland cement, but that it seemed likely that all Portland cements, slag cement and high alumina cement would behave similarly. The following extract is that presented for zinc: “Zinc is probably infrequently used in contact with concrete. It is not so severely attacked as aluminium, and zinc coating has frequently been suggested as a means of protecting steel from corrosion in concrete. Galvanizing, sheradising and electro-plated processes have been used. Galvanized metals are reported to be attacked, but not usually too badly (Cement Bulletin, Switzerland, 6, 1948), protection being regarded as advisable (Obst W, Zement, 15, 1926, 582) rather than essential. A galvanized steel reinforcing bar which had been embedded 54 years in concrete exposed out of doors at the seaside was examined recently. Most of the zinc film was intact and penetration and rusting of the steel had only occurred in a few isolated spots. Zinc-coated steel channel used to bond cementwood wool slabs, on the other hand, has been known to be corroded badly when stored in damp conditions, although satisfactory if dry. A coating of zinc 0.02 mm (20 mm) thick on steel sheets is reported to have been sufficient to resist corrosion from asbestos-cement sheets, but thicker coatings were required when the steel sheets were in contact with cement-sand mortars (Soloy’ev AV, Zh Prikhlm, 22,
148
S.R. Yeomans
1949, 62). Nevertheless, complete destruction of zinc conduits in contact with wet concrete is reported (Schikorr G, Wiss Abh dtsch MatPrufAnst, 2 1941, 51). Galvanized steel reinforcement does not cause spalling of concrete if the zinc is attacked, but solid zinc plates or rods may do so on account of the greater bulk of the corrosion products (Evans RH, Proc Inst Civil Engs (London) 4, 1955, Part III, No 3, 725). For general protection, a coating of bitumen is probably desirable for zinc sheets or other metals thinly coated with zinc.” A number of key investigations of the corrosion of galvanized steel in concrete were reported by the group comprising Bressler, Cornet and Ishikawa (from the University of California, Berkeley) through the mid-1960s to the early 1970s [14–17]. A review of this and related work has been separately published [2,18]. These studies were performed with either black or galvanized bars (coating thickness of about 160 mm) embedded in prismatic concrete specimens (0.5 w/c and 40 MPa). The specimens were variously exposed air, wetting in a 4% salt solution followed by drying, and continuous immersion in salt solution with an impressed DC anodic current. After initial curing, the bars in each of the exposure environments were loaded axially to about 140 MPa and this load was maintained during the exposure. Specimens stored in air showed no cracks or rust stains after 24 months. However, under the influence of the impressed anodic current, specimens with black steel showed cracking after 1 month while those with galvanized steel cracked after 2 months. The black steel specimens had relatively wide cracking while the galvanized specimens showed fine narrow cracks. The data indicated that the average length and width of cracks for galvanized specimens was less than that for the black steel specimens. For specimens subjected to alternate immersion, four with black steel cracked after 9 months, a fifth after 10 months, and a sixth after 18 months. None of the six galvanized specimens had cracked after 15 months, although two showed hair-line cracks after 16 months and one more had cracked after 20 months. The remaining three galvanized specimens were uncracked after 23 months. Slip measurements also showed that the black steel specimens showed greater slip than those with galvanized reinforcement. The main conclusions from this work were that, under highly corrosive conditions, the concrete cracks much earlier and the bond between the concrete and the reinforcement is decreased more rapidly with black steel than with galvanized steel reinforcement. In another early paper, Frazier [19] discussed the use of galvanized steel as a reinforcing medium for concrete structures and also reported on case histories comparing galvanized bar with plain steel. Data from Bermuda going back to 1935 noted that, where steel rods had corroded sufficiently to crack the concrete, galvanized rods, although pitted, were not so severely attacked as to cause distress to the concrete except where it was very thin. This paper also reported on tests
Laboratory and Field Performance of Galvanized Steel in Concrete
149
undertaken on galvanized inserts for concrete, which traditionally were only partially embedded, over concerns around inadequate bond strength. This work led to recommendations that all such inserts should be fully galvanized, for efficiency in handling and fabrication but also because it had been demonstrated that galvanized rods had superior bond strength to carbon steel when cast in concrete. Griffin investigated the suitability of galvanized steel reinforcement in concrete exposed to a marine environment [20]. Eleven small concrete walls approximately 1 m2 were constructed with different reinforcement materials including black steel and galvanized steel. Both air-entrained (20 MPa, 0.644 w/c) and non-airentrained (17 MPa 0.702 w/c) concretes were used. Cover to the face of the walls was 25 mm. The walls were sprayed daily with seawater for a period approaching 3 years. The incidence of cracking of the walls was monitored and the concrete was periodically removed from the reinforcing grid to allow examination of the steel. The conclusions drawn from this large-scale investigation were that galvanized steel reinforcement was, at best, no better than non-galvanized steel reinforcement and air-entrained concrete inhibited corrosion of either galvanized or nongalvanized steel compared with non-air-entrained concrete. It was further observed that painting damaged areas of galvanized steel with zinc-enriched paint was not effective in inhibiting the corrosion of the reinforcement. It should be noted, however, that, for this severe exposure test, the quality of the concrete used was quite low, with high w/c and thus high permeability, and the cover to the bars, at 25 mm, was also very low, especially for marine exposure. Under these circumstances, it is perhaps not surprising that the galvanizing did not survive for longer periods compared with the uncoated bars.
6.3. The 1970s Christensen and Williamson [21] studied the galvanic cell that can develop in ferro-cement construction due to contact between galvanized steel mesh and plain reinforcing steel. The use of galvanized wire of mesh in ferro-cement construction such as boats was favoured for a number of reasons, mainly its resistance to corrosion prior to the mortar being placed, its increased resistance to chlorides and also because, once corrosion had initiated, the zinc would preferentially corrode and cathodically protect the steel. Despite these advantages, the presence of galvanic cells led to the evolution of hydrogen gas at the steel bar resulting in poor bonding, lower strength and reduced durability. Several solutions were presented, but the addition of chromium trioxide to the mix water in order to passivate bars and stifle the hydrogen reaction was shown to solve the problem most effectively.
150
S.R. Yeomans
Okamura and Hisamatsu [22] carried out research on the relationship between corrosion and crack width in reinforced concrete beams. Rectangular beams (150 £ 200 £ 1600 mm) containing galvanized bars (110 mm coating thickness) or black bars were exposed to a 3% salt solution sprayed onto the specimens twice daily for 1 year. The 0.55 w/c concrete had a 28-day strength of 40 MPa. The beams were pre-cracked by bending in a rigid frame and the load was maintained during exposure. After the period of exposure, the beams were subjected to a fatigue test then crushed to examine the extent of corrosion. The results showed that the longer the exposure, the lower the fatigue strength due to corrosion of the bars near the crack opening. For black steel bars, the 2 £ 106 cycle fatigue stress before exposure was 267 MPa, reducing to about 200 MPa at 6 months and 167 MPa at 1 year. In the case of galvanized steel, the fatigue stress reduced from 250 to 200 MPa at 8 months and 195 MPa at 1 year. Since the reduction in fatigue stress was less for galvanized bars than for black steel, it was concluded that galvanizing reduced the effect of corrosion of the reinforcement in cracked concrete. This was confirmed from visual examination of the bars taken from the beams after the fatigue test. Red rust was found on the black bars near the tip of each crack in the concrete while little such rusting was observed on the galvanized bars. Thus, by using galvanized bars, the durability of concrete members with cracks of about 0.3 mm width had the same durability as black steel reinforced concrete with cracks about 0.2 mm wide. It was also noted that reaction between the galvanizing and the concrete caused the concrete to adhere very tightly to the bars. Over the period 1974 –1978, the International Lead Zinc Research Organization sponsored two major projects to follow the case histories of the use of galvanized reinforcement in concrete structures exposed to, or containing, chlorides. These two projects, undertaken by Stark and Perenchio [23] and Stark [24] of the Construction Technology Laboratories of the Portland Cement Association, used a variety of evaluation methods including potential measurements, delaminations surveys, chloride measurements, petrographic examination and metallurgical assessment of the galvanized coating. A summary of the results of these two projects has been published by ILZRO [2] and an overview of the findings is in Table 1. It is apparent that in these structures, despite chloride levels in many cases up to 8 – 10 times higher than threshold values for corrosion of black steel, the galvanized bars had not been affected in conditions where bare steel would have been expected to display severe corrosion. The thickness of the galvanized coating remaining on the bars was also well above minimum specified levels, indicating that only minor reaction between the concrete and the coating had occurred. Specific data from the report by Stark [24] relating chloride content to the remaining thickness of the galvanized coating is given
Laboratory and Field Performance of Galvanized Steel in Concrete
151
Table 1: Summary of findings of surveys of galvanized reinforced structures from ILZRO Projects ZE-206 and ZE-247 [23,24]. Description of structure
Age when inspected
Longbird Bridge Bermuda
21– 23 years Chloride levels at bar level of 1.02– 4.38 kg/m3 pH at bars of 12.7 Average coating thickness 236 mm in range 165 – 442 mm No evidence of distress or reactions between the paste and the galvanized steel 7 years Chloride levels at bar level of 0.3– 0.7 kg/m3 Average pH at bars of 11.8 Evidence of porosity due to gas evolution on some bars No evidence of distress to the zinc coating, which was virtually unaffected No activity in region of mixed galvanized/black bar 3 years Chloride levels at bar level of 1.17 kg/m3 pH at bars in range 12.2 –12.5 Average coating thickness remaining of 130 mm Some frothing due to gas evolution on some bars No evidence of abnormal electrical activity in panels containing both galvanized/black bar 3 years Chloride levels at bar level of 0.85 kg/m3 pH at bars in range 12.4 –12.5 Average coating thickness remaining of 196 mm Some frothing due to gas evolution on some bars No evidence of corrosion, secondary reaction products or distress associated with the galvanized coatings 8 years Chloride levels at bar level of 0.54 kg/m3 pH at bars of 12.7 Minor superficial corrosion on surface of galvanized bar No evidence of abnormal electrical activity in bridge deck 10 years Chloride levels at bar level in range 1.9 –6.0 kg/m3 Average coating thickness remaining of 185 mm No evidence of corrosion on the galvanized bars
Long Dick Creek I35 Ames, IA
Boca Chic Bridge US1 Key West, FL
Seven Mile Bridge US1 Key West, FL
Flatts Bridge Bermuda
Hamilton Dock Bermuda
Findings
(continued)
152
S.R. Yeomans
Table 1: Continued. Description of structure
Age when inspected
Findings
RBYC Jetty Bermuda
8 years
Chloride levels at bar level in range 3.36 – 3.66 kg/m3 Average coating thickness remaining of 94 mm No evidence of corrosion on the galvanized bars Chloride levels at bar level in range 5.5 –12.7 kg/m3 for 1964 construction, 2.4 –4.6 kg/m3 for 1966 construction and 2.4 –2.6 kg/m3 for 1969 construction Average coating thickness remaining in excess of 150 mm No evidence of distress related to corrosion of the galvanized bars despite high chloride levels Chloride levels quite low at 0.30 –0.36 kg/m3 pH at bars of 11.8 No evidence of corrosion on the galvanized bars
Penno’s Wharf 7 – 12 years St George Bermuda
Manicougan River Bridge on I38 Hauterive, Que.
8 years
in Table 2, indicating that between 92 and 100% of the original coating remains after these periods of exposure to chlorides. Further detail of this work is also given in Chapter 7. Gonzalez and Andrade [25] reported on tests in progress on black and galvanized bars embedded in concrete. A number of variables were investigated
Table 2: Summary of chloride and coating thickness data from Bermuda [24]. Structure Longbird Bridge St George Dock
Hamilton Dock Bermuda Yacht Club
Age (years)
Chloride % by weight of concrete
23 12 10 7 10 10 8
0.19 0.27 0.22 0.14 0.08 0.14 0.16
% of zinc coating remaining 98 92 99 98 þ 95 96 100
Laboratory and Field Performance of Galvanized Steel in Concrete
153
including type of cement, carbonation, the use of additives and inhibitors and the humidity of the environment. In short-term tests, the corrosion rates for both black and galvanized steel were similar, although in longer-term tests where the humidity was cycled, a protective layer developed on the galvanized bars, which reduced the corrosion rate.
6.4. The 1980s Long-term studies of the corrosion susceptibility of galvanized steel in concrete were reported by Treadaway et al. [26]. Two major test programmes were described. One programme, which had been underway for 14 years, compared the performance of zinc-coated and plain steel in a variety of concretes made with both dense and lightweight fly ash aggregates. The second, of 5 years duration and where corrosion had been accelerated by the addition of calcium chloride to the concrete, evaluated the performance of galvanized steel and a number of grades of stainless steels. Both programmes employed a large number of concrete prisms with variable cover to the reinforcement and reinforced beams that had been stressed to crack the cover concrete. Concretes with variable w/c (0.5–0.9) and a range of strengths (20 – 40 MPa) were used. The galvanized bar had a mean coating thickness of 89 mm and some were chromate treated. Specimens in both programmes were exposed to an industrial atmosphere and at a coastal site. The principal conclusions from this extensive programme of work can be summarised as follows: *
*
* *
*
at high chloride contents (1.9% and above by weight of cement), although serious corrosion of both galvanized and plain steel occurred, galvanizing was able to delay the onset of corrosion by 6 months to 1 year; in concretes containing up to 0.96% chloride, galvanizing provided a more tangible benefit but it was not possible to determine the full extent of this; in carbonated concrete, galvanizing measurably delayed the onset of cracking; in good-quality chloride-free concrete, galvanizing performed well, as did other steels; and under certain circumstances, galvanizing can considerably lengthen the corrosion phase of the three-stage corrosion process (initiation, corrosion and cracking) and thus can provide a useful benefit.
Because of concerns over conflicting data coming from many laboratory and field studies to this time, which clearly might have arisen due to differences in experimental techniques and types of concretes and exposure conditions used, the FHWA undertook a long-term programme to monitor the corrosion
154
S.R. Yeomans
behaviour of concrete containing either conventional black steel reinforcement or hot-dip galvanized reinforcement [27]. Slabs were cast with top and bottom mats of reinforcement in two different concretes (0.4 and 0.5 w/c). The slabs were ponded with salt solution to encourage penetration of chlorides to the top mat. This allowed indirect estimation of the corrosion rate by the measurement of the macro-cell current between the two reinforcing mats. Either black steel or galvanized steel was used in both mats or a mix of black and galvanized bars was used. The results indicated that, in 0.4 w/c concrete, there was no extra benefit in the use of galvanized bars over black bars. In this concrete, with 0.19% chloride (by weight of concrete) at the level of the bars, the corrosion current for the galvanized bar was 503 and 425 mA/m2 for the black bar. In the 0.5 w/c slabs when both mats were galvanized, the average corrosion current was 1025 mA/m2 at a chloride concentration of 0.3%. In slabs with a bottom mat of black steel and the top mat galvanized, the average corrosion current was 3314 mA/m2, more than three times higher than when both mats were galvanized. In this case, the chloride concentration was 0.27%. In slabs with both mats of black steel, the corrosion current was 4291 mA/m2 at a chloride concentration of 0.4%, some four times higher than for the dual galvanized configuration. In lesser quality concrete (0.5 w/c) with galvanized bar in the top mat and black steel in the bottom mat, corrosive attack on the galvanizing was quite rapid. In this case, the rate of corrosion of the galvanized bar was twice as high as when black bars only were used. When both mats were galvanized, however, the data showed a significant benefit with low corrosion currents and no cracking of the slabs. In contrast, black steel in similar conditions was severely attacked and corrosion-induced cracking of the slabs occurred. A summary of the corrosion potential data measured during this series of experiments is given in Table 3. This shows that, for the galvanized reinforcement, a potential of about 2380 mV (vs CSE) is indicative of the passive condition and about 2 600 mV of the active condition. The main conclusions drawn from this important series of experiments were: *
*
*
better performance of galvanized reinforcement was observed in 0.5 w/c concrete when all the reinforcement (i.e. both top and bottom mats) was galvanized; when the galvanized bar was coupled with black bar, the rate of corrosion was more than twice that compared with black bars; and there was a 36% reduction in macrocell corrosion current when galvanized reinforcement was used. This benefit was, however, far less than would be obtained by using black steel alone in a 0.40 w/c ratio concrete where an 86% reduction in macrocell corrosion current was observed.
Laboratory and Field Performance of Galvanized Steel in Concrete
155
Table 3: Chloride concentration versus Ecorr [27]. Concrete condition
Chloride-contaminated Chloride-free a
w/c ratio
0.5 0.4 0.5 0.4
% Chloridea
0.3 0.4 – –
Ecorr mV (vs CSE) Galvanized bar
Black bar
2 621 2 687 2 385 2 390
2 574 2 400 2 135 2 194
Chloride concentration given as percent by weight of concrete.
It was also apparent at this time that field studies galvanized reinforced concrete had also yielded conflicting results. For example, accelerated field studies in Michigan over a period of 6 years had shown that galvanized bars would retard concrete delamination and spalling but not prevent them, especially in concrete where the cover to the reinforcement was shallow [28]. However, later studies coming out of Bermuda, dealt with in detail elsewhere and reported by ILZRO [2], were indicating that salt-contaminated bridges that had been in service for 20 years or more were not showing signs of corrosion damage, despite the fact that chloride concentrations at the level of the galvanized steel, at up to 4.4 kg/m3, were well above the critical limit for the corrosion of steel [29]. To address this growing dilemma, the FHWA convened an expert panel to examine critically the available literature on the performance of galvanized reinforcement in concrete [30]. In brief, the panel reported that: *
*
*
*
Laboratory studies with aqueous solutions had shown that zinc tolerates higher chloride content than steel before corrosion begins, which at least partly explains why galvanizing delays the onset of corrosion in the outdoor specimens. However, there is indication from the outdoor exposure studies that, once corrosion of galvanized steel bars begins, the rate at which corrosion-induced distress on the concrete occurs is more rapid than for black steel. Owing to a lack of standardization in the methodology (such as specimen design, quality of concrete, exposure conditions, and criteria for distress) used in the various laboratory studies, it is difficult to extrapolate the laboratory results to field conditions for predicting service life. It was deemed impossible to predict their service life since most of the structures were not in service sufficiently long (at the time of the study) for chloride to accumulate in the concrete to a level at which corrosion of the reinforcement would be expected.
156
*
*
*
S.R. Yeomans
In the case of some marine structures (where chloride levels are high), there is the possibility that high moisture contents cause insufficient oxygen to be available at the reinforcement to sustain significant corrosion activity. There is evidence that rapid corrosion will occur when galvanized and black steel (used in the same structures and are electronically) are interconnected in a salt-containing environment. When galvanizing is preceded by chromate treatment, the chemical reaction of zinc in the fresh concrete does not impair bonding of the steel to the concrete or the mechanical properties of the reinforcing steel. Therefore, the same design criteria used for black steel bars can be used for galvanized steel bars.
The FHWA panel was of the view that, for a new concrete bridge deck with a 51 mm cover of 0.45 w/c concrete, and assuming that normal construction practices are used, the use of galvanized steel bars may add 5 more years to the 10– 15 years that is typically required for corrosion-induced distress to be manifested in unprotected bridge decks. In an ILZRO sponsored project, Hoke et al. examine the stresses induced in cement mortars due to corrosion of both galvanized and black steel bars [31]. Cylindrical mortar specimens were cast with a centrally placed 16 mm bar of either galvanized steel or black steel. The bar was enclosed by a thin-walled titanium tube to which strain gauges had been attached to record the expansion forces created as corrosion occurred. A 1:3 mix was used and NaCl was added to the mix. The specimens were immersed in a salt solution of concentration up to 5%. Initially, very little corrosion occurred and to overcome this the specimens were anodically polarized to 1.5 or 6 V for 30 days. In the 6 V tests, a strain of around 900 microstrain was measured around galvanized bar, whereas for black bar it was around 3600 microstrain. The internal pressure developed due to corrosion of the galvanized bar was about 37 MPa, whereas for the black bar it was about four-times higher at 158 MPa. Examination of the corroded bars revealed that the bare steel bars were covered with a black deposit, identified by XRD as Fe3O4. For the galvanized bars, a dark grey layer was covered with a powdery, less-adherent corrosion product. The dark inner-layer was identified as Zn5(OH)8Cl2 and the lighter outer-layer was ZnO. Mineralogical data indicated that the volumetric change accompanying the formation of Fe3O4 and ZnO was 7.80 and 5.36 cm3/mol of parent metal consumed, respectively. The Fe3O4 layer was found to be continuous and adherent. It expanded radially and was effectively locked in place at the bar/mortar interface, thus causing large expansion forces. In comparison, the ZnO layer was powdery and much less adherent. It also had a degree of mobility and was able to diffuse into matrix, thus reducing the expansion at the interface. From these studies, it was clear that, at equal corrosion rates, the corrosion of bare steel bars in
Laboratory and Field Performance of Galvanized Steel in Concrete
157
cement mortars resulted in the generation of significantly larger stresses in the matrix than was the case for galvanized bars. The effect of humidity cycles, in particular, on the corrosion of galvanized steel embedded in carbonated and non-carbonated concrete was studied by Gonzalez et al. [32]. This work, in part prompted by a series of mostly European reports that cast doubt on the usefulness of galvanizing, studied the corrosion rate of bare and galvanized steel in cement mortars that were cyclically subjected to different ambient humidities during hardening and curing. From this work, the corrosion kinetics of the galvanized steel was shown to depend on three principal factors, which, as the authors noted, were largely disregarded in previous research; namely, the type of galvanized coating, the type of cement, and the humidity during curing and hardening. Gonzalez and Andrade also investigated the effect of chlorides, carbonation and relative ambient humidity on the corrosion rate of galvanized bars embedded in cement mortar [33]. Mortars with 0.5 w/c were cast, to which calcium chloride at a dosage of 2% by weight of cement was added. Galvanized bars with an average coating thickness of 60– 80 mm were used. One set of specimens was carbonated for 7 days in carbon dioxide at 50% RH. The specimens were maintained in three different ambient conditions of 50 and 100% RH and partial immersion in distilled water. In all exposures, the galvanized reinforcement initially showed a high corrosion current, which then reduced to a very low value; a maximum current of 1.5 mA/cm2 was observed at 100% RH in non-carbonated chloride-contaminated mortar, but this was one-sixth of the current for uncoated steel in the same conditions. The presence of chlorides had a greater effect on the corrosion of galvanized bar than carbonation. In non-carbonated mortar, maximum currents of 0.1, 1.5 and 0.15 mA/cm2 were observed at 50 and 100% RH and partial immersion conditions, respectively. The influence of ohmic resistance of the mortar on the corrosion behaviour of galvanized reinforcement was also studied. This indicated that, if the resistance was of the order of 104 – 105 V, the lowest corrosion currents were observed, while higher currents were observed for resistance in the range 102 – 103 V. These studies demonstrated that, under the same conditions in chloridecontaminated mortar, uncoated steel corroded 5–10 times faster than galvanized bars. It was also determined that the galvanized coating become more resistant in chloride-carbonated mortar and its corrosion products would not be as damaging to the concrete mass (compared with the corrosion of bare steel) due to their less expansive character. Further, since the corrosion rate is governed by the resistance of the mortar, it was also noted that, when carrying out corrosion tests on reinforcement (and galvanized reinforcement), this should be undertaken in hardened mortar or concrete.
158
S.R. Yeomans
In May 1982, Stark reported on a further ILZRO-sponsored project to evaluate the performance of galvanized reinforcement in concrete bridge decks [34]. The investigation covered eight bridges, built between 1967 and 1975 and subjected to de-icing salt applications, in Iowa, Pennsylvania and Vermont. The work was again undertaken by the Construction Technology Laboratories of the Portland Cement Association. Several of the bridges had been included in previous surveys reported here. A detailed survey of each bridge was undertaken, a brief summary of which is given in Table 4. It was noted by Stark that, although corrosion of galvanized bars was observed, a threshold for the chloride-induced corrosion of the galvanized steel could not be determined. Despite some difficulties in the programme of work, the main conclusions from this study were that: *
*
*
galvanized reinforcement in all but a localized area of one bridge was showing satisfactory resistance to corrosion after these exposure periods; mild or superficial corrosion of the galvanized coating may occur without resultant distress in the concrete; and localized corrosion of the galvanized coating and the base steel may lead to distress and subsurface delaminations where insufficient concrete cover and poor quality concrete are present.
Andrade et al. investigated the influence of the total alkali content of different cements on the corrosion rate of galvanized reinforcement [35]. Mortar specimens (0.5 w/c) were cast with several different cement types. Hot-dip galvanized bar with an average coating thickness 60 –80 mm was used. One series of specimens was kept at 100% RH and another partially immersed in distilled water for 1 year. Corrosion potentials and currents were measured through the experiment and, on completion, the weight loss and morphology of the zinc coating were assessed. In the normal Portland cement (pH 12.60) and the low C3A cement (pH 12.72), the galvanized bars had very low corrosion currents, varying from 0.6 and 0.1 mA/cm2, respectively, at commencement to 0.02 mA/cm2 after 1 year. The coating weight loss was less than 1 mg/cm2, equivalent to 0.3 mm thickness loss per year. The highest corrosion current of 0.27 mA/cm2 was from galvanized bars embedded in normal Portland cement of pH 12.85 and with high initial strength. If homogeneous attack had continued at this rate, the life of the coating would have been about 11 years. Similar behaviour was observed in specimens partially immersed in water and the galvanized coating performed well in Portland cement and low C3A cement. Galvanized bars embedded in low pH slag and fly ash cements (around pH 12.1) showed higher corrosion currents around 0.8 mA/cm2. The metallographic studies revealed that the outermost pure zinc layer of the galvanized coating (the eta layer) was attacked while the underlying zeta
Laboratory and Field Performance of Galvanized Steel in Concrete
159
Table 4: Summary of evaluations of galvanized reinforced concrete bridge decks from ILZRO Project ZE-320 [34]. Description of structure
Age when inspected
Comments and findings
Ames Bridge, Long Dick Creek, IA
14 years
Athens Bridge, PA
8 years
Betsy Ross Bridge, Philadelphia, PA
8 years
Coraopolis Bridge, Alleghany County, PA
9 years
Hershey Bridge, Dauphin County, PA
6 years
Montpelier Bridge, Montpelier, VT
10 years
Only bridges to provide direct comparison between untreated and galvanized reinforcement in the same structure In several areas, water soluble chloride levels were at or above accepted threshold levels Galvanized bars showed mild corrosion of the free zinc layer but with 137– 178 mm zinc remaining on the surface All reinforcing steel galvanized Concrete quite poor (high w/c) and cover , 50 mm Minor transverse cracking and subsurface delaminations Chloride levels in range 1.8 – 8.0 lb/yd3 Some local attack on top side of galvanized bars, but overall virtually no corrosion of the galvanized coating All reinforcing steel galvanized Water soluble chloride levels above threshold values in some locations Mild corrosion of free zinc layer along top surface of galvanized bars, but no evidence of significant corrosion No distress to the concrete All reinforcing steel galvanized Chloride levels below threshold values No evidence of significant corrosion and thus all galvanized bars relatively unaffected All reinforcing steel galvanized Some cracking but no delaminations Corrosive conditions in some areas but galvanized samples retained 120 – 160 mm coating thickness Top mat reinforcement 12 mm below main deck top surface galvanized, with 62 mm concrete overlay (continued)
160
S.R. Yeomans
Table 4: Continued. Description of structure
Age when inspected
Orangeville Bridge, Columbia County, PA
7 years
Tioga Bridge, Tioga County, PA
7 years
Comments and findings Extensive delamination, but not associated with reinforcement corrosion Virtually no corrosion of galvanized coating, even in concrete with chloride levels well above steel thresholds Average remaining coating thickness in range 135 –196 mm All reinforcing steel galvanized No evidenced of surface defects or delaminations No evidence of corrosion of reinforcement Slight corrosion of free zinc layer in coating but thickness remaining in range 188 –259 mm All reinforcing steel galvanized Numerous transverse cracks across full width of deck There was little evidence of overall general corrosion though some areas were locally attacked Coating thickness remaining in range 150 –180 mm
layer remained unaffected. From this, it was concluded that a pH threshold exists (about 12.7), above which there was a risk of dissolution of the galvanizing during the life of the structure. It was noted that further long-term experimentation would be needed to determine whether attack on the coating would continue through the zeta layer as well. Satake et al. [36] reported a detailed study of the performance of epoxy-coated bars with results compared against plain bars and galvanized bars. Concrete prisms (0.55 w/c) were cast in 25 MPa concrete with covers of 20, 40 and 70 mm to the bars. All specimens were pre-loaded to produce cracks of width in the range 0.11– 0.25 mm depending on the cover. The laboratory tests involved immersion in seawater at 60 8C for 6 h followed by drying in air for 6 h. Samples were also exposed to a natural tidal zone. A static tensile stress of 200 MPa was maintained on the specimens throughout testing by means of a reaction frame.
Laboratory and Field Performance of Galvanized Steel in Concrete
161
After 2 years of accelerated exposure, the plain bars with 20 and 40 mm cover had rust covering almost the entire bar surface. At 70 mm cover, after 1 year the plain bars had rust over ,10% of their surface, increasing to about 75% after 2 years. In comparison, the galvanized bars with 20 mm cover had corrosion over about 70% after 2 years, but at 40 and 70 mm cover the extent of corrosion was ,10% with little further increase after 2 years. All of the galvanized bars showed signs of white rusting, indicating corrosion of the zinc layer, especially in the accelerated tests. After 3 years of continuous exposure, the performance of the galvanized bars was very much better than that of the plain steel bar. For the galvanized bars, less than 10% of their surface was affected by red rust (iron) corrosion at covers of 20 and 40 mm. At 70 mm cover, there was negligible such corrosion although zinc corrosion products were present. It was concluded that, although galvanized bars were superior to black bars in the accelerated tests, the galvanizing did not provide complete protection against pitting. In the marine exposure, the level of attack was also less than that on the plain bars, but the presence of localized rusting suggested some progress in the corrosion of the zinc coating. Shimada and Nishi [37] studied the effects of different quality concretes on the performance of galvanized and black steel bars in an offshore marine exposure. Both high- and low-density concrete were used with the high-density concrete being pre-cracked prior to exposure. Variations in w/c ratio (0.56 and 0.60) and cover to the bars (20 and 40 mm) were also incorporated. On removal of the beams after 2 and 5 years, the presence of cracks and the extent of corrosion of the bars were assessed. After 2 years, no cracking of any of the beams was observed, although after 5 years cracks had developed in all portions of the beams exposed at the splash zone. The cracks in the galvanized beams were slightly larger than those with plain steel, indicating that, in such conditions, survival of the zinc coating could not be expected. The corrosion assessment revealed that, after 2 years, the extent of corrosion of the galvanized bars was negligible but, after 5 years, both the black steel and the galvanized steel were attacked due to the high chloride presence. Even though the corrosion rate of the galvanized bars was much lower than that for the plain bars, the tendency to cracking of the beams was about the same. It was noted that this effect might be due to hydrogen liberation that occurred at steel-zinc couples. It was thus concluded that the zinc coating provided protection during the first 2 years of exposure, but its effect could not be expected after long-term exposure in this most corrosive offshore environment. Ru¨ckert and Neubauer [38] studied the reversal of the electro-potential of zinc and its corrosion rate at 60 8C. Cement mortar cylinders (0.5 w/c) were cast with additions of 2% CaCl2 by weight of cement. Each specimen contained one galvanized bar and one black steel bar, which were short-circuited through
162
S.R. Yeomans
a 0.1 V resistor. The specimens were immersed in saturated Ca(OH)2 solution at 60 8C and then exposed in open air. The galvanic current flowing between the galvanized and black steel bars and the corrosion potential were measured over a 30-week period. The instantaneous corrosion potential was measured after disconnecting the bars. For galvanized bars, the Ecorr value was initially in the range of 2 600 to 2900 mV (vs SHE), increasing to 2 300 to 2400 mV after 30 weeks. For black steel, the same value was 2380 to 2190 mV initially, then finally in the range 0 to þ50 mV. In chloride-containing mortars, the stabilized value for galvanized bar was in the range 2 600 to 2800 mV and for black bars 2250 to 2450 mV. In both the chloride-containing and the chloride-free mortars, no potential reversal was observed. Measurement of the galvanic current between the bars showed that, in chloride-free mortars, the initial current was 6 mA/cm2 but that this reduced to zero at the end of the test period. This result indicated that more zinc removal had occurred in the first week of the contact with concrete and then a gradual reversion occurred. In specimens where the galvanized bar was not coupled with the black bar, no effect was observed on the galvanized bar. However, when the bar was coupled, pronounced attack on the zinc coating was observed. In chloride-containing mortars, the initial galvanic current was around 13.8 mA/cm2 and reduced to ,0.5 mA/cm2 after 30 weeks. Metallographic investigations confirmed the current measurements and corrosive attack was observed, which was greater than that in chloride-free mortars. Also in chloride-containing mortar, after decoupling, the galvanized bar reached a potential of 2750 mV while the black steel was at about 2500 mV. From this it was inferred that the zinc was able to cathodically protect the black steel, which was confirmed from visual examination that revealed very few corrosion sites on the black steel when coupled to the galvanized bars. Finally, at an increased temperature of 60 8C, no reversal of potential of the zinc was observed in either the chloride-containing or the chloride-free mortar. A major review and experimental study of zinc-coated concrete reinforcement was undertaken in Finland by Sarja et al. [39]. A variety of tests were performed using Finnish Portland cement and hot-rolled bars plain steel bars and both spraygalvanized and hot-dip galvanized bars. The tests included durability of the bars during transport, handling and bending, chemical reactions with fresh concrete, chromate treatment, bond strength and corrosion studies in seawater under repeated loading. The corrosion testing indicated that both spray and hot-dip coatings protected the bars under moist conditions and in water, but corrosion did eventually occur in seawater. The zinc was able to delay the onset of corrosion, but not prevent it completely in high chloride conditions. It was estimated that, in
Laboratory and Field Performance of Galvanized Steel in Concrete
163
these circumstances, the durability of concrete with galvanized bars can be expected to be between 10 and 30 years. Gonzalez et al. [40] studied the effect of the characteristics of the zinc-alloy coating on the corrosion of galvanized steel in concrete. Three types of coating were used: a conventional hot-dipped coating with a 30 mm eta layer at the surface, a modified coating with a thickened underlying zeta layer, and an annealed coating in which the zeta layer had grown through to the surface by consumption of the eta layer. Mortar samples (0.58 w/c) were cast, some with 2% calcium chloride additions, and exposed for 400 days at 95% RH. In chloride-free mortar, very low corrosion currents were observed for all coatings and it was apparent that the coating structure had little effect. However, the presence of chlorides radically changed the corrosion behaviour of the various coatings. While the overall corrosion current was about 10 times higher in the presence of chloride, the corrosion intensity was much higher for the annealed coating than for the other types. That is, it was apparent that the Fe –Zn alloy layers exposed at the surface of the annealed coating were more corrosion-prone. It was concluded that the coating with a thick and homogeneous pure zinc (eta) outer-layer was superior to the other coatings, especially the annealed coatings, and that the Fe–Zn intermediate alloys are the weakest part of the galvanized coating in contact with concrete. This work also confirmed that, while the presence of chlorides increased the risk of localized attack, the threshold value to cause corrosion is higher for galvanized steel than for bare steel. Mikami et al. [41] reported studies that compared the corrosion resistance of galvanized (80 mm coating) and bare steel reinforced concrete exposed for 2 years with a marine environment in Okinawa. Concrete beams (25 MPa and 0.55 w/c) with covers to the bars of 20, 40 and 70 mm were pre-cracked prior to exposure. Although extensive attack occurred on both the black and galvanized bars under these very severe conditions, estimations of the rust ratio and pitting depth on the bars indicated that galvanized bars performed somewhat better than the black bars. It is to be noted, however, that the record of rusting on the bars was based on red rust for the black steel and white rust (i.e. zinc corrosion products) on the galvanized bars. It is thus quite likely the severity of corrosion for the galvanized bars is in fact over-estimated. Similarly, Makita et al. [42] also reported that galvanizing provided good protective performance in the marine environment, but that this was not always satisfactory in the splash zone of the marine environment. Studies of the fatigue of several types of reinforcement, including uncoated steel bar and galvanized bar, recovered from concrete beams subjected to corrosion fatigue in salt water solutions were described by Roper [43]. Decreases in the fatigue life of corroded bars were explained on the basis of the rapid development of pits from which cracks had propagated, and an increase in the propagation rate of cracks. It was specifically noted that the longevity of
164
S.R. Yeomans
the galvanized bars, in particular, was related to the delay in crack initiation and propagation under cathodic protection conditions. Nu¨rnberger [44] reviewed the corrosion behaviour of galvanized steels in contact with a variety of building materials including concrete, plaster and construction materials containing magnesia. Overall, zinc provided good protection in these cementitious-type materials, although it was noted that attack only occurred when stable protective layers either could not develop or were lost from the surface due to the effects of their environment and exposure conditions, for example poor design and mechanical effects. Hildebrand and Schwenk [45] studied the long-term corrosion behaviour of galvanized reinforcement in mortars immersed in aerated 0.5 M NaCl solution for up to 5 years. Galvanized sheets with a coating thickness of about 50 mm were embedded in Portland cement mortars of 0.57 and 0.77 w/c. One set of specimens was cast using pickled steel sheet for comparison. The 0.57 w/c specimens, which were completely immersed in saltwater, showed a stabilized potential of the galvanized sheet of about 2700 mV (vs SHE) and for black steel about 2300 mV. Similar values were also observed in the 0.77 w/c specimens in the moist curing condition. In the case of partially immersed specimens, the potential of galvanized sheet was around 2500 mV and for the steel sheet it was around 2200 mV in both w/c ratios. From this data, it was concluded that the hardening condition and w/c ratio of the mortar had no significant influence on the corrosion potentials. Visual observation and associated measurements in this work revealed that the galvanized sheet had not suffered significant corrosion damage, irrespective of w/c ratio and curing conditions. For the steel samples, it was observed that the mortar had cracked at the water/air line due to the growth of (steel) corrosion products. The presence of the galvanizing significantly retarded this effect. While it was concluded that pitting corrosion was prevented by galvanizing, it was noted that up to 25 mm of the zinc coating was removed in these exposure conditions, indicating that galvanizing is able to delay, but not completely prevent, corrosive attack. The corrosion properties in concrete of two different types of galvanized bars were studied by Maahn and Sorensen [46]. Bright galvanized steel with a 20 mm pure zinc layer at the surface and a grey coating with iron –zinc compounds in the surface were embedded in 0.5 w/c mortar and exposed to several chloride-rich environments and carbon dioxide. The results showed that, in carbonated concrete, both the bright and grey coatings remained passive and the corrosion rate was of the same magnitude as in non-carbonated concrete. However, the presence of chloride created a different effect. Compared with plain steel bar, the bright galvanized steel was more resistant to chloride than the plain bar, although the grey coating was less resistant. It was also found difficult to inhibit hydrogen evolution from grey coatings embedded in concrete. This work clearly
Laboratory and Field Performance of Galvanized Steel in Concrete
165
demonstrated the importance of the presence of a layer of pure zinc (eta) on the surface of galvanized bar to allow for passivation of the surface in the highly alkaline environment of cement. Pfeifer and his co-workers [47] carried out extensive macrocell corrosion studies on galvanized and black steel reinforcement under accelerated conditions for 50 weeks. For cyclic wet and dry tests, 300 £ 300 mm slabs with two mats of reinforcement were cast with cover to the top mat of 25, 50 or 75 mm. Concretes of variable w/c ratio (0.50, 0.40 and 0.32) were used. In each slab, either galvanized bars or black bars were used in both mats or galvanized bars were used in the top mat and black bars in the bottom. The mats were electrically coupled through a 10 V resistor. A concentrated saltwater solution (15%) was ponded for 100 h followed by 68 h drying at 38 8C. In slabs with black bars in both mats at 25 mm top cover and 0.5 w/c, the corrosion current was 300 mA after 20 weeks reducing to 275 mA after 50 weeks. With galvanized reinforcement in both mats, a peak current of 40 mA was measured after 9 weeks although this dropped to near zero after 30 weeks. With galvanized bars in the top mat and black bars below, the corrosion current after 35 weeks was 300 mA, reducing to 100 mA after 44 weeks. It was noted that none of the slabs with 50 or 75 mm cover to the top mat showed any signs of corrosion during the test period. In slabs with galvanized bars in both mats, an initial potential of 2650 mV (vs CSE) was measured, increasing to 2800 mV then decreasing 2500 mV after 30 weeks. In slabs with only black bars, the initial potential was 2130 mV, increasing to 2580 mV after 30 weeks. A similar behaviour was observed in slabs cast with a mix of galvanized bars in the top mat and black bars in the bottom. In 0.5 w/c slabs with galvanized bars in both mats, zinc corrosion products were observed on the top of the bars. These products were identified as a mix of zinc oxide and zinc hydroxychloride. In slabs with a galvanized top mat and black steel bottom mat, severe pitting of the galvanized bars was noted with red rust (steel) corrosion over about 1.5% of the galvanized surface with zinc corrosion products along the length of the bar. In equivalent slabs with black bars in both mats, although the lower mat bars were free of corrosion, the top mat bars were severely corroded. At the commencement of the experiment, the mat-to-mat resistance of galvanized bars was the same as that of the black bars, i.e. about 200 V. After 48 weeks, this resistance increased to 2000 –2600 V for galvanized bars but was only 450– 660 V for the black bars. This indicated that galvanizing created a 3–4 times increase in resistance across the concrete compared to black steel bars. In cyclic salt-water exposure tests undertaken as part of this programme, reinforced concrete beams were cast with either galvanized bars or black bars embedded at a cover of 25 mm. A 15% salt solution was sprayed for 4 h per day (for 270 days) on the vertical face of the beam, the other faces being exposed to
166
S.R. Yeomans
the air. This created a chloride-induced corrosion cell with the bars on the sprayed side becoming anodic to the bars on the air-exposed faces. After 30 and 370 cycles, the potentials of the galvanized bars were 2 680 and 2 580 mV, respectively. After 270 cycles, the corrosion current for galvanized bars was about 100 mA whereas for black bar it was 1000 mA after 200 cycles. The chloride content measured at the level of galvanized bar was 0.32% by weight of concrete, about 10 times the threshold level for initiation of corrosion in black bars. Active corrosion of the zinc was observed after 25 cycles and, on removal, the bars were coated with both light- and dark-coloured corrosion products. From the results of this dual series of tests, it was concluded that, when only galvanized bars were used in concrete construction, very low corrosion currents developed and only minor attack on the zinc was likely. However, when a mix of galvanized and black bars was used in the same construction, larger corrosion currents developed, the zinc was attacked more quickly and the overall behaviour was more or less similar to that of unprotected black bars. In Japan, Swamy and his co-workers carried out long-term studies on galvanized reinforcement in the aggressive off-shore environment of Tokyo Bay [48]. They studied the effect of concrete quality, concrete cover, crack width and the effect of concrete cleavage due to corrosion of galvanized bar in 100 £ 100 £ 1000 mm cast beams. Both uncoated bars and chromate passivated galvanized bars were used with additions to the mix of 0.2–0.4% NaCl. Concretes of 0.50 w/c (high quality) and 0.60 w/c (low quality) were used and the cover to the bars was 20 and 40 mm. Artificial cracks of widths 0.2 and 0.02 mm were introduced at the tensile zone of the beam. After 2 years site exposure, no visible cracks in the beams were observed although, after 5 years, irrespective of concrete quality, the chloride concentration and type of bars, all beams were observed to have cracked. In galvanized reinforced concrete of 0.56 w/c and additions of 0.2% NaCl, a 2 mm £ 250 mm crack was observed while, in 0.60 w/c concrete, a 3 mm £ 400 mm crack was observed. At both exposure periods, the maximum pit depth on the galvanized bars was much lower than that of uncoated bars; for galvanized bars after 5 years, this was in the range of 1.0–1.9 mm whereas, for uncoated bars, the range was 3.0– 3.3 mm. In the lower quality concrete, it was observed that the area rusted on galvanized reinforcement was greater than that on uncoated reinforcement, i.e. 70– 75% compared with 62–66%. In high-density concrete, however, the behaviour was the same. Based on these long-term field studies, it was concluded that the effect of the galvanized coating on delaying the occurrence of concrete cracking could not be expected in the most corrosive offshore environments. Further studies of the long-term performance of galvanized reinforcement in concrete exposed to various environmental conditions were undertaken in Japan by Kashino [49]. Three concrete types with lean, rich and standard mix
Laboratory and Field Performance of Galvanized Steel in Concrete
167
proportions were used with covers to the bars of 10 and 20 mm in both cracked and uncracked specimens. In some of the specimens, artificial seawater and artificial lightweight aggregate were used. Specimens were exposed for 5 years in various locations including a suburban site, an industrial atmosphere, a cold region and a marine exposure. Galvanized bars removed from uncracked concretes in all mix proportions containing additions of 0.1% chloride (by weight of sand) were in good condition. However, in concretes cast using artificial seawater and with 0.1% chloride, the galvanized reinforcement was severely corroded, even in the uncracked condition. In standard mix and in rich mix concrete with cracks up to 0.1 mm and 0.1% chloride, the galvanized bars were in good condition whereas, in artificial lightweight aggregate concrete, the bars were severely corroded. In uncracked concretes exposed to the industrial atmosphere and in the cold region, the galvanized reinforcement was in good condition at both depths of cover. In the marine environment, at 10 mm cover the galvanized bars were severely corroded although at 20 mm cover no corrosion was observed. From these long-term field exposure studies, the conclusion was drawn that the galvanized reinforcement had performed well in comparatively low chloride content concrete. In the UK, the Building Research Establishment undertook a 5-year field exposure study on galvanized reinforcement at different exposure sites [50]. Various prisms and slabs were cast to assess non-bent coated bars and bars under simulated bending and tying conditions. Both black bars and galvanized bars, coated to BS729 with a minimum coating weight of 700 gm/m2, and cement mortars of 0.6 and 0.75 w/c were used. Sodium chloride at dosages of 5.4 and 1.6% by weight of cement was added to the mix and covers of 10 and 20 mm were employed. After 5 years, bars were recovered and the weight loss calculated. Galvanized bars removed from the prisms containing 5.4% NaCl were almost completely corroded with a weight loss of 13.9 g, which was greater than that of the uncoated bar with a weight loss of 5.9 g. Only in the case of nil-chloride concrete was the weight loss of the galvanized bar (0.31 g) less than the uncoated bar (0.50 g). Estimation of pit frequencies and depth indicated that, in both the lean and rich mixes with 5.4% NaCl, the behaviour of galvanized bars and uncoated bars was the same. For the lower chloride levels (1.6% NaCl), the galvanized bars performed significantly better than the uncoated bars. Visual examination of the slabs indicated that, in highly chloride-contaminated concrete, galvanized steel did not perform as well as uncoated steel. In this case, the entire reinforcement was corroded over 90% of the surface and in many places the underlying steel was severely corroded. Where corrosion had occurred, the pH of the steel/concrete interface was less than 8. In the late 1980s, Pennsylvania DOT inspected a large number of bridge decks, constructed during the previous decade, that had been protected with galvanized or epoxy-coated bars [51,52]. In these reports, both coating systems were found to be
168
S.R. Yeomans
performing adequately at the time of inspection, with no discernable distress to the concrete. Other studies undertaken by the Pennsylvania DOT [53] evaluated the performance of six different bridge deck protective systems. The project surveyed some 169 bridge decks throughout the state and involved a mix of visual inspection, physical testing and chloride determinations. Of this number, 25 decks containing galvanized reinforcement were surveyed 10 years after construction. In all these decks, both the top and bottom mats of the reinforcement were galvanized, as were bar chairs and tie wires. The visual assessment of the galvanized decks produced a rating in the range of 4.2–4.8 (of 5), indicating that the bars were in good condition. Half-cell potentials were in the range of 2210 to 2 310 mV (vs CSE). In most cases, the galvanized reinforcement was in excellent condition, even though some of the potentials indicated probable active corrosion. In areas where potentials ranged from 2500 to 2600 mV and watersoluble chlorides were above 1.1 kg/m3, subsurface delaminations were generally observed and metallographic studies indicated that zinc was consumed and corrosion had progressed to the base metal. For uncoated steel, a chloride content at the bar of 0.06 kg/m3 indicated only moderate corrosion although when in the range 1.7–4.0 kg/m3 severe-to-very severe corrosion was observed. In comparison, galvanized reinforcement remained unaffected by chlorides in the range 2.0–7.0 kg/m3 although in one deck the galvanized bar was severely corroded in the presence of 3.3 kg/m3 chlorides. In locations where mild-to-moderate corrosion was observed, the average coating thickness remaining on the galvanized bars was around 300 mm, indicating that, after 10 years of aggressive field exposure, the galvanized reinforcement more than satisfied the minimum coating requirements of appropriate Standards. From this work, it was concluded that the galvanized reinforcement was in excellent condition, despite high chloride contents in the surrounding concrete, and that, based on deterioration rate and life expectancy, the galvanized reinforced decks were generally performing as well as could be expected. In 1989, Stark published a major review of parameters having an influence on the corrosion resistance of steel in concrete [54]. This extensive report dealt with a range of issues affecting the durability of concrete and included a comparison of plain and galvanized steel. All tests used 300 £ 300 £ 150 mm slabs with a single layer of reinforcement. Two types of cement were used with w/c ratios of 0.35, 0.45 or 0.6. A variety of admixtures and additives were mixed with the concrete. Two-week cycles of ponding with 4% salt solution and drying was continued for up to five and half years, although continuous storage at different RH were used for calcium chloride-contaminated specimens. The results, as they related to galvanizing, were that galvanized steel exhibited slightly greater time to active corrosion than black steel under similar conditions. For example, in a 0.45 w/c
Laboratory and Field Performance of Galvanized Steel in Concrete
169
concrete with 25 mm cover, active corrosion of the black steel began after 30 cycles and after 34 cycles with the galvanized bar. For the 0.6 w/c concrete at 50 mm cover, black steel corroded after about 20 cycles and the galvanized after 36 cycles. These differences were related to the higher chloride tolerance of the galvanized coating. The onset of active corrosion for the galvanized bar was identified by a rapid potential shift to large negative values, with a gradual rise over time towards the levels typical of black steel. This latter shift was taken to represent the progressive corrosion of the coating and the exposure of iron-rich layers in the coating and, eventually, in the underlying steel itself.
6.5. The 1990s Nu¨rnberger and Beul [55] studied the behaviour of black, galvanized and PVCcoated reinforcement embedded in cracked concrete and exposed to a variety of seawater, de-icing and industrial environments. The protective effect of galvanizing exposed to chloride contents up to 1.5% by weight of cement, and in the presence of larger crack widths, was noted. Further protection was provided where the concrete was not cracked. Rasheeduzzafar et al. [56] conducted a 7-year pilot scale programme on mild steel, galvanized steel, epoxy-coated steel and stainless clad steel embedded in concrete contaminated with chlorides in the range 2.4–19.2 kg/m3. Prisms were exposed to a coastal environment in Eastern Saudi Arabia and monitored for cracking. After 7 years testing, the width of cracks that had formed in the concrete specimens was classified as in Table 5. Specifically, 58% of the 2.4 kg/m3 galvanized specimens were uncracked, while only 13% of the black steel specimens were uncracked.
Table 5: Crack width assessment with use of different reinforcing steels [56]. Admixed chloride content (kg/m3)
2.4 4.8 19.2
Crack width classification Mild Steel
Galvanized steel
Epoxy coated steel
Stainless clad steel
Medium Heavy Spalling
Fine Medium Wide
No cracking No cracking Wide
No cracking No cracking No cracking
Code: Fine, Cracks less than 1 mm; Medium, Cracks 1– 2 mm wide; Wide, Cracks 2– 3 mm wide; Heavy, Cracks 3 –5 mm wide.
170
S.R. Yeomans
Further, for the 19.2 kg/m3 concretes, 87% of the galvanized specimens were below the heavy cracking condition and none was at the spalling condition, while all of the black steel specimens were at the heavy cracking condition. In this work, it was found that the amount of cracking was reduced with the use of galvanized bars when compared with black steel and that a delay in the onset of cracking occurred. The black bar samples cracked after 65 days, whereas the galvanized bars cracked after 172 days. The authors concluded that the use of galvanized bars was able to delay or postpone cracking and spalling of concrete by only a finite period of time, and that it was not a permanent solution to corrosion protection. Similar quantitative results were reported for a seawater study using different types of steels, including galvanized, were reported by Baker [57]. In a continuation of the ILZRO-sponsored evaluation of galvanized steel in concrete bridge decks in 1974, 1976 and 1981, a further round of surveys was undertaken in 1991 [58]. In this phase, six bridges in Iowa and Pennsylvania that had been subject to de-icing salts and one bridge in Florida with seawater exposure were again surveyed. The previous studies indicated that more than 90% of the original galvanized coatings remained on the bars in these bridges after 7–23 years exposure to high levels of chloride in the range 0.58–1.95% by weight of cement. A summary of the finding of this present survey follows. *
*
*
*
Boca Chica Bridge, Florida (19 years). In areas of the deck containing uncoated bars, half-cell potentials were more negative than 2350 mV (vs CSE), indicating active corrosion. The chloride concentration at the level of the galvanized bars was in the range 0.264– 0.400% by weight of cement, well above the threshold value of 0.15%. No corrosion was observed. The surface of the galvanized reinforcement was uniform and the overall coating thickness averaged more than 100 mm. Ames Bridge, Iowa (24 years). This bridge also contained uncoated bars at potentials more negative than 2350 mV, indicating severe corrosion. The cover to the galvanized reinforcement was 83 mm and very little corrosion of the zinc was observed. The remaining coating thickness was an average of 119 mm and the chloride concentration at the bar was 0.036–0.714%, of the order of five times the threshold level. Athens Bridge, Pennsylvania (18 years). Galvanized bars were used throughout this structure and chloride levels at the bars were in the range 0.079– 0.750%, but no corrosion was observed. Chloride levels were generally 3 – 5 times the threshold values for uncoated steel. The bars had very thick galvanized coatings, which were virtually untouched. Coraopolis Bridge, Pennsylvania (19 years). Two decks were surveyed where only galvanized reinforcement had been used. Visual and metallographic inspection revealed no corrosion of the galvanized bar and the chloride
Laboratory and Field Performance of Galvanized Steel in Concrete
*
*
171
concentration at the reinforcement was in the range 0.157–0.886%. No corrosion was observed on the galvanized bars and at only a few isolated places were spalling and cracking observed. The surface of the galvanized bars was uniform and in very good condition, with little or no loss of coating. Hershey Bridge, Pennsylvania (16 years). Two galvanized decks were surveyed. Moderate-to-high chloride concentration was observed at the reinforcement, some up to about twice the threshold value, but no corrosion was evident. The galvanized bars were in good condition and no deterioration of the coating was apparent, even in areas where the cover was low at 44 mm. Tioga Bridge, Pennsylvania (17 years). Two galvanized reinforced spans were surveyed. The chloride ion concentration was generally quite low, but in some areas was very high (to about 1.45%) and well above threshold values. Corrosion of the galvanized bar was observed in one sample where severe surface cracking had occurred. The thickness of the coating in this sample was 38 mm. Metallographic examination of other samples revealed the coatings to be in good condition elsewhere.
These surveys highlight the importance of gathering long-term field data. It was concluded that, in these circumstances, galvanized reinforcement would sustain only superficial corrosion in sound concrete, even when the level of chlorides present was high. Moderate corrosion of the galvanized coating may occur if the concrete cover was low and more severe attack would occur in areas of very low coating thickness and the presence of discontinuities in the coating. In nearly all cases, limited corrosion of the galvanizing had occurred after nearly 25 years of service and the amount of zinc remaining on the bars in situ was in excess of the minimum values required for freshly galvanized materials (about 85 mm). It was also apparent that, while the chloride concentration in the various bridge decks had increased continuously over the inspection period, and especially since 1981, the performance of the galvanized reinforcement had not noticeably changed. Long-term exposure tests on galvanized bars in concrete beams were reported by Mu¨ller [59]. A 10-year test programme studied the performance of galvanized and plain steel bars in concrete where black and galvanized bars were coupled, the concrete was cracked and chlorides were present. The galvanized bars had a coating thickness of 120 mm, cover was varied between 10 and 20 mm and the concrete had a 0.6 w/c. The results indicated that the reaction between zinc and wet cement caused zinc loss of 12 mm for concrete containing less than 0.5% chloride (by weight of cement) and 20 mm for 2% chloride concrete. After 2 years in carbonated concrete, black steel started to show corrosion with pits up to 0.5 mm deep. After 6 years, corrosion had extended to about 50% of the surface of the black bars while galvanized bars showed only negligible corrosion over about
172
S.R. Yeomans
5% of their surface. The loss of zinc from the galvanized bars was not more than 30– 50 mm, with the greatest loss on bars coupled to black steel. In this study, chlorides had a greater effect on the corrosion of black bars than galvanized bars. After 2 years, at 2% chloride concrete, corrosion on black bars covered about 40% of their surface while it was about 15% for galvanized bars. At longer times, the corrosion extended but it was always less on the galvanized bars. After 6 years, the total loss of zinc was about 55 mm, only slightly more than at 2 years. These results showed that galvanizing provided efficient protection to the reinforcement in carbonated concrete with and without cracks. To achieve this, however, it was noted that contact between galvanized and black steel must be avoided. In high-chloride concrete, although corrosion is delayed a few years (in this case), once corrosion initiated there was no significant further effect of galvanizing. The pre-condition for effective protection was given as a sufficient thickness of the zinc layer, not less than 100 mm. Northwood et al. undertook laboratory-scale tests on concrete specimens with black and galvanized reinforcement exposed to the atmospheric, tidal and tropical marine environment in Singapore [60,61]. The 80-week programme involved five concrete types embedded with galvanized and black steel wires of 1.8 mm diameter. The coating thickness of the galvanizing was quite thin at 35 mm. The corrosion assessment revealed that the galvanized specimens in all concretes had a lower rust area on the bars (by more than half) and corroded at a later stage compared with the black steel. In atmospheric exposure, the galvanized bars were in good condition and, where corrosion did occur in other samples, this was only in a few of the galvanized bars close to the concrete surface in lower quality concretes. The findings indicated that galvanizing was able to provide protection to the steel by delaying the initiation of corrosion by more than half. However, chloride-induced corrosion tended to cause intense local attack and pitting of the base metal. Saiful Islarn and Kaushik [62] studied the performance of galvanized reinforced concrete under simulated tropical marine conditions. Cylinders in two concrete types (0.40 and 0.48 w/c) were cast using seawater as well as plain water for mixing and curing. Both uncoated and galvanized bars were embedded in the cylinders at covers of 15, 25 and 40 mm. The specimens were subjected to accelerated testing by both alternate wetting and drying and continuous submersion in artificial seawater of concentrations 1N, 5N and 10N NaCl. Each wetting and drying cycle consisted of 12 h immersion in seawater and 12 h drying at 50 8C. After periods of exposure from 3 to 18 months, the performance of the galvanized bar was better in both testing conditions, although it was noted that the amount of corrosion in the submerged test was minimal. Galvanizing delayed the onset of corrosion and the actual amount of corrosion was 30–50% less than on the bare steel. For example, in the alternate wetting and drying test in
Laboratory and Field Performance of Galvanized Steel in Concrete
173
10N solution, the weight loss of the galvanized reinforcement in 0.4 w/c concrete was 0.110, 0.08 and 0.038% at 15, 25 and 40 mm covers, respectively, and the corresponding weight loss of bare steel was 0.146, 0.114 and 0.032%. In the 0.48 w/c concrete, the weight loss at the same covers for galvanized reinforcement was 0.12, 0.095 and 0.032% and for bare steel it was 0.28, 0.248 and 0.120%. Some areas of red rust corrosion were present on the galvanized bars in the highest salt environments, although the remainder of the zinc was covered with a white product, indicating that the zinc was actively corroding. These results showed that the galvanized steel behaved in a similar way to bare steel in the higher quality concrete although, in the lower quality concrete, the weight loss of the galvanized steel was significantly less than that of bare steel. It was concluded that, during long-term exposure in splash/tidal zone situations, galvanizing certainly delayed the onset of corrosion although it may not provide complete long-term protection. Examination of galvanized reinforcement in a 19-year-old replacement bridge deck on Route 495 in New Jersey was undertaken in 1994 [63]. The results of the investigation revealed the galvanized bar in the deck to be in excellent condition in that there was no evidence of attack on the coating or the bars. Chloride levels were moderate in the range 0.17% at 12 mm depth to 0.05% at 37 mm depth. Cracks in the decks were not shown to be associated with the reinforcement. The zinc coating remained relatively uniform and was undamaged with 125–200 mm remaining on the surface after nearly 20 years. Yeomans [64] has undertaken comparative studies of black, epoxy-coated and galvanized steels in concrete via two accelerated tests, one cyclic wetting and drying in a 3.5% NaCl solution, the other in continuous salt fog. Cylinders were cast using 25 MPa concrete with 0.80 w/c. The galvanized bars had a typical coating thickness of 110 mm. Two types of specimens were cast: one with four similar bars embedded vertically at 10 – 15 mm cover used primarily for half-cell potential surveys, the other with four different lengths of similar bar placed horizontally with covers varying from 8 to 32 mm primarily for the corrosion assessment. In the wet and dry test, each cycle consisted of immersion for 3 days in 3.5% NaCl solution, followed by oven drying for 4 days at 60 8C. In the salt fog experiment, specimens were continuously exposed to a 3.5% salt fog at 40 8C and 100% RH. Each test was conducted over about 180 days Of the three reinforcement types, black steel was the most susceptible to corrosion, even at relatively low chloride levels. The average corrosion rate for the black steel, based on weight loss, was 272 mm/year in the wet and dry but only 14 mm/year in the salt fog. After 10 days wet and dry exposure, the corrosion potential for black steel was about 2600 mV (CSE), indicating active corrosion. In comparison, the galvanized bars showed a delay in the onset of corrosion and a much reduced corrosion activity at depth in both exposures. In the wet and dry
174
S.R. Yeomans
test, where the black steel was actively corroding after 10–20 days, galvanizing protected the base steel for 70– 80 days; in the salt fog, corrosion of the black steel commenced after about 45 days while the galvanized was unaffected when the test was discontinued at 180 days. Measurements of the zinc loss from the coating from the wet and dry tests indicated that, at 65 days, about 10 –15% of the coating had dissolved, essentially the eta layer, although by 103 days the coating had all but completely dissolved. Correlation between the onset of corrosion and the chloride content at the bars also indicated that the chloride level required to cause corrosion of galvanized steel, at any depth of cover, was at least 1.5 times that required for corrosion of black steel and realistically was likely to be more than 2.5 times greater. The uncertainty here was that a lower bound value only, not an absolute ratio, could be inferred from the data available for galvanized bars in several cases. A summary of this data is given in Table 6. Half-cell potentials of the galvanized bars showed a shift from the starting potential of about 2600 mV to about 21100 mV, at which corrosion of the zinc was active, followed by a gradual rise in potential over an extended period of time to about 2650 mV when the coating was unable to provide further protection to the base steel. The delay in these potential shifts for the galvanized bars was indicative of the period over which the coating provided protection to the steel. This was of the order of 4– 5 times the period for corrosion of black steel in equivalent concrete and exposure conditions. Another interesting aspect of this work was that the gradual rise in potential (from 21100 to 2 600 mV) for the galvanized bars could be correlated with the progressive dissolution of the coating and exposure of the increasingly iron-rich alloys zeta, delta and gamma in the coating.
Table 6: Depth of corrosion and chloride levels (by weight of cement) for corrosion of black and galvanized steel in concrete [64]. Exposure Condition
Bar type
41 days
98 days
132 days
Depth of Chloride Depth of Chloride Depth of Chloride corrosion (%) corrosion (%) corrosion (%) (mm) (mm) (mm) Wet Black and dry Galvanized
20 0
0.06 . 0.15
30 10
0.10 . 0.15
40 10
0.10 . 0.17
Salt fog
15 0
0.04 0.10
20 10
0.08 . 0.15
15 0
0.10 . 0.15
Black Galvanized
Laboratory and Field Performance of Galvanized Steel in Concrete
175
The following conclusions were drawn from these studies: *
*
*
Galvanized reinforcement can tolerate chloride levels at least 2.5 times higher than those causing corrosion of black steel in equivalent concrete and exposure conditions. Galvanizing protects steel in concrete during the period over which the coating progressively dissolves, thereby delaying the onset of corrosion of the steel substrate. In these conditions, sacrificial protection of exposed steel was afforded to a distance of at least 8 mm. The total period over which galvanizing delays the onset of corrosion of reinforcing steel in concrete was of the order of 4–5 times that for the corrosion of black steel.
Short et al. [65] examined the compatibility of different zinc-alloy coatings on steel embedded in cement pastes, particularly under chloride-contaminated conditions. This involved alloying the zinc with elements such as aluminium, iron, nickel or cobalt, embedding the coated steel in cement paste and taking measurements of corrosion potential and polarization resistance over a 4-month period. The coatings were applied by either hot dipping or electroplating and their thickness was varied from 8 to 30 mm. The effect of chromate treatment was also studied. Chlorides were introduced into the mix at concentrations of 0, 0.4 and 1.0% by weight of cement. A constant 0.4 w/c was used for the pastes and hydroxyl ion concentrations were equivalent to a pH of 13.8, thus providing a highly alkaline environment for the coatings. At the end of the test period, corrosion products were examined using SEM with an energy dispersive X-ray spectrometer. Typical data are given in Table 7.
Table 7: Corrosion data of zinc and zinc-alloy coatings [65]. Coating type
Chloride concentration Pure zinc Pure zinc þ chromate Zn –8% Fe Zn –1% Co Zn –1% Co þ chromate Zn –12.5% Ni Zn –12.5% Ni þ chromate
Thickness (mm)
30 8 12 8 8 10 10
Ecorr mV (vs SCE) average values
icorr (mA/cm2) average values
0% 0.4% 1% 0% 0.4% 1% 2 878 21038 21081 3 9 4.1 2 600 – – 0.05 – 0.1 2 846 2 970 2837 16.9 18.2 21.8 2 846 21029 21040 1.6 2.0 5 2 859 2 819 2913 0.3 0.1 0.3 2 500 – – 0.5 – 1.0 2 350 – – 0.5 – 0.7
176
S.R. Yeomans
These results indicated the beneficial effect of chromate treatment in significantly reducing the corrosion rate, especially for the pure zinc and the zinc-cobalt alloy, but not the zinc –nickel alloy. Variations in the corrosion potentials indicated that, for some alloys, the potentials were close to that for steel in a similar environment (around 2400 mV), such that the level of sacrificial protection would be considerably reduced, if not eliminated entirely. It is interesting to note that the results for the zinc–iron alloy were no better than those for the pure zinc. This confirmed data from other work where there was little difference in corrosion rate for galvanized coatings when either the pure zinc eta phase or the inner zeta phase (, 7% Fe) was the outermost layer in the coating. Examination of the corroded surfaces revealed high corrosion rates and extensive general corrosion for both the pure zinc and zinc –iron alloy in cement containing 1% chloride. General corrosion products of zinc hydroxide and zinc hydroxychloride [Zn5(OH)8Cl2] were evident. In the case of the zinc–nickel alloy, some cracks were observed in both chromate-treated and non-chromate-treated samples, which explains why the electrochemical data were the same. Chromatetreated zinc-cobalt alloy sample showed bright surface with only slight general corrosion. From these studies, it was concluded that zinc and zinc–cobalt coatings with good chromate films had much greater corrosion resistance (in these conditions) than other types of zinc coatings. The corrosion rates of non-chromatetreated coatings were such that it was predicted they would not provide long-term protection. In 1993, Hime and Machin reported the formation of zinc hydroxychloride-II as a corrosion product on galvanized steel components in masonry structures made with mortar containing very high concentrations of chlorides [66]. Cracking of these structures had occurred at the location of corroded galvanized steel, without the presence of red rust corrosion products. It is known that this corrosion product expands by a factor of about 250% when it forms, which is somewhat greater than that of iron corrosion products (100% or more) and far greater than the 50% expansion for the formation of zinc oxide. It was noted that the volume differentials between zinc oxide and zinc hydroxychloride-II depend on the chloride level and that the formation of such a product may generate significant tensile stress around the galvanized bars once corrosion initiates, with resultant cracking of the concrete. These observations may account for the inconsistencies reported in investigations of the performance of galvanized reinforcement where such cracking occurs in masonry or concrete. Burke reported on studies of galvanized bar and plain bar with calcium nitrite inhibitor exposed for 72 months in a marine environment at Key West, Florida [67]. Cylinders in 0.6 w/c concrete were cast with different reinforcement at covers of 12.5, 25, 37.5 and 50 mm. The samples were suspended in the inter-tidal zone.
Laboratory and Field Performance of Galvanized Steel in Concrete
177
Specimens with uncoated bars showed visible staining and cracking from 42 months to the end of the testing. Specimens containing galvanized bars showed no external staining or cracking to the end of the study. The galvanized bar with 50 mm cover had red rust formation over 0.5% of its surface and zinc oxide covered an additional 6% of the surface. There was no evidence that the zinc oxide resulted in cracking of the concrete and this was attributed to the fact that this corrosion product occupied about one-third less volume for a given weight and was loose and powdery. The average weight loss of zinc was 3% for bars with 50 mm cover. At the 25 mm level, the acid-soluble chloride was approximately 9.5 kg/m3, which confirmed that the threshold for corrosion of galvanized bars is higher than that of black bars. These tests clearly showed that the use of galvanized reinforcement extended the life of quite permeable 0.6 w/c concrete in a subtropical marine environment when compared with uncoated reinforcement. Sagues studied the performance of galvanized reinforcement with prior surface damage in a simulated marine substructure environment in the laboratory [68]. Concrete prisms were cast in which bars were embedded horizontally in pairs to form a top and bottom mat. The cover to the outside was 25 mm and the distance between the mats was 75 mm. Three types of reinforcement were used: *
*
*
Undamaged bars of either black steel or galvanized steel without coating damage. Damaged galvanized bar where the coating had been removed by filing so that 1, 10 and 50% of the underlying steel were exposed. A blasted black bar was used to simulate a galvanized bar that had lost 100% of its coating. One mechanically deformed bar bent to U shape but without any other coating damage.
All bars in each specimen were electrically connected so that separate electrical measurements could be made for each bar. The top face of the specimens was ponded with artificial seawater, which was filled once a month. After 2 weeks, the remaining solution was drained and the specimens were dried. Macrocell currents, open-circuit potentials and instant-off potentials were periodically measured. After 350 days of exposure, no specimens showed active corrosion and the damaged bar in each specimen consistently acted as a cathode. Even in the specimen containing four black bars, the average corrosion current was only 1.5 mA initially, reducing to zero with time. In specimens containing damaged bars, the current varied between 1 and 1.5 mA, while the undamaged bar and the deformed bar both showed no current. The open-circuit potentials of the damaged bars were between 2 250 and 2100 mV, whereas both the 100% damaged bar and the black bar had potentials around 2100 mV. For deformed bars, the values stabilized around 2290 mV. Although the damaged bars showed very small macrocell current at this stage, they effectively polarized the exposed steel surface
178
S.R. Yeomans
to a level 50 2 100 mV more negative than that of black steel. This polarizing effect may be enough to increase significantly the time for corrosion initiation compared with that of black bar. This data suggested that, after 300 days, all specimens were still in the initiation stage of corrosion. The damaged portion acted as cathode and a 100 mV potential shift was observed in the protective direction. This trend was maintained up to 50% damage although the shift was insignificant above 50% damage to the coating. Thangavel et al. studied the corrosion resistance of electro-galvanized steel embedded in chloride-contaminated concrete [69]. Black bars were plated to attain a coating thickness of about 95 mm. Two concrete types were used, a lean mix (0.68 w/c) and a rich mix (0.5 w/c), and sodium chloride was added to the mix in the range 0 – 3.5% by weight of cement. Exposure was both atmospheric and full immersion (potable water and seawater) for periods of 2–3 years. Initial potential measurements (vs SCE) for the galvanized bars in all conditions were in the range 2800 to 21000 mV, rising to about 2450 to 2 650 mV after various periods of exposure. The presence of chloride, either in the mix or the immersion conditions, and the lean mix proportions of the concrete, both had a significant accelerating effect on the onset of corrosion. Visual examination revealed that, under immersion conditions, both white and red rust were visible whereas, under atmospheric exposure, only white rust was present. In the most severe conditions, i.e. the lean mix with high chloride content in full immersion, rapid dissolution of the zinc coating and early corrosion of the base steel occurred. In natural weathering conditions, the rate of attack was much slower. The protective efficiency of galvanized steel embedded in concrete of variable w/c immersed in seawater was studied by Bautista and Gonzalez [70]. Bare and galvanized steel were embedded in concrete cubes of varying mix proportions and w/c ratios in the range 0.4 – 0.8. Exposure was in artificial seawater for 1 year. Under each set of experimental conditions, galvanizing and the use of high cement/sand ratio in the concrete both had a high protective effect against corrosion. The corrosion rate of the bare steel increased continuously during the 12 months exposure, while that for the galvanized bars decreased during the first few months then increased over the last few months. The rate of this change for the galvanizing was dependent on the composition of the concrete, but the corrosion current was always less than that for the bare steel. For example, in a 0.6 w/c concrete, typical corrosion currents for the bare steel were in the range 0.20– 10.5 mA/cm2 for potentials varying from 2350 to 2500 mV, and those for galvanizing were 0.25– 2.0 mA/cm2 for potentials over the range 2550 to 2750 mV. Chloride measurements at the depths of the bars (30–35 mm) showed that, in all samples except the 0.4 w/c, threshold levels for corrosion of black steel were exceeded. In chloride-contaminated concrete, the galvanized bars corroded at a slower rate than the bare steel, resulting in an increased time to
Laboratory and Field Performance of Galvanized Steel in Concrete
179
depassivation of the reinforcement. This effect was noted as an increased service life in media with low or moderate chloride concentrations. In higher chloride situations, however, the point was eventually reached when the level of chloride at the bars was such that their effect could not be countered by the inhibitory effect of the hydroxyl ions and, as a result, the corrosion rate increased. The authors summarized their work by noting that, while the critical chloride threshold of galvanized steel is higher than that of bare steel, once this is surpassed, the zinc coating is depleted. They also pointed out that the rate at which this occurs clearly depends on the time taken for the chlorides to reach their threshold level, and this is directly related to the cover to the bars and the quality of the concrete, specifically its cement/sand ratio, and the use of additives to reduce porosity. Fratesi et al. investigated the performance of galvanized steel embedded in cracked concrete immersed in seawater [71]. Concrete prisms of 0.5 w/c were reinforced with plates of bare, galvanized (80 mm coating) and galvanized and chromate passivated steel. Some galvanized plates were damaged prior to embedment. Specimens were cracked to the depth of the bar prior to exposure in natural seawater in Ancona Harbour, Italy. In cracked concrete, the corrosion potential of the bare steel showed it to become active after short periods. The potential of the galvanized plate (21000 to 21200 mV vs SCE) was not affected by the chromate treatment or mechanical damage of the coating and the rate of attack at the apex of the crack remained low with no evidence of penetration of corrosion in comparison to the situation for the bare steel in identical conditions. While the thickness of the zinc coating was progressively reduced at the crack apex, sometimes exposing the underlying steel, no iron corrosion products were observed. This supported the view that zinc cathodically protects adjacent areas of exposed steel. Where attack on the zinc was observed, a thick white compact deposit of calcium hydroxyzincate was present. In sound concrete, the potential shift was delayed and no sign of corrosion was observed after 1 year, although a gradual shift in the potential of the galvanized plate was observed. Thus, galvanizing was effective in preventing corrosion for at least 1 year in aggressive seawater immersion conditions. In tests where different plates were electrically coupled, zero or insignificant currents were measured when both plates were galvanized. The highest currents were measured in cracked specimens where bare steel plates as cathodes were coupled to galvanized steel as anodes; these initially reached values of 80 – 100 mA but decreased to ,10 mA after 1 year. This indicated the importance of not coupling galvanized steel to bare steel, even in sound concrete. In 1996, McDonald et al. reported the results of screening tests for a range of reinforcing materials that had been assessed for inclusion in a major experimental study, sponsored by the FHWA, of corrosion-resisting reinforcement in concrete
180
S.R. Yeomans
[72]. These tests and associated polarization resistance measurements were performed in two solutions: one a 3% salt solution of pH 7 to represent salt water that may be encountered by bars prior to embedment in concrete or at cracks in in situ concrete, and the second a 0.3N KOH þ 0.05N NaOH þ 3% NaCl solution of about pH 13 chosen to represent the pore solution of salt-contaminated concrete. Comment on this by Kinstler [73] makes the point that the likely pH of the second solution is about 13.54, which is well above the pH of saturated calcium hydroxide (about 12.2 –12.6) and also above the threshold pH of about 13.3 at which zinc will not polarize. This solution also lacks the calcium ions necessary for the formation of calcium hydroxyzincate. As such, use of this simulated solution may lead to unreliable results, particularly when such results are to be extrapolated to predictions of field behaviour. The outcome of this screening was that the galvanized bars were rated as having about the same corrosion rate as black bars with a time-to-corrosion of about 1 year. This was considered to be totally inconsistent with the field performance results for galvanized reinforcement. The final report on this research was published in December 1998 [74]. Twelve bar types, including galvanized steel, were tested in concrete slabs (0.47 w/c), some of which were pre-cracked. Two layers of reinforcement were used, the top mat containing either two straight or bent bars and the bottom mat four straight bars. The results for the galvanized bars indicated that, when used in both mats, the average corrosion was 38 times lower than that of equivalent black bars. When tested in pre-cracked concrete with a black cathode, the corrosion rate increased by about 41%. With bent galvanized bars, the corrosion rate increased by a factor of almost 1.8 times, indicating that bending after galvanizing may reduce their performance in corrosive media. The conclusion drawn from this was that coating should be done after fabrication. In uncracked concrete, the performance of galvanized bars with a galvanized cathode was relatively good, although the corrosion rate increased by 24 times when a black cathode was used. This pointed to the potentially damaging effect of mixing black and galvanized steel and to the need to avoid electrical contact between galvanized bars and other metals. Gowripalan and Mohamed investigated the effectiveness of high-performance concrete and galvanizing in reducing corrosion [75]. Plain steel bars and galvanized bars (100 mm coating) were embedded at depths of 20 and 40 mm in both normal-strength (30 and 40 MPa) and high-strength concretes (50 and 80 MPa) and exposed to wetting and drying in a 4% salt solution for 1 year. Comparative rapid chloride penetration tests were conducted to monitor ion penetration through the concrete and the results showed the high-performance concrete reduced chloride ion penetration significantly. Half-cell potential measurements on black steel bars at 20 mm cover indicated that, after about 100 days, the bars in both concrete mixes were actively corroding, although
Laboratory and Field Performance of Galvanized Steel in Concrete
181
somewhat earlier than this (about 20 days) in the normal-strength concrete. Both concretes showed cracking with black steel bar once the threshold potential was reached, which was as early as 3 weeks in normal-strength concretes. With galvanized steel, the gradual shift in the potential from about 21000 mV to about 2600 mV occurred over about 200 days in normal-strength concrete but did not rise above about 2 800 mV in the high-strength concretes after 350 days. The authors concluded from this work that, while the provision of high-performance concrete increases the time to chloride-induced corrosion, the provision of sufficient cover is still necessary in an aggressive environment. It was also noted that galvanizing delayed the onset of chloride-induced corrosion in normalstrength concrete and that the use of high-performance concrete with galvanizing substantially delayed the onset of corrosion. Yeomans undertook a detailed examination of the nature of corrosive attack on the galvanized coating in chloride-contaminated mortars and the migration of the zinc corrosion products into the cement matrix [76]. Galvanized bars (95 mm coating thickness) were embedded in cement mortars (0.45 and 0.6 w/c) and exposed to wetting and drying in 3.5 and 5% salt solutions for periods up to 158 days. Transverse sections were taken from the cylinders and examined in a scanning electron microscope. Contact with wet cement resulted in the initial loss of some 10 mm of the pure zinc eta layer from the surface of the coating. In non-chloride conditions, there was little further attack on the coating over the period of exposure. In chloride-contaminated conditions, continued dissolution of the eta layer progressively exposed the underlying zeta and delta alloy layers of the coating. This dissolution was not uniform and resulted in the formation of many holes and tunnels in the coating, especially where the eta layer intruded around the zeta alloy. Even though the coating was severely attacked by this stage, the remaining thickness of zinc on the surface was 60–70 mm and the underlying zeta and delta alloy layers were still largely intact. Photo micrographs showing this effect are given in Chapter 1. Similar effects are discussed by Andrade and Alonso in Chapter 5. X-ray spectral analysis identified the corrosion product as the mineral zincite (ZnO) with no evidence of the formation of other complex zinc oxychloride-type products. The zinc oxide was highly mobile and migrated away from the bar surface into the adjacent matrix, filling pore space and microcracks to a distance of about 0.5 mm and with an apparent densification of the matrix. The presence of this corrosion product did not cause cracking or other visible distress to the surrounding cement matrix, in contrast to the situation when black steel in concrete corrodes. Further, it was noted that this densification process increases the adhesion between the zinc coating and the cement matrix, thereby increasing bond and reducing load-induced slip compared with black steel. This effect is discussed in detail by Kayali in Chapter 8.
182
S.R. Yeomans
Belaı¨d et al. characterized the steel –cement and zinc–cement interfacial zone (ITZ) adjacent to galvanized and plain steel in contact with cement pastes using surface hardness tests, microscopic examination, XRD and water absorption tests [77]. After 3 days, the abrasion rate and ITZ thickness were systematically lower in the case of contact with zinc than with plain steel. Microstructural observations and XRD indicated that the hydration of the cement paste is delayed in contact with zinc, although the presence of the calcium hydroxyzincate reaction product with zinc provided a higher hardness than the layered portlandite (calcium hydroxide) product on the steel surface. The water-absorption tests showed that initially the absorption was higher for the zinc–cement ITZ (at about 19 h) but that after 28 days there was no difference. This also points to the delay in the hydration of cement due to contact with zinc and the increased porosity (perhaps also due to hydrogen evolution) of the ITZ at this time. The authors noted that the decrease in the difference in the absorption with time means that, in contact with zinc, cement hydration is not permanently inhibited and filling of the porous structure in the ITZ will continue. These same researchers also studied the influence of the chemical composition of cement on the interface between galvanized reinforcement and cement [78]. These tests indicated that the nature of the interface with galvanized steel is strongly dependent on the chemistry of the cement used, in particular its alkalinity, the C3A/C3S ratio and the quantity of gypsum present. It was also apparent that small variations in the composition of the original cement produce significant change in the interfacial zone. As previously noted, the hardness of the interface was attributed to the formation of calcium hydroxyzincate, which is known to crystallize randomly on the interface and increase the bond between cement and the zinc despite the early-age retardation of the hydration reaction. In concretes made with low C3A/C3S ratio cement, the bond of galvanized bars is less than that for black steel bars although with time it is expected that this difference will be overcome due to continued hydration of the cement. Further discussion on the effect of zinc on the bond with concrete is also covered in Chapter 8. In 1999, Sagu¨e´s reported on studies of the performance of galvanized reinforcement in marine substructure service as part of a continuing ILZROsponsored project in association with the Florida DOT [79]. This investigation used concrete with modern mix designs (including 8% silica fume replacement), such as for marine bridge applications, in field marine exposure tests of a group of 18 piles installed at Matanzas Inlet, Florida. Both plain steel bars and galvanized bars were used in separate piles. Preliminary results were reported in 1994 and testing was planned to continue for a number of years. In the FDOT-sponsored tests (also reported by Sagu¨e´s), apparent corrosion rates of the galvanized steel were very low (, 0.3 mm/year) with minimal
Laboratory and Field Performance of Galvanized Steel in Concrete
183
wastage of the zinc coating after 2 years exposure. This suggested that a significant portion of the pure zinc eta layer could remain in placed for a period of 50 years before the arrival of the chloride contamination front. The concrete was also below the pH 13.3 limit suggested for lack of stability of galvanized bars in concrete. After 6 years testing, reinforcement potentials and polarization resistance data indicated that both the plain steel and the galvanized bars had reached passive behaviour. Sampling of one column after 9 years also showed a greatly improved performance over plain steel bars. In the continuation tests, other issues were addressed. These included the effect of partial loss of the zinc coating, either due to dissolution or mechanical damage, on the performance of galvanized bar in chloride-contaminated concrete and the need for field performance of galvanized reinforcement placed in currently formulated concrete in marine substructures. In addition to the field specimens that were continuously surveyed, laboratory specimens consisting of 0.33 w/c concrete prisms with 20% fly-ash replacement were cast with pairs of reinforcing bars. Three configurations of galvanized bar were embedded: all undamaged, one damaged on top and one bent on top. The specimens were exposed to simulated seawater. The results from the laboratory testing, from 1993, indicated that: *
*
*
during initiation, small but measurable currents develop between the bars steel and the galvanized portions of the reinforcement surface; exposed steel surfaces behaved as cathodes, with potential shifts (in the protective direction) of about 100 mV for specimens with as much as 50% surface damage; and while all specimens were still in the initiation stage of corrosion, chloride profiling suggests that corrosion was likely to begin in the near future.
The field testing indicated that the reinforcement potentials for both plain and galvanized steel had reached and maintained passive behaviour in both types of concrete, one with 20% fly-ash replacement, the other with 20% fly ash and 8% silica fume after a few months of exposure and during the 6 years exposure to date. Although active corrosion was yet to be observed, this work, and the infrastructure associated with it, provided the basis for future long-term assessment of the comparative behaviour of plain steel and galvanized steel under conditions of actual concrete quality and real-life development of corrosion.
6.6. 2000 and Beyond During 2000, Kinstler compiled an extensive research update on galvanized reinforcing steel [73]. This work, which was a collection of experimental and anecdotal evidence, explored the efficacy of galvanized reinforcing steel in concrete
184
S.R. Yeomans
construction. It reviewed significant experiences with galvanized reinforcement up to that time and synthesized these into a new paradigm, based on models for the initiation and propagation of corrosion, which was shown to be consistent with demonstrated and field behaviour. Kinstler summarized the superior performance of galvanized bars compared with black bars on the basis of a higher chloride threshold for corrosion initiation, the reduced corrosion kinetics and a lowering of the stresses induced in concrete as a result of the mobility of the zinc corrosion products. Belaı¨d et al. [80] examined the nature of the corrosion products formed on galvanized steel in contact with concrete. Their work involved exposing concrete cylinders (0.7 w/c) with embedded black steel or galvanized steel (with 45, 50 or 100 mm coating thickness) to cyclic salt water wetting and drying over a period of 20– 60 months. Very high chloride levels accumulated at the interface between the bars and the reinforcement (in the range 0.4–0.7% by weight of concrete) resulting in severe corrosion of black steel although the attack on each of the galvanized coatings was somewhat less. By this, it was clearly demonstrated that galvanizing increases the corrosion resistance of reinforcement in chloride-contaminated concrete. XRD identified both zinc oxide and zinc hydroxychloride as products of corrosion, presented as a white layer covering the surface of the galvanized bars. In very long-term cyclic testing, longitudinal cracks were observed in the cover concrete and zinc hydroxide was also formed. For coatings thicker than 50 mm, although the coating was partly consumed, continued protection of the steel was provided. However, in coatings thinner than 50 mm, the zinc was completely dissolved and corrosion of the base steel commenced even before cracking of the cylinders. This result indicated the importance of ensuring a minimum coating thickness of zinc to provide adequate protection. These observations were also used to explain reports of different zinc corrosion compounds identified by other researchers. Particular mention was made here of variations in the exposure times of specimens tested and variations in the coating thickness of the types of galvanized products used. Of the various corrosion products identified, it was noted that the zinc hydroxychloride was the most damaging since, should it occur, it causes cracking of the cover concrete and limits the service life of the galvanized coating. In other work by Belaı¨d et al., the porous structure of the interfacial zone around galvanized and black steel bars was investigated [81]. Small prisms of rapid hardening cement paste (0.5 w/c) were cast against either a galvanized (80 mm coating) or black steel plate. The specimens were fog-room cured for 19 h or 28 days, after which they were stored until constant mass was reached. The pore-size distribution and water distribution in the interfacial zone were determined on companion samples. A mercury intrusion porosimeter was used for the pore-size determination with data from the core of the bulk cement paste
Laboratory and Field Performance of Galvanized Steel in Concrete
185
taken as reference. After 19 h hydration, the results indicated a modification of the pore-size distribution and an increase in the porosity of the transition zone between cement and galvanized steel compared with that of the cement steel interface. After 28 days, however, the porous structure of the interfacial zone around galvanized and black steel was not significantly different. This change was attributed to the formation of calcium hydroxyzincate, which is the favoured hydration product with galvanized steel. This compound, together with cement hydration products, was able to fill the pore space in the interfacial zone, thereby reducing the size of pores and the connectivity of capillary networks. In 2001, Saravanan et al. [82] reported on tests on galvanized bars in simulated concrete of two pH levels, 12.7 and 10.5, to which 0 and 0.5% chloride had been added. Electrochemical impedance spectroscopy measurements were undertaken on three bar types, black, galvanized with a coating thickness of 95 mm and chromate-treated galvanized bars. The results indicated that both galvanized and chromate-treated galvanized bars gave excellent corrosion resistance compared with black steel at both pH levels in the presence and absence of chlorides. The average corrosion rate for the black steel specimens was about 2.5 times higher than that of the galvanized bars. The corrosion rate for the chromate-passivated bars was slightly lower than the rate for the plain galvanized bars, indicating the additional beneficial effect of chromate treatment. Andrade et al. [83] reported on a series of tests that compared the bond of deformed bare steel bars with galvanized bars that had been exposed for up to 10 years to the chloride environment of natural seawater. The hot-dip galvanized steel used had a coating thickness of 100 mm. Concrete with a 0.66 w/c ratio, some having calcium chloride additions, was used and cover to the bars was 25 mm. As far as the bond results are concerned, it was found that, over time, galvanized reinforcement maintained bond strength with concrete much higher than that required by Standards and cements with high alkali contents or the presence of chlorides in the concrete did not have a significant influence on the bond. It was also found that the use of a hydrogen inhibitor, in this case potassium chromate added to the mixing water, was determined to be unnecessary to maintain bond. The other aspect of this work to be commented on was the observation that the galvanized reinforcement, after 10 years immersion in seawater, did not show any signs of corrosion, even at parts of the bars protruding from the concrete beams. In contrast, bare steel used on stirrups, which were electrically isolated from the main reinforcement, showed extensive corrosion. This indirect observation confirms the superior corrosion resistance of galvanized steel in this severely corrosion situation, both embedded in the concrete at depth and at points of exposure with minimal cover. A follow-up study of the long-term performance of galvanized steel in concrete bridge decks in Pennsylvania was again undertaken by Construction Technology
186
S.R. Yeomans
Laboratories [84]. Earlier surveys over the period 1974– 1999 (previously cited) had examined bridges in Florida, Iowa, Pennsylvania and Vermont. This survey was of the Athens and Tioga bridges in Pennsylvania, both galvanized reinforced, which had been previously examined by CTL in both 1981 and 1991. *
*
Athens Bridge (age 28 years). The average chloride level was 0.41% (by weight of cement), which is more than 2.5 times higher than the ACI threshold value for black steel (0.15%). In areas where the potential was low (about 2 700 mV), no signs of corrosion of the galvanized bars were evident. Metallographic examination revealed the majority of bars exceeded the minimum galvanized coating thickness required for new bars of about 84 mm. In some areas (noted above), the remaining coating on the bars was in the range 56–66 mm. Tioga Bridge (age 27 years). The average chloride level was 0.40% (by weight of cement), again more than 2.5 times higher than the ACI threshold value. There were no signs of corrosion on any of the galvanized reinforcement, even from high potential areas. The thickness of zinc remaining on the bars greatly exceeded the minimum specified thickness of 84 mm.
The overall results from this survey confirmed the satisfactory performance of galvanized reinforcement in these bridge decks after long-term exposure to both calcium chloride, used as an admixture at the time of casting, and roadway deicing salts. The report concluded that the “galvanized reinforcing bars generally showed satisfactory resistance to corrosion” and visual inspection revealed “no signs of corrosion on any of the steel reinforcement” except in one area of the Athens bridge. Further, “cracking, delamination and spalling or evidence of active corrosion was not generally observed”. A Swedish investigation of the effect of residual chromate levels in raw cement was undertaken by Vinka [85]. In 1983, a decision was taken that all cement for sale in Sweden and other Nordic countries must be chromate reduced (from Cr6þ to Cr3þ) in order to lessen the health hazards of personnel working with concrete. Since the natural levels of chromates in many cements are able to passivate zinc, even if only temporarily, the issue of concern was whether the reduced chromate levels (to a maximum of 2 mg water soluble Cr6þ per kg of dry cement) would affect the corrosion performance of galvanized reinforcement in concrete. A twopart experiment was conducted. In the first, test panels were exposed in concrete to determine whether the zinc surface would be passivated and over what period hydrogen evolution would occur from the surface while, in the second, the corrosion of bare and galvanized steel were compared in chloride-containing concrete. Both grey (107 mm coating) and bright (78 mm coating) galvanized bars were used. Two types of Swedish cement were used to produce concrete of 0.44 w/c with 35 mm cover to the bars. The results of the first experiment indicated that the galvanized bars were passivated in concrete made with both of
Laboratory and Field Performance of Galvanized Steel in Concrete
187
the chromate-reduced cements. Passivation was observed for both the grey and bright zinc coatings with slightly less coating loss for the bright coating. One important conclusion noted by the authors was that the “dangers that have been associated with hydrogen evolution on zinc coated steel in concrete seem to be clearly exaggerated”. The second series of tests showed that, on splitting of the blocks, the adhesion of the concrete to the galvanized bars was superior to that for the black steel. Concerning the extent of corrosion, the steel bars were more severely attacked than the galvanized bars. Galvanized bars performed well in concrete with a chloride content #1% by mass of cement and the threshold value seemed to be about 1.5%. A summary of several ILZRO-sponsored projects commenced in the mid-1990s to evaluate the suitability of galvanized reinforcement in different tropical environments and, in particular, the performance of locally produced bars was given by Ba¨bler et al. [86] and Srinivasan [87]. Long-term tests were carried out in the laboratory and at different tropical marine exposure sites in south-east India, the Yucatan Peninsula in eastern Mexico and at Matanzas Inlet in Florida (previously cited). Concretes of different qualities (strength, w/c, composition) were used with black bars, galvanized and galvanized plus chromate-treated bars for comparison. In all cases, including the laboratory tests, chloride was chosen as the aggressive component. At the time of this report, on-site tests had been underway for 5 years. With data from the various tests arrangements, the effects of concrete quality (w/c ratio), the influence of cover depth, the effect of different treatments of the reinforcement surface treatment and damage to the zinc coating were assessed. Despite the difference in both the exposure conditions and experimental set-ups, the projects all independently showed that, in tropical marine environments, galvanized reinforcement has a better corrosion resistance than plain steel in concrete. Galvanized coatings up to 250 mm thick were able to retard the corrosion process by more than 5 years (to date) once aggressive media (chlorides) reached the reinforcement. It was also clear that additional chromate treatment of the galvanized steel did not increase the corrosion resistance significantly. It was noted, however, that an increased risk of stress corrosion cracking might exist and so, for this reason and environmental concerns, it was suggested that chromate treatment should not be used for galvanized reinforcement. Further details of this comparative study are yet to be published. Wheat and her co-workers recently reported on an investigation of the corrosion behaviour of a number of reinforcing materials compared with that of black steel, including galvanized steel that had been coated before and after bending [88]. A modified version of the ASTM G109 macrocell test procedure was used. A low w/c concrete (0.4) was used and the specimens were ponded with a 3.5% NaCl solution on a 2-week exposure, 2-week dry cycle for a period of 3 years. Some specimens contained the corrosion-resistant steel in both the top and bottom mat,
188
S.R. Yeomans
while most had the corrosion-resistant steel in the top and black steel in the bottom. After some 30 months exposure, preliminary results show that, in concrete of this quality (low w/c and not deliberately cracked), only limited corrosion was occurring, with all samples showing corrosion currents ,10 mA. On examination, only the black steel specimen showed any likely corrosion and the galvanized bars were in good condition, although there was evidence of some corrosion of the zinc because of the presence of white corrosion product. Some anomalies were noted, however, in that the galvanized bars bent after coating performed better than bars bent before coating in both the salt-free calcium hydroxide solution test as well as the calcium hydroxide solution containing the threshold level of chloride. In the presence of 3.5% NaCl, however, the bars bent before coating performed better. Testing in this programme is ongoing Finally, in an extensive programme of coordinated research jointly sponsored by the United Nations Common Fund for Commodities and the International Lead Zinc Study Group under Program LZSG/03, a series of projects in India and Mexico have been undertaken to investigate the performance of galvanized and zinc-based alloy-coated bars in tropical marine conditions. These projects, undertaken by CINVESTAV at Merida, Mexico, the Indian Institute of Technology at Chennai, India and the Central Electrochemical Research Institute at Karaikudi, India studied the behaviour of black, galvanized and galvanized and chromate-treated bars in a range of environments, including atmospheric exposure and accelerated testing in both continuous salt fog and saltwater wetting and drying. These studies, conducted over a 3–5-year period, included various types of concretes (ordinary Portland cement and Pozzolanic cement), quality of concrete (low w/c of 0.4 –0.5 and high w/c of 0.6–0.7), grades of concrete (15 –45 MPa), some with chloride additions, and varying covers to the reinforcement (15–50 mm). A range of evaluation techniques was used, including potential-time studies, electrochemical impedance spectroscopy, gravimetric studies, time-to-cracking and surface and metallurgical examination. At the time of writing, the final reports of these investigations were being collated by ILZRO in preparation for publication and dissemination through seminars and other forums. The broad conclusions from these studies, as they specifically relate to the use of galvanized reinforcement, can be summarized as follows: *
*
*
in all testing conditions, galvanized and galvanized and chromate-treated bars showed better corrosion performance than black steel; in lean concrete mixes, the chromate treatment had a beneficial effect, even in cracked concrete; in cracked concrete, black bar corroded more with increased crack width but the attack was retarded by the presence of the galvanizing and corrosion of the substrate was prevented;
Laboratory and Field Performance of Galvanized Steel in Concrete
*
*
*
*
*
*
*
*
189
although both the galvanized and chromate-treated bars were attacked in highchloride conditions, the substrate was not attacked to any severe extent and a considerable amount of the coating remained on the surface at the conclusion of testing; the extent of attack on the coating in high-chloride conditions was reduced for the chromate-treated bars; in low-chloride conditions, attack on the coating was limited to the pure zinc outer-layer although in higher chloride conditions attack had progressed to the inner alloy layers; in atmospheric exposure tests simulating industrial and tropical marine conditions, the galvanizing was found to be in tact after 2 years testing; in salt-fog tests, corrosion of the galvanized coating initiated in 0.5 w/c OPC concrete at a chloride level 3.5 times higher than for black steel and in similar PZ concrete at a level 2.9 times higher; in the more severe wetting and drying tests, corrosion of the galvanized coating initiated in 0.5 w/c OPC concrete at a chloride level 1.8 times higher than for black steel; in the range of accelerated corrosion tests, the time to cracking of the cover concrete for galvanized bars was of the order of 4–12 times longer than for black steel bars; and in higher strength concretes, in particular, the presence of galvanizing substantially increased the time to cracking compared with black steel.
Overall, these investigations have clearly shown that galvanized reinforcement results in a several-fold increase in corrosion resistance compared with black steel and significantly delays the time-to-cracking of concrete. In particular, the chloride threshold for galvanized steel in concrete is of the order of 2–4 times higher than that for black steel in equivalent concrete and exposure conditions. Further, in superior quality concretes with increased depth of cover to the reinforcement, a further and significant increase in corrosion resistance is achieved. It is likely that further work will be undertaken arising from these studies.
6.7. Summary The research that has been reviewed in this chapter represents a sustained effort over some 40 years, and a significant investment in time and resources, to address a range of issues relating to improvements in the durability of reinforced concrete construction, specifically the use of galvanized reinforcement. This work, and the conclusions to be drawn from it, form the basis of an extensive body of knowledge allowing an improved understanding of the characteristics and performance of galvanized reinforcement in cementitious materials.
190
S.R. Yeomans
What is initially most striking is the enormous variety of research that has been undertaken in so many different countries, research that varies from the highly scientific to the very applied and practical. Some of the research is also clearly market-focused and as such falls into the category of product development, but this should not be dismissed because it also makes a valuable contribution. Overall, what this points to is a worldwide interest in galvanizing as a means of corrosion protection for reinforcement, and recognition of its relevance and presence as one of a number of corrosion-resisting reinforcing materials available for use in concrete construction. What is apparent is that the research that has been undertaken, especially since the mid-1980s, has generated a large body of laboratory data and field evidence that is much more concise and reliable in its methodology and that is repeatable, comparable and, to a very large extent, consistent in its outcomes and conclusions than was previously the case. As noted in the discussion above, it was largely the lack of consistency and reliability in the early published data that led to the FHWA notifications in the mid-1970s and again in 1982 that cast so much doubt on the efficacy of galvanizing and restrictions on its use at that time. The weight of evidence from this research has now provided sufficient data to substantiate the superior performance of galvanized reinforcement compared with black steel in concrete. Specifically, it is clear that galvanizing affords long-term protection to steel reinforcement that extends the time to the initiation of corrosion of the reinforcement compared with black steel and, as a result, significantly delays and perhaps even avoids cracking of the concrete in many types of structures for 30– 50 years and beyond. In this way, galvanizing may eliminate the need for the first cycle of repair and maintenance of traditionally reinforced concrete, which would more than recover the costs associated with its initial use. The extension of life through galvanizing is the result of a fortunate combination of circumstances, which act in concert to protect the concrete mass. From the moment of casting, zinc in contact with wet cement chemically reacts and is quickly passivated by the formation of adherent surface crystals of calcium hydroxyzincate. This film effectively isolates the zinc from the alkaline environment of concrete and stabilizes it in much the same way as steel in concrete is also passivated. During service, galvanizing is able to resist the natural effects of a reduction in pH of the cover concrete due to carbonation, which black steel is unable to do. In fact, the corrosion resistance of zinc improves as the pH drops below the threshold at which black steel would normally corrode. Further, it has also been widely demonstrated that zinc has a significantly higher threshold for chloride-induced corrosion than black steel, such that, in marine exposure or de-icing conditions, zinc will remain in the passive condition and thus not actively corrode until a much higher concentration of chloride ions accumulates at the depth of the reinforcement. Quite what the actual chloride threshold value for zinc
Laboratory and Field Performance of Galvanized Steel in Concrete
191
may be is not specifically known as it is dependent to some extent on the metallurgical nature of the zinc coating itself. Evidence to hand does strongly suggest, however, that the chloride threshold is several times higher than that for black steel; an amount of individual data indicates a value typically 2–4 times that of black steel but, in some studies, especially field surveys, chloride levels up to 10 times the threshold for black steel is reported as not causing attack on the galvanized coating. The research has also shown that, once corrosion of the galvanized coating does initiate, additional protection is afforded through the progressive, but quite slow, dissolution of the bulk of the alloy coating covering the steel. However, even when the coating is completely lost in local areas, the exposed regions of the steel substrate are cathodically protected via the sacrificial dissolution of the adjacent coating. The extent, or range, of this protection is dependent on the conductivity of the electrolyte but there is evidence to suggest that in normal concrete this protection extends to distances of perhaps 10 mm. Another important aspect of the behaviour of zinc in concrete that has been reported is that, during the dissolution of the coating, the zinc-rich corrosion products that typically develop are known to be friable and able to migrate away from the bar/matrix interface into the body of the concrete. This distance over which the migration occurs varies considerably but there is evidence to suggest it may be up to 0.5 mm and thus clearly beyond the interfacial zone. The significant advantage of this behaviour is that attack on the coating and the precipitation of zinc-rich corrosion products does not cause the same level of distress to the cover concrete as in the case for black steel. In this way, early cracking of the concrete can be avoided, as previously noted. It is also clear from the research that, in good (or superior) quality concrete with an adequate depth of cover to the bars, such as may be expected in general building and construction, the advantages to be gained by the use of galvanized reinforcement can be fully optimized. What is equally apparent, however, is that in poor-quality concrete exposed to very aggressive test conditions, galvanizing does not always provide the extension of life that is expected. This is all the more likely in poorly designed concrete specimens with low cover to the bars or exposed sections and in specimens that have been artificially pre-cracked or mechanically damaged. It is to be noted, of course, that, in such conditions, black steel reinforcement would itself quickly corrode and that distress to the cover concrete would almost immediately follow. To conclude, it is worth noting that there are many issues that must be addressed when designing and interpreting the results of accelerated testing programmes. Unfortunately, this is not well understood and is perhaps the most obvious reason why such disparate data is sometimes presented and contradictory conclusions drawn from research programmes that purport to address similar issues. In the design of experiments, then, researchers should be conscious of the need to be able
192
S.R. Yeomans
to compare data and outcomes from one test programme with that from another. In this context, there is considerable benefit to accrue from side-by-side experiments where different materials are compared in similar concretes and exposure conditions. Equally, laboratory studies that can be benchmarked to long-term natural exposure conditions are also of extreme value. Studies of galvanized reinforcement in concrete, as with any corrosion-resisting reinforcement, should therefore be undertaken with these outcomes clearly in mind. Only when experiments are designed and conducted such that correlation and cross-referencing to other published data and real-life performance is achievable, will it be possible to answer the many important questions that remain concerning the corrosion protection of steel reinforcement in concrete.
References [1] ILZRO. (1970). Galvanized reinforcement for concrete. Zinc Institute Inc and International Lead Zinc Research Organization, NY, USA, 120 p. [2] ILZRO. (1981). Galvanized reinforcement for concrete — II. International Lead Zinc Research Organization, NC, USA, 280 p. [3] Andrade, M. C., & Macais, A. (1988). Galvanized reinforcements in concrete. In: A. Wilson, J. Nicholson, & H. Prosser (Eds), Surface Coatings — 2 (pp. 137–182). Elsevier, London, Chapter 3. [4] Comite´ Euro-International du Beton, Coating protection for reinforcement: State of the art report (1992). CEB Bulletin d’Information No. 211, Chapters 2 and 5. Also published by Thomas Telford Services Ltd, 1995, 51 p. [5] ILZIC. (1995). Protection of reinforcement in concrete: an update. Indian Lead Zinc Information Centre, New Delhi, India. [6] Building Research Station Digest (UK). (1969). Zinc-coated reinforcement for concrete. Digest, 109. [7] Concrete Institute of Australia (1984). The use of galvanized reinforcement in concrete. Current Practice Note 17. [8] Galvanizers Association of Australia. (1999). After-fabrication hot dip galvanizing. GAA, Melbourne, VIC, Australia. [9] American Galvanizing Association. (2002). Hot-dip galvanizing for corrosion prevention: a guide to specifying and inspecting hot-dip galvanized reinforcing steel. AGA, Centennial, CO, USA, various editions. [10] Yeomans, S. R. (1987). Galvanized steel reinforcement in concrete. First National Structural Engineering Conference, Institution of Engineers, Australia, ACT, Australia, 662– 667. [11] Yeomans, S. R. (1993). Considerations of the characteristics and use of coated steel reinforcement in concrete, NISTIR 5211. National Institute of Standards of Technology, Gaithersburg, MD, USA. [12] Yeomans, S. R. (2001). Applications of galvanized rebar in reinforced concrete structures, NACE Corrosion 2001, Paper 01638. National Association of Corrosion Engineers, Houston, TX, USA.
Laboratory and Field Performance of Galvanized Steel in Concrete
193
[13] Halstead, P. E. (1957). Corrosion of metals in buildings. Chemistry and Industry, 1132– 1137. [14] Bresler, B. & Cornet, I. (1964). Galvanized steel reinforcement in concrete. Proc. 7th Congress, International Association of Bridge and Structural Engineers, Rio de Janeiro, August, 1964. [15] Bresler, B., & Cornet, I. (1969). Corrosion protection of steel in concrete and tentative recommendations for use of galvanized steel reinforcement in concrete, ILZRO Report TS-69-04. International Lead Zinc Research Organization, NC, USA. [16] Cornet, I., & Bresler, B. (1966). Corrosion of steel and galvanized steel in concrete. Materials Protection, 5, 4, 69–72. [17] Cornet, I., Bresler, B., & Ishikawa, T. (1968). Mechanism of steel corrosion in concrete structures. Materials Protection, 7, 3, 44–46. [18] Cornet, I., & Bresler, B. (1980). Critique of testing procedures related to measuring the performance of galvanized steel reinforcement in concrete. In: D. E. Tonini, & J. M. Gaidis (Eds), Corrosion of reinforcing steel in concrete (ASTM STP 713, pp. 160–195). American Society for Testing and Materials, Philadelphia, PA. [19] Frazier, K. S. (1965). Value of galvanized reinforcement in concrete structures. Materials Protection, 4, 5, 53–55. [20] Griffin, D. F. (1969). Effectiveness of zinc coating on reinforcing steel in concrete exposed to a marine environment, Technical Note N-1032. Naval Civil Engineering Laboratory, Port Hueneme, CA, USA. [21] Christensen, K. A., & Williamson, R. B. (1971). The galvanic cell problem in ferrocement, Report No UCESM 71-14. Structural Engineering Laboratory, University of California, Berkeley, 56 p. [22] Okamura, H., & Hisamatsu, Y. (1976). Effect of use of galvanized steel on the durability of reinforced concrete. Materials Performance, 15, 7, 43 –47. [23] Stark, D., & Perenchio, W. F. (1975). The performance of galvanized reinforcement in concrete bridge decks, ILZRO Project ZE-206. International Lead Zinc Research Organization, NC, USA. [24] Stark, D. (1978). Galvanized reinforcement in concrete containing chlorides, ILZRO Project ZE-247. International Lead Zinc Research Organization, NC, USA. [25] Gonzalez, J. A., & Andrade, C. (1979). A preliminary study of the behaviour of galvanized reinforcing bars in carbonated concrete. Revista Metallurgica, 15, 2, 83 –90. [26] Treadaway, K. W. J., Brown, B. L., & Cox, R. N. (1980). Durability of galvanized steel in concrete. In: D. E. Tonini, & J. M. Gaidis (Eds), Corrosion of reinforcing steel in concrete (ASTM STP 713, pp. 102–131). American Society for Testing and Materials. [27] Clear, K. C., Virmani, Y. P., Jones, W., & Jones, D. (1981). Time-to-corrosion of reinforcing steel in concrete: vol. 4. Galvanizing reinforcing steel, FHWA Report No. FHWA/RD-82/028. Federal Highway Administration. [28] Arnold, C. J. Galvanized steel reinforced concrete bridge decks: progress reports. FHWA Report No. FHWA/RD-78-R1033, Federal Highway Administration. [29] Cook, A. R. De-icing salts and the longevity of reinforced concrete, Corrosion ’80. Paper No 132. National Association of Corrosion Engineers, Houston, TX, USA. [30] Manning, D., Escalante, E., & Whiting, D. (1982). Galvanized rebar as a long-term protective system, FHWA Report No. DTFH61-82-P-300-30041-2/3. Federal Highway Administration.
194
S.R. Yeomans
[31] Hoke, J. H., Pickering, H. W., & Rosengarth, K. (1981). Cracking of reinforced concrete, ILZRO Project ZE-271. International Lead Zinc Research Organization, Research Triangle Park, NC, USA. [32] Gonzalez, J. A., Vazquez, A. J., & Andrade, C. (1982). Effect of humidity cycles on the corrosion of galvanized reinforcement embedded in carbonated and uncarbonated concrete. Mate´riaux et Constructions, 15, 88, 271– 278. [33] Gonzalez, J. A., & Andrade, C. (1982). Effect of carbonation, chlorides and relative ambient humility on the corrosion of galvanized reinforcement embedded in concrete. British Corrosion Journal, 17, 1, 21 –28. [34] Stark, D. Evaluation of the performance of galvanized reinforcement in concrete bridge decks. ILZRO Project ZE-320. International Lead Zinc Research Organization, NC, USA. [35] Andrade, A., Molina, Herete, A., & Gonzalez, J. A. (1983). Relation between the alkali content of cements and the corrosion rates of galvanized reinforcements. In: A. P. Crane (Ed.), Corrosion of reinforcement in concrete construction (pp. 343 – 356). Ellis Hardwood Limited, England. [36] Satake, J., Kamakura, M., Shirakawa, K., Mikami, N., & Swamy, N. (1983). Longterm corrosion resistance of epoxy-coated reinforcing bars. In: A. L. Crane (Ed.), Corrosion of reinforcement in concrete construction (pp. 357–377). Ellis Horwood Ltd, England. [37] Shimada, H., & Nishi, S. (1983). Seawater corrosion attack on concrete blocks embedding zinc galvanized steel bars. In: A. P. Crane (Ed.), Corrosion of reinforcement in concrete (pp. 407–418). Ellis Horwood, England. [38] Ru¨ckert, J., & Neubauer, F. (1983). Contact behaviour of galvanized and ungalvanized steel in concrete at higher temperature. Werkstoffe und Korrosion, 34, 295–299. [39] Sarja, A., Jokela, J., & Metso, J. (1984). Zinc-coated concrete reinforcement, Research Report 306. Technical Research Centre of Finland, Espoo, 92 p. [40] Gonzalez, J. A., Vazquez, A. J., Jauregui, G., & Andrade, C. (1984). Effect of four coating structures on corrosion kinetic of galvanized reinforcement in concrete. Materiaux et Constructions, 17, 102, 409–414. [41] Mikami, N., Arai, T., Yamazaki, A., & Tsukinuki, Y. (1985). Study of the corrosion resistance of rebars in concrete. Trans Japan Concrete Institute, 7, 181– 188. [42] Makita, M., Mori, Y., & Katawaki, K. (1980). Performance of typical protection methods for reinforced concrete in marine environments, Performance of Concrete in Marine environments, ACI SP65. American Concrete Institute. [43] Roper, H. (1985). Fatigue and corrosion fatigue failure surfaces of concrete reinforcement, Microstructural Science, 12. In: Proceedings 16th Annual Technical Meeting, International Metallographic Society, ASM. [44] Nu¨rnberger, U. (1986). Corrosion behaviour of galvanized steel in contact with building materials. Werkstoffe und Korrosion, 37, 6, 302–309. [45] Hildebrand, H., & Schwenk, W. (1986). Effect of galvanizing on the corrosion of steel in concrete immersed in NaCl solution. Werkstoffe und Korrosion, 37, 4, 163–169. [46] Maahn, E., & Sorensen, B. (1986). The influence of microstructure on the corrosion properties of hot-dip galvanized reinforcement in concrete. Corrosion, 42, 4, 187– 196. [47] Pfeifer, D. W., Landgren, R. J., & Zoob, A. (1997). Protective systems for new prestressed and substructure concrete, FHWA Report No. FHWD/RD-86/193.
Laboratory and Field Performance of Galvanized Steel in Concrete
195
[48] Swamy, R. N., Koyama, S., Arul, T., & Mikami, T. (1988). In: V. M. Malhotra (Ed.), Durability of steel reinforcement in marine environment (SP109, pp. 147–162). ACI, Detroit, MI, USA. [49] Kashino, N. (1988). Applicability of newly developed corrosion resistant reinforcement in reinforced concrete construction. Second Australia –Japan Coordinated Workshop on Durability of Reinforced Concrete. Concrete Institute of Australia, pp. 733–741. [50] Treadaway, K. W. J., & Davies, H. (1989). Performance of fusion bonded epoxy-coated steel reinforcement. The Structural Engineer, 67, 99 –108. [51] Turgeon, R. Evaluation of epoxy and galvanized reinforcing bars in Pennsylvania bridge decks, Corrosion ’87, Paper Number 140. [52] Malasheski, G. J. (1989). Bridge deck protection with epoxy and galvanized rebars — Pennsylvania’s experience. NACE NE Regional Meeting, Baltimore MA, September 1989. National Association of Corrosion Engineers, Houston, TX, USA. [53] Malasheski, G., Maurer, D., Mettott, D., & Arellano, J. (1988). Bridge deck protective systems, Report No. FHMIA PA-88. Commonwealth of Pennsylvania Department of Transportation, Harrisburg, PA, vol. 85-17. [54] Stark, D. (1981). Influence of design and materials on the corrosion resistance of steel in concrete, PCA Research and Development Bulletin RD098.01T. Portland Cement Association. [55] Nu¨rnberger, U., & Beul, W. (1991). Influence of galvanizing and PVC coating of reinforcing steel and of inhibitor on steel corrosion in cracked concrete. Werkstoffe und’Korrosion, 42, 10, 537–546. [56] Rasheeduzzafar, Dakhil, F. H., Bader, M. A., & Khan, M. M. (1992). Performance of corrosion resisting steel in chloride-bearing concrete. ACI Materials Journal, 89, 439–448. [57] Baker, E., Money, K. & Sanborn, C. (1977). Marine corrosion behaviour of bars and metallic-coated steel reinforcing rods in concrete. Chloride corrosion of steel in concrete. ASTM STP 629 (pp. 30–50). American Society for Testing and Materials, Philadelphia, PA. [58] Stejskal, B. G. (1992). Evaluation of the performance of galvanized steel reinforcement in concrete bridge decks, CTL Project 050324. Construction Technology Laboratories Inc, Skokie, IL. [59] Mu¨ller, H. H. (1993). Behaviour of galvanized rebars in concrete. In: S. Nagati, T. Nireki, & F. Tomosawa (Eds), Durability of building materials (pp. 147– 156). E&FN Spon, London. [60] Northwood, D. O., & Roy, S. K. (1994). Corrosion protection of reinforcing bars by galvanizing, Corrosion and Prevention ’94. Australasian Corrosion Association, Adelaide, 10 p. [61] Roy, S. K., Liam, K. C., & Northwood, D. O. (1993).Cement and Concrete Research, 23, 6, 1289 –1306. [62] Saiful Islam, Md., & Kaushik, S. K. (1994). A study of corrosion of reinforced concrete in a simulated tropical marine environment. In: R. N. Swamy (Ed.), Corrosion and corrosion protection of steel in concrete, vol. 2, (pp. 1267–1276). Sheffield Academic Press, UK. [63] Lucius Pitkin Inc (1994). Bridge deck galvanized rebar evaluation — Route 495 viaduct over US 1 & 9, New Jersey. Report No. M94402. [64] Yeomans, S. R. (1994). Performance of black, galvanized, and epoxy-coated reinforcing steels in chloride-contaminated concrete. Corrosion, 50, 1, 72– 80.
196
S.R. Yeomans
[65] Short, N. R., Zhou, S., & Dennis, J. K. (1994). Corrosion behaviour of zinc-alloy coated steel in hardened cement pastes. In: R. N. Swamy (Ed.), Corrosion and corrosion protection of steel in concrete, vol. 2, (pp. 1287 –1298). Sheffield Academic Press, UK. [66] Hime, W. G., & Machin, M. (1993). Performance variances of galvanized steel in mortar and concrete. Corrosion, 49, 10, 858–860. [67] Burke, D. F. (1994). Performance of epoxy-coated reinforcement, galvanized reinforcement and plain reinforcement with calcium nitrite in a marine environment. In: R. N. Swamy (Ed.), (Corrosion and corrosion protection of steel in concrete, vol. 2, pp. 1254 –1266). Sheffield Academic Press, UK. [68] Sagues, A. A. (1994). Performance of galvanized reinforcement in marine substructure service, ILZRO Project — ZE-418, Part 1. International Lead Zinc Research Organization, NC, USA. [69] Thangavel, K., Rengaswamy, N. S., & Balakrishnan, K. (1995). Corrosion resistance of galvanized steel in concrete. Materials Performance, 34, 9, 59–63. [70] Bautista, A., & Gonzalez, J. A. (1996). Analysis of the protective efficiency of galvanizing against corrosion of reinforcement embedded in chloride contaminated concrete. Cement and Concrete Research, 26, 2, 215– 224. [71] Fratesi, R., Moriconi, G., & Coppola, I. (1996). The influence of steel galvanization on rebars behaviour in concrete. In: C. L. Page, P. B. Bamforth, & J. W. Figg (Eds), Corrosion of reinforcement in concrete construction (pp. 630–641). Royal Society of Chemistry. [72] McDonald, D. B., Pfeifer, D. W. & Blake, G. T. (1996). The corrosion performance of inorganic, ceramic, and metallic-clad reinforcing bars and solid metallic reinforcing bars in accelerated screening tests. FHWA Report FHWA-RD-96-085. Federal Highways Administration, 104 p. [73] Kinstler, T. J. (2000). Research and update on galvanized reinforcing steel. Industrial Galvanizers America, Midlothian, VA, private communication. [74] McDonald, D. B., Pfeifer, D. W. & Sherman, M. R. (1998). Corrosion evaluation of epoxy-coated, metallic clad and solid metallic reinforcing bars in concrete. FHWA Report FHWA-RD-98-153. Federal Highways Administration, 127 p. [75] Gowripalan, H., & Mohamed, H. M. (1998). Chloride-ion induced corrosion of galvanized and ordinary steel reinforcement in high-performance concrete. Cement and Concrete Research, 28, 8, 1119–1131. [76] Yeomans, S. R. (1998). Corrosion of the zinc alloy coating in galvanized reinforced concrete. NACE Corrosion 98, Paper No 653. National Association of Corrosion Engineers, Houston TX, USA. [77] Belaı¨d, F., Arliguie, G., & Franc¸ois, R. (1998). Comparison of ITZ characteristics around galvanized and ordinary steel rebars. In: A. Katz, A. Bentur, M. Alexander, & G. Arliguie (Eds), The Interfacial transition zone in cementitious composites (pp. 198–203). E & FN Spon, London. [78] Belaı¨d, F., Arliguie, G., & Franc¸ois, R. (1999). Influence de la composition du ciment sur l’interface acier galvanise-beton. Mate´riaux et Techniques, 7– 8, 31–36. [79] Sagu¨e´s, A. A. (1999). Performance of galvanized rebars in marine substructure service. ILZRO Project ZE-418. International Lead Zinc Research Organization, NC, USA. [80] Belaı¨d, F., Arligue, G., & Franc¸ois, R. (2000). Corrosion products of galvanized rebars embedded in chloride containing concrete. Corrosion, 56, 9, 960 p.
Laboratory and Field Performance of Galvanized Steel in Concrete
197
[81] Belaı¨d, F., Arligue, G., & Franc¸ois, R. (2001). Porous structure of the ITZ around galvanized and ordinary steel reinforcements. Cement and Concrete Research, 31, 1561– 1566. [82] Saravanan, K., Muralidharan, S., Natesan, M., Venkatachari, G., & Srinivasan, S. (2001).Transactions of the Institute of Metal Finish, 79, 4, 146–150. [83] Andrade, C., Arteaga, A., Lo´pez-Hombra, & Va´squez, A. (2001). Tests of bond of galvanized rebar and concrete cured in seawater. Journal of Materials in Civil Engineering, 319–324. [84] Olson, C. A., & Nagi, M. A. (2002). Evaluation of the performance of galvanized steel in concrete bridge decks, ILZRO Project ZC-10. Construction Technology Laboratories, Inc, Skokie, IL, 39 p. [85] Vinka, T.-G. (2002). Corrosion of zinc coated steel in chromate reduced concrete with and without chlorides, Project SE3, COST 521 Workshop. IST-Luxembourg University of Applied Sciences, Luxembourg, pp. 59– 64. [86] Ba¨ßler, R., Lamers, D., & Alhassan, S. (2003). Suitability of galvanized rebars in tropical marine environment, Paper 03250, Corrosion. National Association of Corrosion Engineers, Houston, TX, USA. [87] Srinivasan, S. (2003).Transactions of the Indian Institute of Metals, 56, 3, 315–317. [88] Wheat, H. G., Fowler, D. W., & Jirsa, J. O. (2003). Challenges in evaluating different reinforcing materials, Paper 03297, Corrosion. National Association of Corrosion Engineers, Houston, TX, USA.
This Page Intentionally Left Blank
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 7
The Bermuda Experience: Leading the Way on Galvanized Reinforcement Neil D. Allan Bath University, UK
7.1. Introduction This chapter discusses the efficacy and utility of galvanized reinforcement in a variety of concrete structures on the islands of Bermuda. For over 50 years, galvanized reinforcement has been used extensively in everyday construction. Bermuda has an aggressive marine environment and experience has shown that galvanized reinforcement is a cost-effective solution to the concrete-corrosion problem. In preparing this paper, the author has drawn on extensive experience gained working on the design and construction of civil and maritime projects throughout the world, including 8 years from 1987 as a senior engineer with the Ministry of Works and Engineering (MW&E) in Bermuda. Reference is made to the unique geographical and climatic conditions of Bermuda and the problems this creates for reinforced concrete structures. Also presented are the results of specific investigations and current research. Using practical examples and illustrations, the author portrays an engineer’s viewpoint of the wider application of galvanized reinforcement.
7.2. The Islands of Bermuda Bermuda is a group of small oceanic islands that lie in the northwest Atlantic about 1000 km east of North Carolina. An aerial view of the islands is shown in Fig. 1. The islands, of which there are about 138 in number, are volcanic in origin with a coral limestone cap. The mainland, as it is called, is some 39 km long and 2 km wide at its widest point and comprises seven of the largest islands linked by bridges. Longbird Bridge and Watford Bridge are two of the most important of these bridge crossings and will be referred to later.
200
N.D. Allan
Figure 1: Aerial view of the islands of Bermuda. (See colour plate 6).
The climate is sub-tropical and frost-free. The maximum air temperature in the summer is 328C, dropping to about 108C in the winter months of January and February. The average relative humidity is 80%. Nowhere in Bermuda is more than about 1 km from the sea and so salt spray and salt-laden air is a constant challenge. In short, it is quite evident that Bermuda has an aggressive marine environment and that the prevention of steel corrosion, including steel reinforcement, has important economic implications for the island. The Bermuda MW&E is responsible for the design, procurement and maintenance of the island’s infrastructure and effectively acts as functional authority for all engineering issues. Over the years, the Ministry has built up a considerable body of knowledge and experience in issues relating to corrosion and corrosion control and the design and construction of durable, low-maintenance structures. Any visitor arriving in Bermuda for the first time is struck by the outstanding natural beauty of the islands but, for a civil engineer arriving to take up a senior post with the Ministry, something else rather unique about the islands became apparent. As a general observation, and considering the severity of the environment, the civil infrastructure was in surprisingly good condition and the docks, wharves and harbors were in particularly good condition compared with similar facilities visited in Africa, the Caribbean, Europe and North America. Another feature was that most of the reinforcement steel used on the island was hot-dip galvanized.
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
201
It was not initially obvious that all these observations were inextricably linked and this chapter attempts to make that connection clear and, more explicitly, to answer two simple questions: (1) Why does the MW&E in Bermuda resolutely continue to specify exclusively hot-dip galvanized reinforcement for all of its construction works? (2) Is the practice that has been adopted for over 50 years still applicable in the light of experience, recent research and changing construction methods? While the insight presented here into the behavior of galvanized reinforcement gained from Bermuda is specific to sub-tropical island conditions, it is the author’s view that there are no commonsense reasons why the practices developed and the lessons learnt in Bermuda should not have a wider application.
7.3. Previous Investigations 7.3.1. Sponsored Research — 1970s During the mid-1970s, several ILZRO-sponsored studies were undertaken by David Stark, from the Construction Technology Laboratories of the Portland Cement Association, on the performance of galvanized reinforcement in bridge decks and marine structures in Bermuda [1,2]. Four structures built between 1953 and 1968, each built by different contractors and for different organizations, were examined. These were: * * * *
a dock wall in Hamilton Harbor; a jetty at the Royal Bermuda Yacht Club; a dock wall at Pennon’s Wharf, St George’s; and an approach span of Longbird Bridge near the airport.
In these investigations, the size of bar and depth of concrete cover and the chloride content at the bar depth were determined. Metallographic examination of sections of the galvanized bars extracted from concrete cores was also undertaken. A summary of the results is given in Table 1 [1]. Stark concluded that, despite exceedingly high average levels of chloride ions (1.7–3.6 kg/m3 of concrete) present in the concrete surrounding the reinforcing bars, all were performing well in service with no signs of distress to the concrete. Stark also reported that, in all the structures investigated, which ranged in age from 10 to 23 years, the maximum deterioration of the zinc layers was only 8% as a percentage of the original coating thickness. The actual mechanism for calculation of the zinc loss used by Stark was not clear and recent work by Yeomans [3] indicated that the zinc-corrosion process is non-uniform. Therefore, using random core samples may not enable reliable
202 N.D. Allan
Table 1: Summary of Stark’s findings (CTL, 1975). Location of structure
Hamilton Harbor RB Yacht club Penno’s Wharf Longbird bridge
Date of build
1966 1968 1964 1953
Bar type and diameter (mm)
Deformed, 25 Round, 10 Deformed, 16 Deformed, 20
Concrete cover to bars (mm)
Age at time of inspection (years)
Chloride ion concentration (kg/m3) of concrete
Zinc loss as % of original thickness
155 –70 60 55– 70 50 –115
10 8 12 23
1.7 – 2.0 1.7 – 2.0 1.7 – 3.6 1.7 – 4.3
5 Nil 8 2–3
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
203
predictions to be made of the future performance of galvanized reinforcement in concrete structures. Nonetheless, Stark’s rigorous study clearly showed that, even under the most aggressive marine conditions, galvanized reinforcement appeared to be contributing significantly to the longevity of the concrete structures.
7.3.2. Bermuda MW&E Surveys — 1991 To supplement Stark’s earlier work and to provide an understanding of the longterm performance of galvanized reinforcement in practice, a visual survey of the four structures mentioned above was again undertaken in 1991, some 15 years after Stark’s initial investigation. The inspections were carried out in the inter-tidal zone, splash-zone and below water. A brief summary is provided below.
7.3.2.1. Hamilton Harbor — Dock Wall The wall extends about 1.5 m above high-tide level and was constructed in 1966 using 25 mm diameter, twisted galvanized bars with a concrete cover ranging from 55 to 70 mm. At the time of Stark’s research, 10 years after construction, the corrosion to the reinforcement was less than 5% of the original zinc coating. This was in concrete with measured chloride concentrations between 1.7 and 2.0 kg/m3 of concrete that would be potentially corrosive to untreated steel. Almost 25 years after construction, the visual survey in 1991 reported no sign of distress to the concrete due to any corrosion of the reinforcement (Fig. 2).
Figure 2: Dock #1, Hamilton Harbor.
204
N.D. Allan
7.3.2.2. The Royal Bermuda Yacht Club The jetty was built in 1968 using 10 mm diameter galvanized bars at a depth of 60 mm. The dock was designed and supervised by a local consulting engineering firm. At the time of Stark’s measurements, there was essentially no indication of corrosion to the bars after 8 years of exposure, again with high chloride levels of 1.7 –2.0 kg/m3. The 1991 survey, 23 years after construction, confirmed that the concrete was in very good condition with no signs of any distress caused by corrosion (see Fig. 3). 7.3.2.3. Penno’s Wharf — Vertical Concrete Dock Wall This structure is essentially a capping beam to sheet piles and extends 1.5 m above high-tide level. The wall had been built in stages but the oldest section was constructed in 1964 using 16 mm galvanized deformed bars with 55–70 mm concrete cover. Stark found a maximum 8% loss of the original zinc coating from the bars with chloride-ion concentrations peaking at 3.6 kg/m3. The visual survey in 1991 noted that, despite considerable physical abrasion caused by fender chains, the concrete was generally in good condition (Fig. 4). Where the physical damage had exposed the reinforcement, there was only minor rust staining and no signs of distress or spalling of the concrete, despite the fact that the exposed bars had been exposed to an aggressive marine environment for a considerable period. This property of galvanized reinforcement, i.e. to provide protection to exposed bars due to the presence of the zinc coating, provides an unintentional but valuable maintenance feature, particularly in high-visibility areas where rust staining and spalling of concrete is highly undesirable.
Figure 3: Jetty at the Royal Bermuda Yacht Club.
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
205
Figure 4: Penno’s Wharf, St George’s.
7.3.2.4. Longbird Bridge Built in 1953 by the US Navy, the bridge consists of a swinging asymmetrical steel span and a single concrete approach span. Although full design details were not available, it is thought that 20 mm diameter deformed bars with a concrete cover ranging from 50 to 115 mm were used. Stark identified that the bars in the support beams were in a highly corrosive environment with chloride levels up to 4.3 kg/m3. There was also a low chloride ion gradient across the concrete cores, suggesting that significant chlorides may have been present at the time of placement, as well as evidence of poor compaction of the concrete, particularly around the bars. Despite these onerous conditions, the zinc coating had been only
206
N.D. Allan
slightly affected by corrosion, with just 2 –3% loss of the original coating thickness and no evidence of concrete deterioration. Considering this, it was therefore quite puzzling to find in the annual bridge inspection records evidence of cracks to the soffit of the approach span beams in 1984, only 8 years after Stark’s tests. These cracks appeared to be the result of reinforcement corrosion but, significantly, no rust staining or corrosion products were present. At the time of the 1991 inspection, repairs to two of the beams were underway by the US Navy as part of a $US2M refurbishment work of the main steel bridge structure and swing mechanism. Interestingly, the bars exposed as part of the repairs to the concrete were still covered with an intact zinc coating. There were signs of minor corrosion in isolated areas along the bars and this was thought to be caused by salt introduced by sea-spray on the bars or salt in the concrete introduced during construction.
7.4. Longbird Bridge Revisited — 1995 In 1995, after 42 years of service, another investigation was instigated by the Ministry to determine the anticipated useful life of the Longbird Bridge concrete approach structure. This was part of an overall assessment of the condition of the bridge, due in part to the planned withdrawal of the US Navy from Bermuda and the subsequent transfer of property back to the Bermuda Government. It was hoped that tests on the concrete and reinforcement might shed some further light on the reasons for the cracking of the approach spans previously noted. A detailed investigation was thus initiated to determine the long-term performance of the galvanized reinforcement and the properties of the concrete used in the construction of the bridge. Two concrete cores were extracted, each containing a portion of galvanized reinforcement, from non-critical, low-stress zones well away from the repaired concrete. The results of the petrographic analysis of the concrete, chloride analysis and metallurgical examination of the reinforcement are detailed in Appendix A. In essence, the tests demonstrated that there was no evidence of corrosion of the reinforcing steel beneath the galvanized coating, nor any cracks that might indicate corrosion of the steel. This was despite very high levels of chlorides in the concrete adjacent to the bars, which were well above that generally considered necessary to induce corrosion of uncoated steel. The investigation confirmed that, with the exception of the minor superficial repairs to the concrete soffits in 1991, the 42-year-old concrete approach span bridge was in good condition and should continue to provide excellent service for many years. This did still leave the question, however, of why the concrete approach beams had cracked between 1976 and 1984, although with no further signs of distress at other locations.
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
207
One plausible explanation is that the cracks were the result of overloading and that they were structural and not corrosion-related. Around the same period that the cracks must have occurred, there were several reported cases of serious overloading of Longbird Bridge by abnormal vehicles, such as cranes and heavy equipment used in the construction of the nearby civil air terminal. This is well documented in the Ministry’s files. In addition, overloading by mixer trucks and container vehicles was a growing problem on all the bridges around the island at about this time. Longbird Bridge was particularly vulnerable to this casual overloading because it was one of the heaviest used bridges with the lowest safe axle-load rating. It was not until the early 1980s that the Ministry’s structural engineers properly regulated the heavy vehicle users on the island. This hypothesis does also help to explain why there was no evidence of the rust staining so commonly associated with advance reinforcement corrosion. Therefore, apart from the curious cracks that developed on Longbird Bridge in the 1980s, the initial investigations into the long-term durability of hot-dip galvanized reinforcement had been very positive and had been confirmed by later investigations.
7.5. Hamilton — Old Bus Depot Additional evidence of the longevity of galvanized reinforcement became available because of the demolition of the old bus garage in Hamilton. Built in 1943 using hot-dipped galvanized reinforcement, the Old Bus Depot was leveled in 1989 to make way for a new road. However, due to planning delays, the concrete rubble was still on site in early 1991 and inspection of the bars strewn around the site at that time clearly demonstrated the benefits of the galvanizing of reinforcing steel. Both the concrete and the reinforcing bars were in good condition after 45 years of service, as shown in Fig. 5, despite having been exposed for 18 months and subjected to severe physical damage during demolition. In addition, no evidence of bond or zinc/concrete interface problems was observed, but this matter is covered in more detail in a later section.
7.6. Performance of Black Steel in Bermuda Since there has not been a long history of the use of uncoated (black) steel on the island, it was quite difficult to find suitable examples to investigate. Also, because problem-free concrete structures tend to go unnoticed, there is, of course, a natural bias toward failures. Despite these reservations, the following examples are graphic reminders of the potential for reinforcement corrosion in Bermuda.
208
N.D. Allan
Figure 5: Cold twisted galvanized reinforcement at the Old Bus Depot demolition site. Note that steel mesh reinforcement around the steel column had not been galvanized.
7.6.1. Sargasso Fish Processing Facility The building, erected in 1943, was situated on the US Navy annex to the south west of the island where many of the reinforced concrete buildings had been built with plain black bars. Fig. 6 shows the all-too-familiar reminders of what can go wrong with a reinforced concrete structure should corrosion of the steel occur. The severe corrosion is indicative of chloride attack, possibly introduced via aggregates or mixing water. These buildings were condemned in 1991 as unsafe, unfit for habitation and beyond economic repair, and they were subsequently demolished.
7.6.2. Ordinance Island Bridge Ordinance Island Bridge was designed by a Florida-based company without the benefit of galvanized reinforcement. The bridge was the responsibility of the town of St George, not the Ministry, but this does not fully explain why local practices and methods of construction were not adopted. The bridge consisted of pre-stressed concrete beams spanning between tubular steel piles, with an in situ concrete deck.
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
209
Figure 6: External reinforced columns without galvanized protection showing advanced state of corrosion.
Neither the tendons, stirrups nor the main deck reinforcement was galvanized although the foundations and pile caps, designed separately by local consultants, did utilize galvanized reinforcement. The bridge was only 10 years old when it was considered beyond economic repair. It had suffered massive spalling and severe corrosion to both the reinforcement and pre-stressing tendons although the pile caps were in excellent condition. In 1991, the bridge beams and deck were replaced with pre-cast concrete beams using galvanized reinforcement throughout.
7.6.3. Commentary Of course, in all the examples quoted so far, there are other issues involved, such as the appropriateness of the design, the quality of the concrete, the level of site supervision and workmanship, the severity of the exposure conditions and the client’s brief. However, the underlying intuitive conclusion drawn from the examples cited is that, in practice, galvanizing the reinforcement does appear to help arrest corrosion for many years, even in very aggressive marine exposure conditions such as in Bermuda.
7.7. The Watford Bridge — A Case Study This section deals with another bridge structure that suffered badly in Bermuda’s climate — the Old Watford Bridge. By describing the design of the replacement
210
N.D. Allan
bridge, this case study should provide a clear way forward for designers utilizing galvanizing both for the concrete reinforcement and for superstructure steelwork. Since 1903, there have in fact been three Watford Bridges, the last two being of particular relevance. The 1957 bridge is a classic example of what can happen if local corrosion conditions and construction practices are not fully understood, which in this case led to inadequate protect of the reinforcement. In contrast, the designers of the new Watford Bridge, built in 1982, were meticulous about corrosion protection and developed a quite revolutionary design adopting galvanized steel beams and reinforcement to create a low maintenance, 120year design life structure. Watford Bridge forms the sole road link between Somerset and the Watford Islands, leading to the old Naval Dockyard and commercial port on Ireland Island. The link is required for commercial, industrial, residential and tourist traffic.
7.7.1. The Old Watford Bridge A fixed road link of some 140 m has been in existence between the islands since 1903, when a steel truss bridge was built by the Admiralty. The bridge was replaced in 1957 by a post-tensioned concrete structure. To be fair to the designers, this type of bridge design has been plagued by tendon-corrosion problems and was, until recently, banned by the Highways Agency in the UK. Only 3 years after construction, rust spots appeared on the flanges of the concrete beams and there then began an ongoing program of chipping, repairing and sealing in an attempt to protect the embedded steel reinforcement. Despite the remedial work, deterioration of the superstructure continued and became progressively more severe. In 1977, an in-depth investigation was carried out by the Cement and Concrete Association to determine the cause and extent of the corrosion. A full survey of the structure was completed and 25 core samples of concrete were extracted from the webs of selected beams for subsequent examination and analysis. The samples were examined to determine the depth of penetration of carbonation, the general quality of the concrete and its ability to provide protection to the reinforcement. Further tests were carried out to determine the chloride concentration at various depths from the weathered faces to discover where, if anywhere, threshold chloride levels may have been exceeded. Jacobson and Churchman [4] reported carbonation penetration to be consistently of the order of 10– 15 mm. Chloride contents were greater near the weathered faces than in the heart of the samples and, except in two cases, there was a marked drop in chloride level beyond the limit of carbonation penetration. In the remaining two cases, there were isolated zones of higher concentrations beyond the limit of carbonation. The degree of carbonation of the concrete indicated that,
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
211
at the locations of the samples, the concrete was of good quality, well compacted and capable of providing adequate protection to the reinforcement. The relative concentrations of chloride indicated that no undue quantities of chloride had been included in the concrete at the time of mixing; rather, the chlorides present had penetrated the hardened concrete from the outside. Because of the survey, it was possible to identify two types of corrosion. The first consisted of widely distributed areas of corrosion centered on the secondary reinforcement. This was concentrated mainly on the bottom flanges of the beams where it showed in the form of horizontal cracking, principally on the sides of the flanges and, to a lesser extent, on the underside of the flanges. The second type of corrosion was seen to be concentrated in the area of the pre-stressing tendons, causing general destruction of the ducts and, in places, the tendons. This type of corrosion was consistent with their having been penetration of the pre-stressing ducts by sea water, causing internal corrosion of both the duct sheathing and the tendons. It is believed that salt from the atmosphere and intermittent washing from the sea penetrated the movement joints between spans and seeped into the anchorage pockets and thence down the pre-stressing ducts that were not completely sealed by grout. This was determined upon demolition [4]. The survey confirmed that there was sufficient evidence of widespread and severe corrosion, as shown in Fig. 7, to indicate that the bridge would eventually become unserviceable. It was also considered unlikely that a system of repair and maintenance could be devised that would extend the life of the bridge for any reasonable time. The bridge was demolished in 1982 after only 25 years of service.
Figure 7: Severe corrosion in beams of the Old Watford Bridge constructed in 1957.
212
N.D. Allan
7.7.2. The New Watford Bridge The lessons learnt from the rapid deterioration of the 1957 bridge were carefully analyzed before a brief was compiled for a replacement bridge. Construction of the New Watford Bridge (see Fig. 8) commenced in 1979 after much deliberation about alternatives and corrosion protection. The (then) Public Works Department called for a 120-year life span with no maintenance for up to 10 years and only minor maintenance up to 20 years. The final design put forward used galvanized structural steel beams supporting a galvanized reinforced concrete deck. Considerable attention was given to detailing to minimize maintenance and reduce corrosion throughout the design and construction of the new bridge. The use of galvanized reinforcement was endorsed by the consultants despite at that time still being quite controversial elsewhere in the world. The consultants relied heavily on their own research and the fact that Bermuda had been successfully using galvanized reinforcement for over 40 years. Some of the design provisions regarding the reinforced concrete deck included: *
* *
*
all reinforcement hot-dip galvanized to BS729 with a minimum coating thickness of 140 mm; all tie wire also hot-dipped galvanized; any damaged surfaces of the galvanizing touched up with two coats of zinc-rich paint; a minimum concrete cover to all reinforcement of 60 mm; and
Figure 8: The New Watford Bridge. (See colour plate 7).
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
*
213
a specified 28-day concrete strength of 37.5 MPa with minimum OPC cement content of 400 kg/m3.
For the bridge structure itself, the main structural steel plate girders were hotdip galvanized and painted with a four-coat paint system. All nuts, bolts and washers were also galvanized, spun and painted. Throughout the structure, joint details which may trap condensation were avoided; for example, bolted connections between steel beams were kept to a minimum and substantial dripinducing grooves were located beneath edge beams and movement joints. Further, additional concrete casing to the piles was provided within the splash zone, circular columns were used for all piers to eliminate chipping of corners in the splash zone and beams were supported on laminated rubber bearings that would not deteriorate. In 1992, over 10 years after construction, the concrete and steelwork were in excellent condition. Visual inspection revealed no sign of corrosion or deterioration of the reinforced concrete, except for some hairline cracking on the edge beams over the abutments, which was probably due to shrinkage restraint from the abutment wing walls. To ensure that there was no insidious deterioration of the structure, an investigation was initiated to determine the level of chloride penetration and carbonation to the concrete and the degree of corrosion, if any, to the galvanized reinforcement steel. Eight cores were taken from the structure at different locations with different exposure conditions. The results of the analysis of the cores carried out in 1992 are summarized in Table 2. A petrographic examination of Core #1 identified that the concrete had a moderately low water/cement ratio in the range 0.40–0.45 and was well consolidated. The top of the core was slightly carbonated and the bottom had the indentation of a piece of reinforcement (secondary reinforcement was included in five of the eight cores). Around the indentation, there appeared to be some air entrainment. There were no indications of corrosion products on the concrete surface of the imprint. The chemical analysis showed a high chloride level at the surface that decreased rapidly into the core, as would be expected. The chloride levels were all below the 1% threshold limit. The metallurgical examination revealed the galvanized coating to be in very good condition and there was no evidence of corrosion of the reinforcement, even in the most aggressive splash zone. The chemical and metallurgical analysis of the cores confirmed observations from the visual inspections of the bridge. Both the galvanized reinforced concrete deck and the galvanized plate girders were performing extremely well and the bridge was on track to meet and surpass the design criteria for a 120-year life. The choice of galvanized reinforcement for this bridge was not accidental or
214
Table 2: New Watford Bridge — summary of test results. Orientation Exposure zone
1
Inclined to vertical
Spray zone
0.05
0.01
0.01
193
Horizontal surface Vertical surface Horizontal surface Horizontal surface Inclined to vertical
Tidal zone
0.61
0.25
0.04
153
Splash zone
0.5
0.21
0.04
N/A
Tidal zone
0.68
0.11
0.08
N/A
Splash zone
0.58
0.21
0.08
120
Spray zone
0.07
0.02
0.09
128
Spray zone
0.07
0.08
0.01
N/A
Spray zone
0.02
0.07
0.04
138
2 3 4 5 6
7 8
Edge beam, south side of west abutment Foot of west abutment Wall of west abutment Foot of west abutment Wall of west abutment Edge beam, north side of west abutment Fishing embrasure, north side of deck South wing wall
Horizontal surface Vertical Surface
Chloride ion Chloride ion Chloride ion Thickness of content at content at content at galvanized surface (%) 30 mm depth (%) 50 mm depth (%) coating remaining (mm)
Notes: (1) Specified galvanized layer thickness of 140 mm. (2) Water soluble chloride ion concentrations are by weight of concrete.
N.D. Allan
Core Location of # element
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
215
merely the result of precedents. The decision was not taken lightly by the Public Works Department, who were very much working in the public spotlight as a result of the early demolition of the previous bridge, or by the consultants who had their international reputation at stake.
7.8. Economic Case for Galvanized Reinforcement In 1997, the average price of galvanized reinforcement ex-factory was approximately 50% more expensive than plain uncoated bar. This appears a high premium but, once shipping, transport, handling, tax, profit and labor are included, the additional cost of galvanized bars in situ is typically less than 10% above that of plain black steel. This premium equates to about 0.25–1.0% of the total capital cost of a typical building project, which is fairly insignificant compared with the whole-life cost savings that can be expected from the use of galvanized reinforcement. These values will obviously vary depending on market conditions, local rates and the type of construction. What is important is to view the initial modest additional expenditure on galvanized reinforcement, whatever that may be, as an investment with the future savings on maintenance costs providing the dividends. Typical maintenance costs for a variety of structures are given in Table 3 (as published by the ICE, UK). The life-cycle costs of long-duration products such as a bridges and buildings are notoriously difficult to ascertain, let alone the contribution of individual components such as the reinforcement. However, the payback period for an initial additional investment in galvanized reinforcement of less than 1% of total costs will typically be very short indeed when viewed against the values given in Table 3. Further discussion of the economics of galvanizing has been given by Yeomans in Chapter 1.
Table 3: Typical maintenance cost of structures (ICE, UK). Type of structure
Administration buildings Wharves and dolphins Reinforced concrete bridges
Annual maintenance costs (% of initial capital cost) 0.2 –0.4 0.3 –0.5 0.3 –0.4
216
N.D. Allan
7.9. Practical Considerations In the following sections, some observations are given on practical issues concerning the choice, processing and use of reinforcing steel for building and construction. This is based on the author’s experience gained in Bermuda and elsewhere over many years.
7.9.1. Alternate Coatings Epoxy-coated reinforcement has been used in Bermuda by private companies, although usually only in office buildings or similar projects. It has been tested by the MW&E in a number of trials over the years but has never been accepted as an equivalent substitute to galvanizing. The MW&E continues to specify hot-dip galvanized reinforcement in favor of other protection systems, namely epoxycoated, primarily because of its superior resistance to damage during shipping, storage and site handling.
7.9.2. Special Practices Experience has shown that reinforcing steel destined for Bermuda can sit on a dock for many weeks before shipment and then spend up to 3 weeks on board ship. After unloading at the dock, where again it may remain for some time, it is then transported to a storage yard before delivery to site. On average, a bar will have been handled at least eight times from leaving the galvanizer to arrival on site; it is then usually handled several more times before it is in its final position. While a small and isolated location such as Bermuda may increase the requirement for handling, it would not be unexpected that other mainland countries experience similar high handling rates; certainly, congested city sites would exhibit similar problems to an island. The fact that the galvanized coating is more robust and much less likely to be damaged or scratched during shipment and placement is the single most important and compelling reason for its continued preference in Bermuda. Epoxy-coated bars, on the other hand, require special handling during transportation, storage and treatment on site and, from experience, it is known that this level of care is not generally forthcoming from stevedores or site operatives. On site, epoxy bars are often chipped or scratched and showing signs of corrosion prior to being embedded.
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
217
7.9.3. Damage to Coating Galvanized bars may occasionally chip or scratch but the areas of damage are usually small and do not normally require repair because of the self-healing and protective, sacrificial properties of galvanizing. In Bermuda, repairs are required to damaged areas greater than 6 mm using two coats of a zinc-rich paint; this was seldom required, however, unless the bars had been cut or bent on site. MW&E specifications require all reinforcement be bent before galvanizing. This did not appear to cause local contractors any problems or delays since most reinforcing bars were bent off-site as a matter of routine and so the extra planning required (when galvanizing) was minimal. Of course, it was recognized that some site bending would be inevitable and these cases were treated on an individual basis and usually only required the application of zincrich paint as mentioned above. Galvanizing compares favorably with epoxy coatings when it comes to site bending. For example, Fig. 9 shows a galvanized bar and an epoxy-coated bar that were removed from a local construction site in Bermuda, both having been bent on site. The galvanized bars had light crazing and flaking on the outer zinc layers but, underneath, the zinc– iron alloys were sound and obviously still adequately protecting the steel. The zinc and the accompanying alloys are, by the nature of the galvanizing process, an integral part of the bar. The epoxy bar, on the other hand, had considerable rusting at several locations at the bend and along the adjacent ribs. It may be argued that much of the damage to the epoxy coating was caused by the bending machinery, not the bending itself, but that is essentially irrelevant as the damage to the coating still occurred.
Figure 9: Comparison of epoxy coated (upper) and galvanized bars (lower) after site bending.
218
N.D. Allan
7.9.4. Coating Mass Another important advantage of hot-dip galvanized reinforcement is its natural inbuilt quality control. During the dipping process, a minimum amount of zinc deposition is virtually guaranteed. Despite a range of quality-control tests and coating-thickness measurements that can and are used, it is certainly comforting for the engineer on site to know that, if galvanized bar looks right, it probably is.
7.9.5. Bond Strength This issue seems to have caused great concern for some engineers contemplating using galvanized reinforcement. Initial investigations and guidance in Bermuda relied heavily on the work done by the British Building Research Station [5]. Their research concluded that galvanized reinforcement bars demonstrated an increase in bond strength compared with ordinary bars, an observation that has since been endorsed by other research [6,7]. The reason for this increase in bond stress is believed to be caused by the presence of calcium hydroxyzincate crystals, which act as a bridge between the zinc coating and the concrete, thereby increasing the adhesion and the formation of non-expansive zinc-corrosion products and densification of the adjacent matrix, as discussed by Yeomans [3]. Epoxy-coated reinforcement, on the other hand [8,9], shows reductions in bond stress of up to 20% and, when Cairns [10] analyzed bond strength in relation to cracking in the concrete, the reduction was closer to 50%. This puts epoxy-coated deformed bars in the same category, for the purposes of bond and lap lengths, as plain round bars. Cairns goes on to report that this reduction in bond stress may double the design crack width and increase beam deflections at serviceability conditions by 20%. In summary, the bond strength of galvanized reinforcement is at least as effective as normal, deformed black steel bars and offers real advantages over epoxy-coated bars that require design modifications if used. These issues are discussed in more detail by Kayali in Chapter 8.
7.10. Continued Use of Galvanized Reinforcement in Bermuda It may be perceived from this paper that the use of galvanized reinforcement in Bermuda is historical and small scale, but this would be far from the truth. In the 8 years to 1997, the Bermuda Government embarked on an ambitious capital construction program costing in the order of US$300M. Galvanized reinforcement was specified for all of these projects.
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
219
Figure 10: Tynes Bay Waste-to-Energy plant foundations with 100% galvanized reinforcement. (See colour plate 8).
Figure 11: Tynes Bay construction showing a heavily reinforced ground beam. (See colour plate 9).
220
N.D. Allan
The two largest construction projects ever undertaken on the island, the Tynes Bay Waste-to-Energy Facility valued at US$70.5M and the new Secondary School at US$50M þ , together accounted for over 3000 t of galvanized reinforcement. Fig. 10 shows the scale of the foundations of the Tynes Bay site where all reinforcement was hot-dip galvanized, while Fig. 11 shows the detail of a heavily galvanized reinforced ground beam. This is quite a testimonial to the continued confidence that local and international consultants, contractors and clients place in the use of galvanized reinforcement. Specialist applications, such as the recently constructed and state-ofthe-art Shark Pool commissioned by the Bermuda Aquarium and Zoological Society, further demonstrates the versatility of galvanized reinforcement. In addition, there has been a decline in the use of epoxy-coated reinforcement in preference to galvanized steel in the private sector, mainly for the reasons stated above. Certainly, no sensible client or designer would entertain specifying an uncoated bar in Bermuda, a situation so very different to that encountered in the UK or USA.
7.11. Conclusions Civil engineers are, by nature and training, analytical, logical and cautious. They usually need to have considerable confidence in any new product or technique before it is fully accepted. Galvanized reinforcement is slowly beginning to gain their confidence in the UK and the USA. Galvanizing as a process has been around for over 100 years and is well proven to delay significantly the onset of steel corrosion. Despite this, it is quite mystifying why designers who would happily specify galvanized handrails appear to baulk at the thought of using galvanized reinforcement. The Bermuda experience confirms the reliability and effectiveness of using galvanized reinforcement in a range of structures. Engineers can place their confidence in a product that has been successfully used in Bermuda for over 50 years and whose performance has been scientifically evaluated and found to be sound.
References [1] Stark, D., & Perenchio, W. (1975). The performance of galvanized reinforcement in concrete bridge decks, ILZRO project no. ZE-206. International Lead Zinc Research Organisation, NC, USA. [2] Stark, D. (1978). Galvanized reinforcement in concrete containing chlorides, ILZRO project no. ZE-247. International Lead Zinc Research Organisation, NC, USA. [3] Yeomans, S. R. (1998). Corrosion of the zinc alloy coating in galvanized reinforced concrete, Corrosion 1998. NACE International, Houston, TX, Paper No. 653.
The Bermuda Experience: Leading the Way on Galvanized Reinforcement
221
[4] Jacobson, J. K., Churchman, A. E. (1984). Demolition of a pre-cast post-tensioned bridge, Bermuda. Concrete Society Conference on Demolition of Special Structures, UK. [5] Building Research Station. (1969). Zinc coated reinforcement for concrete, digest no. 109. HMSO, London. [6] Fratesi, R., Moriconin, G., & Coppola, L. (1996). The influence of steel galvanization on rebars behaviour in concrete. In: C. L. Page, P. B. Bamforth, & J. W. Figg (Eds), Corrosion of reinforcement in concrete construction, The Royal Society of Chemistry, Special Publication 183. [7] Kayali, O., & Yeomans, S. R. (2000). Bond of ribbed galvanized reinforcing steel in concrete. Journal of Cement and Concrete Composites, 22, 6, 459–467. [8] Mathey, R. G., & Clifton, J. R. (1976). Bond of coated bars in concrete, Structural Engineering Division. American Society of Civil Engineers, 102, ST1, 215–229. [9] Kobayashi, K., & Takewaka, K. (1984). Experimental studies on epoxy coated reinforcing steel for concrete protection. International Journal of Cement Composites and Lightweight Concrete, 2, 99 –116. [10] Cairns, J. (1996). Performance of epoxy-coated reinforcement at the serviceability limit state. Proceedings Institution of civil engineers, structures and buildings, 104, 61–73.
Appendix A
Evaluation of Galvanized Reinforcing Steel in the Longbird Bridge, Bermuda David Stark Construction Technology Laboratories (CTL), Inc., Illinois, USA
April 1995 Reproduced with permission from David Stark and CTL. Editorial note: Equivalent SI values included for comparison.
A.1. Introduction The present investigation was intended to evaluate the performance of the galvanized reinforcing steel in the Longbird Bridge, after 42 years of service. For this purpose, two concrete cores were extracted and forwarded to CTL. These cores were nominally 5 12 in: (140 mm) in diameter and 12 in. (300 mm) and 6 in. (150 mm) long. Core No. 1 was taken horizontally into the bridge deck from an outer exposed formed surface, while Core No. 2 was taken vertically through a curbsidewalk component of the structure. Core procurement was done in the fall of 1994.
A.2. Scope of Work Three lines of evaluation were requested to characterise the performance of the galvanized reinforcing steel in the concrete, as follows: 1. One core, No. 1, was subjected to petrographic examination in accordance with procedures in ASTM C856, “Standard Practice for Petrographic Examination of Hardened Concrete.” Both finely lapped and freshly fractured surfaces were examined to characterise the quality of the concrete and identify any other features such as abnormal microcracking and secondary reaction products. Also, depth of carbonation was determined by phenolphthalein applications to freshly fractured surfaces. 2. Determine the total (acid soluble) chloride contents of the concrete at selected depths in the two cores. For this purpose, dry powder samples were obtained
Evaluation of Galvanized Reinforcing Steel in the Longbird Bridge, Bermuda
223
using a drill and 14 in: (6.4 mm) diameter bit, and saving the sample obtained from inside a 14 in: wide outer ring of concrete in the cores. The holes were drilled into the cylindrical faces of the cores to avoid contamination from previous wet coring. Chloride contents were determined in accordance with procedures in ASTM C 1152, “Standard Test Method for Acid-soluble Chloride in Mortar and Concrete.” Non-evaporable water contents also were determined for each sample and used to correct for differences in paste – aggregate ratios among the samples. 3. Metallographic analysis was done to determine the thickness and compositions of the galvanized coatings on the embedded reinforcing steel. This was done on one section of steel in each concrete core.
A.3. Results of the Investigation The following sections describe the findings of this investigation.
A.3.1. Petrographic Examination Results of the petrographic examination of Core No. 1 are reported below. The examination was conducted on both finely lapped and freshly fractured surfaces of the core. The coarse aggregate is a crushed dense to porous limestone consisting virtually entirely of calcite. The colour ranges from light to orange-buff. Particle shapes are angular to subangular and blocky to elongate. Maximum particle size is 1 in. (25 mm). The fine aggregate appears to be of the same type of limestone but more consistently of a dense, fine grained texture. Both coarse and fine particles are uniformly distributed through the concrete. The concrete is well consolidated with tight, intimate bond between aggregate particles and the hydrated cement paste matrix. The matrix contains numerous voids, generally in the size range characteristic of intentionally entrained air. These voids are uniformly distributed through the matrix. A few, somewhat larger, voids are scattered throughout the matrix and are considered entrapped air voids. The air content of the concrete was estimated at 2.5–3.5%. The microscopic examination revealed no abnormal microcracking through the full length of the core, including that due to drying shrinkage or processes that cause progressive deterioration of the concrete. Treatment of freshly fractured surfaces with phenolphthalein indicates neither detectable carbonation of the cement paste matrix in the interior of the concrete, nor particularly at the exposed outer surface of the concrete. This surface carries a light grey to white painted
224
D. Stark
coating, beneath which is a dark, dense surface zone in the concrete where the phenolphthalein caused the appearance of a bronze colour on fractured surfaces. This surface zone extends to a maximum depth of 14 in: (6.4 mm) and may represent the application of a surface coating or sealant. It was not present in, nor typical of, concrete deeper in the core. The core also contained a section of galvanized reinforcing bar, 12 in: (12.7 mm) in diameter and located 2 78 in: (73 mm) below the coated external surface of the core. There was no evidence of corrosion of the steel substrate either on the surface of the steel bar or in the cast of the bar in the concrete. Treatment of the cast with phenolphthalein also revealed no evidence of carbonation. Most of the cast displayed sharply defined features of the embedded steel. However, an 34 in: (19 mm) long section of part of the cast displayed a frothy texture that may represent either localised inadequate consolidation of fresh concrete along one side of the steel, or reaction coating with the highly alkaline solutions in the fresh concrete.
A.3.2. Chloride Contents Results of the measurements for total chloride contents, corrected for variations in paste-aggregate ratios among the samples, are given in Table 1. Because there were no cement content determinations or mix design data available, results are expressed as mass per unit volume of concrete, wherein the unit mass of the concrete was determined in the 1976 investigation [1,2] to be 3765 lb/yd3 Table 1: Chloride concentrations in longbird bridge concrete. Core No.
Core location
Depth in inches (mm)
Chloride lb/yd3 (kg/m3)
1
Taken horizontally into side of bridge deck
0.0 –0.25 (0 – 6.4) 1.50 –1.75 (38 –45) 2.75 –3.0 (70 – 76)a 6.0 –6.25 (152 – 159)
5.12 (3.05) 3.95 (2.35) 3.24 (2.10) 3.05 (1.81)
2
Taken vertically into sidewalk
0.0 –0.25 (0 – 6.4) 1.25 –1.50 (32 – 38) 2.75 –3.00 (70 – 76)a
8.73 (5.18) 10.39 (6.18) 8.81 (5.23)
All chloride values are corrected for differences in paste– aggregate ratios among the individual samples. Mass of chloride ion is based on a unit weight of the concrete of 3765 lb/yd3 (2240 kg/m3), determined in the previous investigations. a Depth of reinforcing steel.
Evaluation of Galvanized Reinforcing Steel in the Longbird Bridge, Bermuda
225
(2240 kg/m3). Also, all values were corrected for differences in past-aggregate ratios among the seven samples. The threshold acid-soluble chloride concentration above which corrosion is likely on untreated steel is 0.20% by mass of cement. For a concrete containing 500 lb of cement per cubic yard (297 kg/m3), which is estimated to be close to that in the Longbird Bridge concrete, the threshold chloride level would be 1.0 lb/yd3 (0.6 kg/m3). From Table 1, it is seen that chloride levels in the samples from Longbird Bridge are well above the threshold level for untreated steel. In Core No. 1, the chloride concentration at the level of the steel was 3.24 lb/yd3 (1.9 kg/m3). In Core No. 2, the concentration was 8.81 lb/yd3 (5.2 kg/m3) at the level of the steel. These concentrations were well above the threshold level, and therefore provide an environment favourable to corrosion of untreated reinforcing steel. However, it should be recalled that oxygen also must be present to sustain active corrosion. Whether this condition is met at these sample locations is not known.
A.3.3. Metallurgical Analysis Metallurgical analyses were performed by the Zinc Corporation of America (ZCA) on a section of reinforcing steel in each of the two cores used by CTL for petrographic and chloride analyses. A summary of findings for each of the five types of analyses is given below. 1. Elemental mapping at paste– steel interface. This work, using the scanning electron microscope, mapped the diffusion of zinc from the galvanized coating 15– 20 mils (380– 500 mm) into the surrounding cement paste. This diffusion is reported to help prevent a build-up of internal pressure that might lead to spalling of the concrete. 2. Metallographic examination of galvanized coating. One examination of steel in Core No. 1 revealed a 7 – 10 mil (175 –250 mm) thick coating consisting of a blocky delta layer of zinc– iron alloy adjacent to the steel, overlain by a columnar zeta alloy which, in turn, is covered by a layer of pure zinc. Another area revealed a mild attack on the coating which, at the time of examination, was 5.5 – 6.8 mils (140 – 173 mm) thick and displayed an irregular surface profile. The coating on the steel in Core No. 2 has undergone locally more severe corrosion which has exposed the steel surface at isolated locations. Coating thickness ranged from 0 to 2.0 mils (0–50 mm) at certain locations, and up to 1.3–10 mils (33– 250 mm) at other locations. 3. Average coating thickness on bar circumference. Optical microscopic determinations revealed average coating thickness of 7.1 (180 mm) and 4.9 mils (124 mm) on the reinforcing bars in Core Nos. 1 and 2 respectively.
226
D. Stark
This was based on 12 equally spaced measurements along the full peripheries of the steel, with the greatest and smallest thickness measurements being discarded. 4. Semi-quantitative analysis of corrosion product. Scanning electron microscope-elemental dispersive X-ray analysis revealed that, after normalising to 100% and expressing as oxides, 55% ZnO was present in corrosion products, together with 31% CaO. This confirms the other observations of localised corrosion of the galvanized coating. 5. X-ray diffraction analysis. This examination revealed the presence of zinc oxide and calcium zinc hydroxide hydrate and major and minor corrosion products. Traces or possible traces of zinc hydroxide and iron oxide also were identified.
A.4. Summary and Conclusions This project covers an investigation of the corrosion resistance of galvanized reinforcing steel in Longbird Bridge in Bermuda, built in approximately 1952. Petrographic examination of concrete cores, determination of acid-soluble chloride contents, and various metallurgical analyses of the galvanized coating were included in the investigation. The petrographic examination revealed the concrete to be in good condition with no evidence of progressive degradation. There was neither evidence of corrosion of the reinforcing steel substrate beneath the galvanized coating nor cracks associated with steel that might reflect corrosion of the steel. Acid soluble chloride contents of the concrete near the reinforcing steel in the cores sampled indicate concentrations far greater than the threshold level generally considered necessary to induce corrosion of untreated steel. In these samples, the acid-soluble chloride concentrations were calculated as 3.24 and 8.81 lb/yd3 of concrete (1.93 and 5.24 kg/m3), which is well above the 1.0 lb (0.6 kg/m3) threshold level for active corrosion of untreated steel. It should be noted that, in these cases, moisture and chloride conditions may have been so uniform, or oxygen may not have been sufficiently available, that local differences in electrical potential could sustain active corrosion cells. Metallurgical analyses indicated that minor localised corrosion of the galvanized coating has occurred on the reinforcing steel in Core No. 2, where acid soluble contents were calculated as 8.81 lb/yd3 of concrete (5.24 kg/m3). Nevertheless, at this location, there was no distress observed associated with the steel. Overall, the galvanized steel coatings on reinforcing steel in the samples of concrete from the Longbird Bridge have provided excellent service for more than
Evaluation of Galvanized Reinforcing Steel in the Longbird Bridge, Bermuda
227
40 years without signs of impending or existing distress associated with steel, despite high levels of chloride ion adjacent to the steel.
References [1] Stark, D., and Perenchio, W. (1975). “The Performance of Galvanized Reinforcement in concrete Bridge Docks,” Project No. ZE-206, International Lead Zinc Research Organisation, Inc., and American Hot Dip Galvaniser Association, Inc., October. [2] Stark, D. (1978). “Galvanized Reinforcement in Concrete Containing Chlorides” Project ZE-247, International Lead Zinc Research Organisation, Inc., and American Hot Dip Galvanisers Association, Inc., April.
This Page Intentionally Left Blank
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 8
Bond of Steel in Concrete and the Effect of Galvanizing Obada Kayali University of New South Wales, Canberra, Australia
8.1. The Significance of Bond Bond between concrete and reinforcement is essential for developing the full capacity of the reinforcement and, as such, is the most important property contributing to the successful functioning of a reinforced concrete system. The distribution, magnitude and nature of stresses which develop in the surrounding concrete are complex and are affected by many factors, among which are the concrete strength, the bar diameter, the absence or presence of bar-surface deformations (ribs), the geometry of the ribs, the presence or absence of confining reinforcement, the cover to reinforcement and the position of the bars in the member. The main sources of bond strength depend upon whether the bars are ribbed. If the bars are smooth, the main contributions to bond strength come from the chemical adhesion and the friction resistance occurring between the bar and the concrete. If, however, the bars are ribbed, two additional sources of bond strength become more significant. These are the bearing capacity of the concrete between the lugs and the shear strength of the concrete cylindrical surface located between the lugs, as represented in Fig. 1. Bond stresses are presented in the following three common situations: * * *
anchorage; changing moment along the beam; and constant moment along certain beam sections.
The first situation can be clearly represented by a cantilever beam example. The cantilever shown in Fig. 2 is reinforced with a bar of diameter db ; anchored at a distance L into the support. If the stress in the bar at the section a –a was fst ; the force
230
O. Kayali
Figure 1: Stresses contributing to the bond stress in a deformed bar [2] (with permission of John Wiley & Sons, Inc.).
Figure 2: Bond stresses in anchorage situation.
Bond of Steel in Concrete and the Effect of Galvanizing
231
applied at the bar is Ast fst ¼ ðpdb2 =4Þ £ fst
ð1Þ
This force must be resisted by equal and opposite force in the direction of BA. This is provided by the action of the concrete surrounding the bar and applied in nonuniform distribution starting from zero at the end A to a maximum at the section B. The action exerted by the concrete on the surfaces of the bar may be given the description of bond action and the stresses resulting from it are the bond stresses. If bond stress is denoted by sb ; and the average bond stress between sections A and B is denoted sbav ; then sbav pdb L ¼ ðpdb2 =4Þ £ fst
ð2Þ
sbav ¼ ðdb =4Þð fst =LÞ
ð3Þ
from which From this, it can be concluded that, if there is an ultimate value of bond stress for the particular concrete and reinforcing bar which can be sustained, such value may be called the bond strength existing between the particular bar and concrete and may be denoted by sbu : It also follows that the length of anchorage AB needed, such that there will be no failure in bond, will be equal to Ld ¼ db fst =ð4sbu Þ
ð4Þ
where Ld is called the development length needed for that situation. The second situation is represented in Fig. 3 where there is a change in the moment between sections a –a and b –b, which are separated by the differential
Figure 3: Forces and moments in a differential length of a beam.
232
O. Kayali
distance dx: Since the moment in the differential distance dx has changed from M to M þ dM; it follows that the tension applied on the bar changes from T at section a – a to T þ dT at section b –b. If now the bar was isolated as a free body, it can be seen that, for equilibrium in the horizontal direction, dT must be opposed by action acting on the bar in the direction opposite to that of dT: This action must be supplied by the bond between the reinforcement and the surrounding concrete. Thus dT ¼ sb pd dx
ð5Þ
z dT ¼ dM
ð6Þ
but
where z is the moment arm at the section. Hence, dT ¼ dM=z
ð7Þ
dM=z ¼ sb pd dx
ð8Þ
sb ¼ ðdM=zÞ=ðpd dxÞ; or
ð9Þ
sb ¼ ðdM=dxÞ=ðpd zÞ
ð10Þ
and therefore
and so
However, dM=dx is the value of the shear force at the section in consideration and therefore sb ¼ V=ðpd zÞ
ð11Þ
It can thus be seen that the bond stress developed between the bar and the surrounding concrete is a function of the shear force acting at the section in consideration where the shear force is the rate of change of the moment at that section. It may be concluded also from here that, in sections of beams where a high shear force value is expected, a large bond stress value can also be expected to occur. At this instance, one is tempted to consider the situation of constant moment to be one that is free of bond stresses. This, however, is not the case, as has been demonstrated by Goto [1]. Figs. 4 and 5 show the approach taken by Goto where equal tensile forces were applied to both ends of a ribbed steel bar encased in concrete. Using dye injection to follow the development of the cracks, Goto was able to show the crack trajectories emanating from the ribs and their behaviour between the major cracks in the concrete where the tensile strength of concrete had been exceeded. The trajectories of these small cracks were in the direction of the bearing stresses created by the ribs and perpendicular to the direction of
Bond of Steel in Concrete and the Effect of Galvanizing
233
Figure 4: Crack development in concrete in bond with steel at constant tension [1] (with permission of the American Concrete Institute).
the principal tensile stresses, which eventually caused cracking when the tensile strength was exceeded. It can thus be seen that, at the position of the large major crack, the whole tension is carried by the bar only while, just beyond the major crack, tensile stresses start to develop in the concrete. As a consequence, there will be a change in the tension in the bar from a maximum at the major crack to a minimum somewhere between two major cracks and to a maximum again at the next major crack.
Figure 5: Deformation and cracking of the concrete surrounding tensioned steel [1] (with permission of the American Concrete Institute).
234
O. Kayali
It is now recalled that the rate of change in the tension experienced by the bar creates bond stresses dT ¼ sb pd dx
ð5Þ
sb ¼ dT=ðpd dxÞ
ð12Þ
sb ¼ ðdT=dxÞ=ðpdÞ
ð13Þ
and or This situation can be represented in Fig. 6 [2]. It can thus be concluded that the factors contributing to bond are the adhesion between the concrete and the bar surfaces, the friction between the surfaces and the concrete, the mechanical interlock provided mainly by the ribs, the shear strength of the concrete — which plays a significant role in the possibility of bond failure, and the tensile strength of the concrete — which governs the formation and spacing of cracks.
Figure 6: Changes occurring in moments, tension in the concrete, tension in the steel and stiffness in the vicinity of flexural cracks [2] (with permission of John Wiley & Sons, Inc.).
Bond of Steel in Concrete and the Effect of Galvanizing
235
8.2. Failure in Bond Bond strength has only been indirectly defined. ASTM [3] describes a standard test for comparing concretes on the basis of the bond developed with reinforcing bars in which the average bond stress is calculated at any stage of a pullout test until: * * *
the yield point of the reinforcing bar has been reached; the enclosing concrete splits; or a slippage of at least 2.5 mm has occurred at the loaded end.
The bond stress calculated by this method is equal to the load on the bar, as recorded at any stage in the test divided by the nominal surface area of the entire embedded length of the bar. If the bar yields before splitting or excessive slip occurs, the bond stress calculated must be below the value of the bond strength. Although not fully identified, the slip after which consistent drop in load occurs should be the point at which failure of bond by pullout is defined. Also, if splitting of the concrete encasing the bar occurs, the load at which such failure happens should also be the point at which failure of bond by splitting is defined. Failure in bond has thus been associated with the phenomena of splitting and slip. These phenomena are inter-related. In undeformed bars, failure is usually associated with large slip resulting from the breakdown of chemical adhesion followed by the failure of frictional resistance. For deformed bars, however, the situation differs significantly due to the presence of the ribs. Lutz and Gergeley [4] described the situation for ribbed bar as being a manifestation of the dominance of one of two actions. The first is a wedging action exerted by the rib on the concrete by trying to lift that part of the concrete upward and simultaneously pushing it against the slope of the rib. This action would induce ring stresses in the surrounding concrete that would result in splitting if the tensile strength of concrete is exceeded. It may be concluded here that the surface condition of the rib, the extent of chemical adhesion between the rib face and the concrete, the spacing between the ribs and the value of the angle of the rib all have a significant effect on the slip of the bar with respect to the concrete. The second action is that of the ribs bearing against the adjacent concrete between pairs of ribs. This action may result in the crushing of the concrete in the area between the ribs. Movement of the bar would then occur as a result of the failure of the resisting concrete layer in front of it. This would also cause slip until the crushed concrete powder in front of the rib is itself compacted and starts to act as an inclined face of a rib, albeit with a different coefficient of friction between it and the remaining concrete. This situation may produce a wedge action causing the remaining concrete to be pushed upward. This effect would create ring
236
O. Kayali
pressure around the bar, eventually resulting in splitting of the concrete and sliding of the bar with respect to it. This mode of failure is summarized in Fig. 7. Rehm [5] and Lutz and Gergely [4] found that a rib-face angle smaller than 408 would be more inducive to the wedge type of action, while crushing of the concrete in front of the ribs would be more dominant where the rib-face angle is larger than 408. Unusual surface conditions, such as the presence of grease, for example, would make wedge action more likely to dominate, even at rib-face angles larger than 408 [4]. The exact process by which splitting occurs in conjunction with excessive slip has been the subject of much research. Several models have been postulated to arrive at a clearer understanding of the way in which cracks form around bars in concrete, their types and order of formation [1]. Analysis has also been undertaken of the stresses developed in the concrete around ribbed bars, taking into consideration various cover and placement relationships [6–8]. The ribs, inclined at an angle, transmit compressive forces to the concrete in a manner similar to the situation of water pressure being transmitted to a thickwalled cylinder. Here, the inclined radial pressure emanating from the bars is similar to the water pressure, and the thick-walled cylinder is represented by a cylinder of concrete surrounding the bar and having a thickness equal to the smallest cover. As a result of the radial forces, the concrete in this thick cylinder experiences circumferential ring stresses which, if they exceed the tensile strength of concrete, will cause splitting parallel to the bar length. The planes of cleavage of such splitting can be of several orientations, as shown in Fig. 8. If the concrete cover is sufficiently thick, splitting may not occur as the bar may slip and be pulled out accompanied by shearing of the concrete between the ribs from the layer of concrete just above the ribs. In this case, the bar is deemed to have failed in bond, albeit at an upper-limit bond strength [9]. The influence of rib geometry on the development of stresses in the surrounding concrete has also been widely studied, with most researchers finding significant effects due to differences in the ratio of rib height to rib spacing
Figure 7: Wedge action failure between two ribs [2] (with permission of John Wiley & Sons, Inc.).
Bond of Steel in Concrete and the Effect of Galvanizing
Figure 8: Splitting and cleavage orientation as affected by bar arrangement and position: (a) single bars and (b) spliced bars [8] (with permission of the American Concrete Institute).
237
238
O. Kayali
[10 – 12] and the effect of the face angle of the rib [6,10]. The effect of confinement on bond strength has also been studied [13]. Other concrete-related parameters, such as casting position and mix workability, have also been found to affect bond strength [14].
8.3. Evaluating Bond Strength Bond strength has usually been evaluated by the pullout test, a standard procedure, which is detailed in ASTM C234 [3]. The principle of the method is shown in Fig. 9 [15]. In essence, a cube or cylinder of concrete or mortar is reinforced with a defined length of the bar to be studied and tensile load is applied to pull the bar from the hardened mix. Slip of the bar may be measured on both the loaded and the free ends of the cube or cylinder. In such a test, slip may start quite early at the loaded end although no slip would be observed at the free end. Together with the onset of slip, the bond stress between the concrete and the bar rises in the vicinity of the loaded end but decreases to zero along the bar before the free end is reached. As loading increases, the magnitude of the stress and the extent of slip both increase along the bar towards the free end. At any instant, the value of bond stress is taken as the average stress over the full embedded
Figure 9: Pullout test arrangement [15].
Bond of Steel in Concrete and the Effect of Galvanizing
239
length of the bar. This is, however, far removed from the true stress distribution, which is more realistically represented in Fig. 10 [16]. Failure in bond is usually deemed to have occurred by the time slip is recorded at the free end. Failure can, however, take more than one mode. For example, if the bar was ribbed then it is expected that failure will occur by splitting of the concrete. This is due, as has been discussed earlier, to the ring forces developing in the concrete surrounding the steel bar. In actual structural situations, this type of bond failure can be observed in flexural members, especially when the cover to the reinforcement is relatively thin. If the bar was plain (i.e. smooth), failure is expected by pullout of the bar and this will be more probable the shorter the embedded length of the bar. If the embedded length of the bar is somewhat longer, a tensile failure of the bar may be obtained. In all of these cases, as the bar is loaded in tension, the surface of the concrete at the loaded end is restrained in compression. This loading regime has the effect of magnifying the actual slip and closing any cracks that might develop. It may thus be concluded that the pullout test, in this simple form, is not representative of real loading conditions, especially in beams. The test may serve only as a basis for comparison between bars of different sizes, embedment length, rib geometry, surface condition and variations in concrete type, etc. Recognizing that the stress derived from the pullout test is inaccurately assumed to represent uniformly distributed bond stress, researchers have attempted to overcome this inaccuracy by performing pullout tests on bars with very short embedment length [17]. However,
Figure 10: The variation in bond stress and slip along the bar in pullout test [16] (with permission of John Wiley & Sons, Inc.).
240
O. Kayali
Figure 11: Bar size effect on bond stress and slip [19] (with permission of the American Concrete Institute).
this method is believed to yield an unrealistically high estimation of bond strength and give a large variability [18]. Abrishami and Mitchell [19] have addressed this problem by proposing a new testing technique, which enables the complete bond stress versus slip response to be determined. Their work clearly demonstrated the two major bond failure modes, namely splitting and pullout. They also showed (see Fig. 11) that bars with smaller diameter exhibit larger bond strength and that a slip value of about 0.36 mm associated with bond failure is similar for both small and large bars. Fig. 11 also indicates a splitting-type failure where the bars continue to show slippage after the peak load has been achieved. On the other hand, when pullout failure was observed, a sudden brittle failure occurred in the sample and no post-peak load was able to be recorded. These effects are shown in Figs. 12 and 13. Fig. 13, in particular,
Figure 12: Pullout failure representation [19] (with permission of the American Concrete Institute).
Bond of Steel in Concrete and the Effect of Galvanizing
241
Figure 13: Pullout-type failure and splitting-type failure [19] (with permission of the American Concrete Institute).
illustrates the difference between the pullout and splitting modes of failure and shows that the ultimate bond stress in pullout is much higher than that achieved with the splitting failure mode for the same size of steel bar. Also, in the pullout shear failure mode, smaller diameter bars exhibited larger bond stress capacity. The above discussion serves to illustrate the difficulty in devising a test that can give a reliable estimate of bond with minimum influence from those parameters governing changes in the concrete and the reinforcement. It also shows that bond strength is not a single value for a certain combination of concrete and bar; rather, it is a variable that depends upon many factors, among which is the mode of failure, whether by splitting or pullout. The mode of failure itself is also dependent upon several factors, which include the cover depth, the concrete strength, the reinforcement size, the presence of coatings on the steel, the size of the concrete member and the confinement of the main reinforcement.
8.4. The Bond Stress-Slip Relationship Slip of the bar with respect to the surrounding concrete is inevitable, even at very low levels of loading. What is somewhat difficult to define is the amount of slip
242
O. Kayali
that is associated with failure in bond. This is because, as discussed above, failure in bond can be initiated by either splitting or pullout, the slip associated with each at failure is different and many factors are significant in influencing the mode of failure. As also noted, the ultimate bond stress that can be experienced is different for the same type of failure depending upon a number of factors such as, for example, the size of the bar. What may be encouraging, however, is that there is evidence that the amount of slip associated with ultimate bond capacity for the same type of bond failure is approximately similar [19]. Ferguson [16] has demonstrated that, in a pullout test, the free end would not be experiencing any significant slip while maximum slip would be obtained at the loaded end. ASTM C234 pullout test [3] specifies the slip value as the average of two dial gauges, both mounted against the loaded end, and the slip at which a pullout type of failure has occurred as 2.5 mm. This of course does not apply if the type of failure was by the reinforcing bar reaching its yield point or by splitting of the enclosing concrete. Some researchers have preferred to consider the value of slip as the average between the measured values at the loaded and the free ends of the pullout specimen [19]. Clearly, this value would be less than the value obtained for the same specimen if the slip at the loaded end only was considered. Other researchers have taken the slip values at the free end only to represent slip in the bond– slip relationship. While such values will be much smaller at failure than those considering the loaded end alone, or the average between the two ends, this approach may be significant in identifying specimens whose failure was by pullout rather than splitting. If bond failure was by splitting, very small slip may be measured at the free end at the time of failure and therefore the comparison between different concretes and/or steel reinforcement may be obscured if only the free end slip was taken. Moreover, it has been observed that for plain (i.e. smooth) bars, an average value of bond strength may reasonably be calculated by dividing the force in the bar at bond failure by the nominal circumferential area acting in bond. This is not the case, however, for deformed bars where it is recognized that 70– 90% of the bond strength is transmitted by the ribs and such calculation is less meaningful [20]. Windisch [20] also points out that, for deformed bars, the unloaded end slip is not characteristic for the whole slip situation along the embedded length. On the other hand, if the failure is initiated by pullout, meaningful results [21] may be obtained from this procedure, although a certain slip value has to be defined as the demarcation of failure by pullout and this value must be less than the 2.5 mm at the loaded end, as suggested by ASTM C234. This approach also provides a basis of comparison with beam specimens where it is relatively easy to measure slip values at the free ends of the reinforcing bars. Mathey and Watstein [22] suggested either a value of 0.25 mm for the loaded end or 0.05 mm for the free end as the value of slip that defines critical bond stresses. They advised that, if these slip values are
Bond of Steel in Concrete and the Effect of Galvanizing
243
not exceeded, the bond stresses corresponding to them are sufficiently low such that under-reinforced beams would fail by yielding of the reinforcement. This same criterion has later been applied to compare the performance in bond of steel reinforcement with different coatings by measuring the slip at the free end of the bar [23].
8.5. Evaluation of Bond in a Flexural Member As discussed above, the bond stress is evaluated from sb ¼ ðdT=dxÞ=ðpdÞ
ð13Þ
and therefore bond stress is usually not uniformly distributed along the bar. Thus, expressing bond stress as sbav ¼ ðdb =4Þ £ ð fst =LÞ
ð3Þ
is a rough estimation of average stress, which is much less than the maximum stress occurring where dT=dx is a maximum. One can say at this time that the equation which expresses bond stress in the flexure situation sb ¼ V=ðpd zÞ
ð11Þ
is more representative of the bond stress occurring at a particular point of the flexural element. It may, however, be more accurate to conclude that, in a beamflexural situation, two kinds of bond stresses occur [4]: the anchorage bond stress expressed as Eq. (3) and the flexural bond stress due to shear generally expressed as sb ¼ V=ðNpd zÞ
ð14Þ
where N is the number of bars. Recognition of these two kinds of behaviour lies at the core of developing equations for design recommending minimum development length and splice length requirements. Orangon et al. [8] proposed the following expression for steel stress pffiffiffi ð15Þ fs ¼ ½1:2 þ 3C=db þ 50db =ls þ Atr fyt =ð500sdb Þ £ ð4ls f 0c Þ=db where C is the smaller of the bottom (or top) cover or one half of clear spacing between splices, Atr the area of the transverse reinforcement per developed bar, fyt the yield strength of transverse reinforcement, s the spacing of transverse reinforcement centre-to-centre or the development/splice length divided by the number of stirrups, and db is the diameter of the developed/spliced bar.
244
O. Kayali
Note that in Eq. (15) all stresses are in psi and lengths in inches. It may be noted here that this expression represents the stress that can be developed in a length ls of the bar. The above expression has been the basis of the ACI design requirement for development length [24], as stated in the equation ld =db ¼ ð3=40Þ £ ð fy =ð f 0c Þ0:5Þ £ ðabgl=ððc þ Ktr Þ=db ÞÞ
ð16Þ
where ld is the required development length, db the bar diameter, fy the specified yield strength of the reinforcement, f 0c the specified compressive strength of concrete, a the position of steel factor, to take into account the adverse effect of top reinforcement position, b a coating factor to account for epoxy coating, g the reinforcement size factor to reflect the more favourable performance of smaller size bars, l the lightweight concrete factor to reflect the adverse effect of generally lower tensile strength of lightweight concrete, Ktr a factor representing the contribution of confining reinforcement, and c is a factor to represent the smallest of the side cover, cover over the bar or one half the centre-to-centre spacing of the bars. Since, as explained above sbav ¼ ðdb =4Þ £ ð fst =LÞ
ð3Þ
the expression of Orangun et al. may be presented to express the average bond stress that can be obtained in the developed bar or splice as pffiffiffi sbav ¼ ½1:2 þ 3C=db þ 50db =ls þ Atr fyt =ð500sdb Þ £ f 0c ð17Þ It may also be noted here that, as the expression for fs above represents the stress that can be developed in the length ls ; it follows that the expression for sbav represents the bond stress that can theoretically be obtained between the bar and the concrete for the given length ls and the particular conditions of cover, transverse reinforcement and placement. This value is therefore the maximum average bond stress that can be obtained between a particular bar and a concrete of particular strength when the stress in the steel has reached a certain fs corresponding to the embedment length ls : Hence, the value of the bond stress that can be obtained from the expression above is limited by the value of bond strength corresponding to the actual situation and the concrete strength. A value which may be adopted for bond strength is subject to definition. Such definition depends upon the acceptable crack width associated with a certain slip of the bar. Thus, the term useable bond strength [2] may be employed to define a critical bond stress at a certain value of slip. Mathey and Watstein [22] defined the critical bond stress in flexural situations as the lesser of the bond stress value associated with a free-end slip of 0.002 in. (0.05 mm) or a loaded-end slip of 0.01 in. (0.25 mm), this being equal to half the acceptable crack width. Ferguson et al.
Bond of Steel in Concrete and the Effect of Galvanizing
245
[25] used this criterion in pullout tests and their results have been represented by Park and Paulay [2] as in Fig. 14, which shows the tensile stress in bars at the critical slip value. It also shows the tensile stress in bars at ultimate capacity. While the criterion of critical slip value is quite convenient for defining bond failure in pullout tests, it has been pointed out that a lower slip value may be more appropriate in beam tests. This is especially so when such tests measure slip values at the free end of the beam in flexure [23]. Evaluating bond strength by means of beam testing has been the subject of significant research in recent years. The importance of this is mainly because the standard pullout test subjects the concrete to compression and the steel to tension, which is not the case in a flexural member where the concrete surrounding
Figure 14: Tensile stresses in bars at values of critical slip and ultimate strength [2] (with permission of John Wiley & Sons, Inc.).
246
O. Kayali
the tensile steel is also under tension. It is therefore not surprising that results relating slip to load on the steel bar are quite different for the two types of tests. Chana [26] and Clarke and Birjandi [27] demonstrated the inadequacy of some standard pullout test results, which they have shown to give unrealistically high bond strength values and result in failure modes that are not typical of those occurring in practice. This has been especially the case for deformed bars [20,26,27]. Beam tests have also shown that slip values may be similar at ultimate flexure loads, while at service loads they differ significantly and thus reflect different crack and deflection behaviour [23]. RILEM has described a Standard Beam Test [28] where the bond strength, which is assumed to be uniform along the embedded length, is calculated from Eq. (3) and where fst is the stress in the bar at the load causing failure in bond. The RILEM Standard Beam Test, however, does not exactly specify when a failure by bond is considered to have occurred. Quite recently, a variety of research using different configurations of the beam test has been reported [11,14,29–38]. ASTM has also introduced a beam end test [39] that simulates the behaviour of reinforcement in flexural members. This test is based on extensive research by Darwin and Graham [11], and has proven satisfactory in providing meaningful results on various aspects of the bond phenomenon of deformed reinforcement in structural concrete. Their method has been very recently employed with some modifications to study the bond stress and slip in galvanized, epoxy-coated and black steel reinforcement [40].
8.6. New Developments in Concrete 8.6.1. Higher Strength Concretes The rapid progress in cement chemistry and in chemical admixtures has made it possible to produce concrete of strengths that were not envisaged by current Standards. It is now normal to produce concrete whose characteristic strength exceeds 50 MPa by ordinary ready-mix concrete firms. It is also relatively easy to produce concrete of flowable workability that possesses compressive strength in excess of 100 MPa. Recently, very high-strength concrete has been produced under the trade name of Compresit. This contains a high amount of cement and silica fume with a maximum aggregate size of 4 mm, up to 10% by volume of straight steel fibres, and with compressive strengths in the range 140–170 MPa [41]. Reactive Powder Concrete, produced in a similar way to Compresit and heat treated, gives compressive strength in the range 200–800 MPa [42]. While these types of concretes are now taking their position in the construction industry, it remains that few, if any, of them have been fully investigated as far as
Bond of Steel in Concrete and the Effect of Galvanizing
247
bond is concerned. In a recent international conference on high-strength concretes, it was concluded that bond strength does increase with compressive strength [43]. It has also recently been reported that bond strength, being dependent on the tensile strength, increases as the compressive strength of concrete increases and has been expressed in terms of the square root of compressive strength [44]. It has, however, been reported that the increase in bond strength with compressive strength occurs at a diminishing rate until the strength reaches 95 MPa [45]. This is contrary to the previously held view, based on experimental results, that the increase in bond strength becomes smaller above a compressive strength of about 20 MPa [46] and even negligible after about 35 MPa [47]. This indicates that the newer high-strength concretes differ substantially from previous generation concretes, even though their strengths may be comparable. Such differences are believed to be due to the use of new ingredients, necessary to develop the high strength, but which cause changes in the bond property to an extent that warrants special consideration in Codes of Practice. Such changes in codes and standards in the area of bond and spliced reinforcement have indeed been suggested for those situations where high-strength concrete is used [36]. Reports on this work, however, are still not sufficient to make general recommendations for design purposes. High-strength concretes usually gain their properties because of the use of super plasticizers, condensed silica fume, and ultra-fine cements. It is therefore expected that these ingredients will have some effects on the three main components of bond strength, namely adhesion, friction and mechanical interlock. Indeed, it has recently been found that the use of super plasticizer in the mix substantially increased the bond strength [48]. The effect of silica fume on bond has also come under investigation. Hwang et al. [49] found that silica fume reduced bond strength. At this time, it is also appropriate to re-examine the role of adhesion as it is likely that, with the new generation of concretes, there will be a greater chemical role in the concrete-bar interface zone. Investigations along the lines of Khalaf and Page [50] on the microstructural features of the interface may contribute to this.
8.6.2. Structural Lightweight Concrete The concrete industry is witnessing a resurgence of interest in lightweight concrete, in particular that made from industrial by-products, for structural applications. Attention has therefore been duly directed towards the bond properties of such concrete. The general view has been held that lightweight aggregate concrete is brittle and therefore subject to splitting at a much lower bond stress than in the case of normal weight concrete [9]. Provisions in the Codes of Practice had been made for modification of the bond strength, usually by a reduction factor, when
248
O. Kayali
lightweight structural concrete is used [51–53]. Whether a reduction in bond strength is really necessary, however, is still a matter of dispute, especially so with high strength, structural lightweight concrete. For example, Rossetti [54] has indicated that such concretes behave satisfactorily in bond while Mor [55] has proposed that lightweight aggregate concrete should show better bond characteristics than normal weight concrete, mainly because of the presence of superior interfacial zones between the paste and the aggregates and between the paste and the steel. It has variously been reported that condensed silica fume does not have a significant effect on bond in normal strength concrete [45] and that it even reduced the bond strength [49]. In contrast, Mor [55] found that the steel–concrete bond in high-strength, lightweight concrete, when taken at a slip level of 0.25 mm in 152.4 mm (6 in.) cubes, being the maximum acceptable for structural performance, increased by 2-fold as a result of using condensed silica fume. Mor attributed most of the gain in bond strength at the serviceability load to improved adhesion, which was more pronounced in lightweight concrete where condensed silica fume was used [55,56]. In related work, Ezeldin and Balaguru [57] found that silica fume resulted in higher bond strength although the bond failure in such cases was brittle.
8.6.3. Fibre-Reinforced Concrete As a result of the expanded use of high-strength concrete, the problem of brittleness of such concretes has become prominent, especially in its effects on bond and shear strength. In the endeavour to solve this problem and enhance the fracture toughness of high-strength concrete in particular and concrete in general, fibre reinforcement has been utilized. The contribution of fibres to the fracture toughness of concrete has been widely recognized and this contribution undoubtedly affects the bond strength. Ezeldin and Balagaru [57] reported a significant increase in the bond strength of high-strength concretes when reinforced by steel fibres and that the general ductility of such concrete was enhanced. They attributed the increase in brittleness in highstrength concrete to the incorporation of silica fume and credited the fibres for the reversal of such effect. In recent developments in high-strength and ultra-high-strength concretes, 0.4 mm diameter £ 12 mm long steel fibres were used to produce concretes of strengths ranging from 140 to 800 MPa with flexural strengths as high as 140 MPa [41,42]. This represents a radical change to the conventional flexural strength of concrete and would be expected to have some consequences on the bond characteristics. Relatively recent research has shown that the addition of thin and short steel fibres in very high-strength concrete has resulted in a 3-fold increase in the bond strength between the reinforcing bars and the concrete [41].
Bond of Steel in Concrete and the Effect of Galvanizing
249
8.6.4. The Role of Galvanized Reinforcement To date, there has been no reporting on the use of galvanized reinforcement in high-strength concrete. Similarly, there are no data on the use of galvanized fibre reinforcement in high-strength, ultra-high-strength or structural lightweight concrete. Since the aim of the use of the new materials and production techniques is to produce concrete of high performance and superior durability, it is expected that the utilization of galvanized reinforcement, whether conventional or fibre, may result in significant improvement in the structural and durability performance. The role of galvanizing in imparting superior corrosion resistance has been well documented in this publication and elsewhere. In contrast, however, the role of galvanizing in improving the mechanical performance of conventional and fibre reinforcement has not been fully investigated. The following sections address this issue from the perspective of bond strength.
8.7. Coated Reinforcement and Bond The corrosion of steel reinforcement in concrete and methods of protecting steel from corrosion have been extensively discussed in other chapters in this book. What is of interest here is whether such protection methods, which involve a coating to the steel, affect the bond characteristics and to what extent. As noted elsewhere, the most common methods of protecting steel reinforcement with coatings is by coating with either organic films, mainly in the form of fusionbonded epoxies, or with zinc by galvanizing. The effects of these coating systems on the bond-slip characteristics of reinforcement are considered below.
8.7.1. Fusion Bonded Epoxy Coating Of the organic coatings, the most versatile and compatible is the fusion-bonded epoxy coating to the extent that this has become virtually the only non-metallic coating used for steel reinforcement in concrete [58]. The effect of this method on bond has been the subject of continuous research. Early researchers in this area suggested that the bond strength of epoxy-coated bars became unacceptable [59]. As previously noted, Lutz and Gergeley [4] predicted that the presence of grease on the surface of deformed bars would favour wedge action even at rib-face angles larger than 408. Epoxy coating provides a very smooth surface, which may resemble grease in its effect and therefore slip is expected to occur earlier and the resulting ring pressure would be expected as the mode of failure.
250
O. Kayali
The first pullout tests performed on epoxy-coated ribbed bars embedded in concrete and compared with black steel bars were reported by Mathey and Clifton [60]. In their research, it is evident that the question of determining the slip value associated with “bond failure” was a matter of definition. A critical bond stress was taken as the value corresponding to a slip of 0.25 mm at the loaded end or 0.05 mm at the free end [22]. According to this definition, the results indicated that ribbed bars with epoxy coatings approximately 250 mm or less in thickness were not different from uncoated bars in bond strength. Thicker coatings were found to significantly reduce the bond strength. At the present time, the usual specified film thickness after curing is 175– 300 mm, based on the fact that the corrosion resistance of the bar increases with the film thickness while adhesion of the film to the bar is reduced [61,62]. Investigation of the bond characteristics of epoxy-coated reinforcement in concrete has been quite extensive. Recently, the bond capacity of ribbed bars was found to be reduced significantly because of epoxy coating [63]. Accordingly, the ACI Code requires the basic development length for ribbed epoxy-coated bars to be increased due to the loss of bond strength by the presence of the epoxy coating [24]. It has also been found that the geometry and profile of the surface deformations influence the bond strength [64]. In other work, Cairns and Abdullah [65] concluded that the reduction in bond strength due to the epoxy coating can be lessened with an increase in relative rib area. They did indicate, however, that concrete strength and rib-face angle did not influence the ratio between the bond strength of epoxy-coated and non-coated bars. This observation is, however, an area of some disagreement among researchers. For example, Hamad [66] found that the bond strength of epoxy-coated bars was significantly affected by the rib geometry and recommended certain optimum geometrical characteristics to improve the bond strength of epoxy-coated bars. Such optimization would result in steeper rib surfaces and closer and higher ribs. In this work, Hamad acknowledged the difficulties that the implementation of these recommendations would pose as far as the application of a uniform and efficient epoxy coating was concerned. Failure to achieve such application would contravene the requirements of the standards and result in corrosion problems.
8.7.2. Hot-Dip Galvanizing 8.7.2.1. The Question of Bond of Galvanized Reinforcement It is to be expected that the reduction in bond strength observed with epoxy-coated reinforcement will not apply to the use of galvanized reinforcement. On the contrary, an increase in bond strength with galvanized reinforcement should not be dismissed out of hand. Research in this area, however, has not been as exhaustive as
Bond of Steel in Concrete and the Effect of Galvanizing
251
that for epoxy-coated reinforcement and some conflicting results have been reported. In the 1920s, Slater et al. [67] studied the bond strength of galvanized and non-galvanized bars in concrete and found that, in some cases, the time required to develop full bond strength may be longer for galvanized than non-galvanized bars. Some 30 years later, Robinson [68] conducted pullout and short-beam tests on galvanized bars, clean and smooth mild steel bars and rusted and appreciably pitted mild steel bars. All bars in this test programme were non-ribbed (i.e. smooth bars). The results showed a large reduction in bond strength in galvanized bars when compared with the rusty and pitted bars, but also a large increase in bond strength of galvanized bars when compared with the clean, smooth bars. These tests indicate the sensitivity of the bond strength to the roughness of the bar surface. Kayali and Yeomans [23] conducted a number of flexural tests on beams that were reinforced with either black, epoxy-coated or non-chromate-treated galvanized bars. These tests indicated that there may not be a significant difference between the ultimate capacity in flexure between beams reinforced with ribbed bars of the same rib geometry whether they were black, epoxy-coated or galvanized. In more recent work, Kayali and Yeomans conducted beam-end bond tests on 28-day-old specimens containing ribbed reinforcement of either black bars, epoxy-coated bars or galvanized bars [40]. Continuous measurements of slip at the free end of the bar were made while loading in a flexure-like mode, whereby both the bar and the surrounding concrete were under tension such as in actual load situations. Six beams were tested for each of the black and epoxy-coated bars, while 12 beams were tested with galvanized bars. Based on a rigorous statistical analysis, their results showed that there was no significant difference at ultimate load between the free-end slip of galvanized bars and black bars. However, a significant difference between epoxy-coated and black or galvanized bars was observed. If the galvanized and black bars were statistically considered as one group, the mean value of slip of the epoxy-coated bars was about 200% greater than the value for the galvanized and black group. A summary of these results is represented in Fig. 15.
8.7.2.2. The Phenomenon of Hydrogen Evolution Bird [69] reported the results of work by Lewis and Laurie (1958), which showed no reduction in bond strength for galvanized bars. Bird indicated that the hydrogen evolution at the surfaces of galvanized steel immersed in Portland cement paste occurs on surfaces where iron and zinc are in contact but not on the surface of pure zinc similarly immersed. This suggested that it was the zinc–iron alloy layers reaching the surface of the galvanized coating that initiated the formation of hydrogen resulting in a porous contact layer that subsequently impaired the bond. Other work in this area by Swamy [64], Porter [70] and Yeomans [71] has confirmed this view.
252
O. Kayali
Figure 15: Load – slip relationship for black, epoxy-coated and galvanized ribbed bars with 150 mm embedment [40].
The process of hot-dip galvanizing, discussed in detail elsewhere in this book, results in the formation of metallurgically bonded zinc–iron alloy layers. Each successive layer from the steel substrate outwards contains a higher proportion of zinc and the outermost layer is relatively pure zinc [71,72]. Thus, the evolution of hydrogen in such circumstances is not expected to be significant if the outer layer of the coating was predominantly pure zinc, even if it were quite thin [69]. It has been shown that the cathodic evolution of hydrogen in the favourable condition of exposed steel in contact with zinc lasts for less than about 1 h because the galvanic cell between the iron and zinc is rapidly polarized [69]. It therefore follows that, to prevent hydrogen formation, it is necessary to maintain the presence of a pure zinc layer at least for the first hour of fresh concrete being in contact with galvanized steel. Once the concrete has started to harden and set, the loss of the pure zinc layer at this time will have no effect on the bar/paste interface as far as hydrogen evolution is concerned. Hence, the risk of creating a situation where hydrogen evolution may occur, resulting in a potential loss of bond, is mainly dependent on the presence of pure zinc in the outer layer coating.
Bond of Steel in Concrete and the Effect of Galvanizing
253
8.7.2.3. The Role of Calcium Hydroxyzincate The situation may, however, be more complex than outlined above. Zinc is known to react with fresh concrete, resulting in the formation of calcium hydroxyzincate accompanied by the evolution of hydrogen [73]. Calcium hydroxyzincate has been shown to be a fibrous hydration product whose presence adjacent to the surface of the reinforcement is believed to positively contribute to the bond between the reinforcement and the surrounding concrete [74]. Arliguie et al. [75] used microscopy to show that this particular hydration product acts as a physical anchorage. Blanco et al. [76] demonstrated the presence of the dense formation of a very compact layer of calcium hydroxyzincate crystals on the galvanized steel surface. They also related the size of these crystals to the pH level while Fratesi et al. [77] produced further SEM/EDXA evidence of the formation of calcium hydroxyzincate at the surface of galvanized bars. The structure of calcium hydroxyzincate in these tests was crystalline, growing in irregular orientation with the bar surface. They suggested that these crystals act as bridges between the metal surface and the concrete, thereby strengthening the adhesion of the bars. This effect is also qualitatively confirmed in the observation that Portland cement mortar droppings tend to adhere strongly to zinc and, when fully hardened, become very difficult to remove without damaging the metal [78]. What is likely, therefore, is that the adhesion gained from the formation of calcium hydroxyzincate may compensate for the loss of bond due to hydrogen evolution and that this accounts for the reported comparability between black and galvanized bars as far as bond is concerned [21,23,79,80]. The bond strength between metal surfaces and cement paste has been shown to vary significantly for different metals. Khalaf and Page [50] studied the adhesion/ cohesion bond strength of mild steel, stainless steel, copper and brass to cement paste and mortars and reported that the adhesion of the paste or mortar to copper or brass is significantly higher than it is for steel. They attributed this effect to the partially acidic character of the oxide films on copper and brass and showed that the failure surface in the case of steel was located at the steel–paste interface and revealed a predominant formation of calcium hydroxide crystals. On the other hand, the fracture surfaces for copper and brass were predominantly away from the interface and through the zone of hydration products adjacent to the interface; that is, the failure is a cohesion rather than adhesion failure. This observation is similar to that obtained from pullout tests on galvanized samples [21,81]. Arliguie et al. [82] microscopically examined the Portland cement–zinc interface. They produced evidence that the surface of the metal is largely covered by calcium hydroxyzincate, which very strongly adheres to the zinc. They also found that, in the presence of zinc, the formation of calcium hydroxyzincate is preferential to that of portlandite to the extent that no portlandite crystals were to
254
O. Kayali
be found at the zinc-paste interface. The formation of portlandite crystals occurs after the formation of calcium hydroxyzincate and hence the interface is free of portlandite crystals. Moreover, the tortuous relief and irregular orientation of the calcium hydroxyzincate crystals prevent the portlandite crystals that eventually develop within the transition zone from forming in orientation parallel to the nearby metallic surface [82]. This, of course, results in a strong adhesion bond and also increases the strength of the cohesion bond in the transition zone by virtue of the disruption in the portlandite crystal orientation. Nevertheless, the adhesion bond between the zinc and the paste is stronger, as evidenced by the failure surfaces being through the transition zone rather than the interface. There is now ample evidence that the adhesion between the zinc and cement paste is relatively stronger than that between the steel and the paste [21,78,81,82]. Pullout tests conducted on zinc, iron and copper wires embedded in small C3A1 paste cylinders showed that the bond between the zinc and paste was much larger than that between the iron wire or the copper wire and the paste [83]. Tashiro and Ueoka [83] showed that the bond strength of zinc to cement paste continued to rise with age in a straight-line relationship, while little further gain in bond strength was obtained after the first 7 days in the case of iron wires. Similar observations were made by Belaı¨d et al. [81] in pullout tests on smooth galvanized steel bars compared with smooth black steel bars. In these tests, the superiority of the galvanized bars was significant after 7 days age although, prior to 7 days, the results for galvanized bars were not significantly different from those of black steel [81]. Black steel is surrounded mainly by CH hexagonal crystals, which possess planes of cleavage that are expected to result in weakening the adhesion bond of the hardened paste to the steel surface, even in mature hardened paste. It is generally held that the C – S – H gel, which constitutes about 50–60% of the completely hydrated paste is the major source of strength in the hardened cement paste while the CH, which constitutes about 20 –25% of the volume of solids in the hydrated paste, has considerably less strength [84]. It has been suggested that the tendency of CH crystals to cleave under shear limits its contribution to strength [85]. There is also evidence from scanning electron microscopy and energy dispersive X-ray analysis that the steel – hydrated cement interface is enriched with calcium hydroxide in excess of that for material beyond the interface [50,82, 86,87]. Portlandite crystals have the tendency to grow in an orientation such that their cleavage plane is parallel to the surface of contact, thereby weakening the interface [82]. It has also been shown that the microhardness of the hardened cement paste, and thus its compressive strength within a region of 250 mm from the interface, is 1
In cement chemistry, C ¼ CaO, A ¼ Al2O3, S ¼ SiO2, CH ¼ Ca(OH)2 and H ¼ H2O.
Bond of Steel in Concrete and the Effect of Galvanizing
255
significantly less than in the bulk of the material. The decrease in this property has been associated with an increase in porosity of approximately 25% [87]. The greater porosity in this region and the presence of weak cleavage planes as a result of the enriched calcium hydroxide interface may provide the limit of the adhesion component of the bond strength between steel and concrete.
8.7.2.4. Retardation Effect of Zinc and the Transition Zone Early research suggested that zinc may have a retardation effect on the set and early hardening of concrete and thus may exert an important influence on bond with galvanized steel [88]. Retardation had been shown to occur when zinc powder was added to cement and it has now been reasonably established that zinc and its oxide cause retardation in the early hydration process of cement paste [83,89,90]. This retardation is caused by the precipitation of zinc hydroxide onto the anhydrous grains of cement. The zinc hydroxide exists in both crystalline and amorphous forms and is not readily dissolved until the prevailing pH exceeds values between above about pH 12. The formation of zinc hydroxide is also preferential to that of portlandite. The retardation effect is thus caused by both the formation of a physical barrier between the cement grains and the pore solution, thereby preventing the hydration of C3S and C3A, and by the preferential formation of the zinc hydroxide. When the alkali concentration of the medium increases beyond 12.8, very soluble zincate ions of Zn(OH)-3 and Zn(OH)24 form, which react readily with calcium ions and water to form a solid precipitation of calcium hydroxyzincate, CaZn2(OH)6·2H2O [89]. In more recent work, Belaı¨d et al. [91] studied the pore-size distribution in the transition zone for galvanized steel and black steel in contact with cement paste. At an early age, they found that the transition zone next to the galvanized steel was more porous than in the case of black steel. At 28 days, however, the difference disappeared. Their work confirmed that, due to the early retardation effect of zinc and/or the hydrogen evolution at the interface, the transition zone becomes porous. However, with the cessation of the retardation effect, calcium hydroxyzincate precipitates and fills the pores to the extent that no significant difference is recorded between the two cases. This discussion sheds further light on the observation by Slater et al. [67] that the development of full bond strength requires more time in the case of galvanized bars when compared with non-galvanized reinforcement. In metallographic studies of the reaction of zinc with cement mortars, Yeomans [92] observed that the interfacial/transition zone between the galvanized bar and concrete was rich in zinc. This finding adds a new dimension to the observation of Hofsoy and Gukild [88] that zinc retards the hardening of concrete. If zinc ions do migrate, even within the small distance of about 10 mm as indicated by Yeomans [92], the retardation effect will be manifested within the interfacial/transition zone
256
O. Kayali
and thus have a significant reducing effect on the bond strength between the bars and the concrete at early age. This conclusion has been indirectly supported by the results reported in the work of Belaı¨d et al. [81,91], who found that, up to the age of 7 days, ribbed galvanized bars had a lower bond with concrete than ribbed black steel bars. In recent work, Hamad and Mike [93] reported a significant reduction in bond strength with galvanized bars compared with black steel bars. Their conclusions, however, were based on results obtained after only 7 days of curing. In view of the evidence of retardation caused by zinc as presented here, their results are entirely expected. Nevertheless, by the age of 28 days, any effect on bond that may have been caused by early retardation has been shown to completely disappear [40]. Belaı¨d et al. [81,91] also showed that, after 7 days, the galvanized bars continued to gain bond strength while the black steel bars did not and, after nearly 2 weeks, the galvanized bars achieved superior bond strength to black steel bars. As has been previously noted, however, in the case of ribbed bars it is the mechanical interlock that governs the failure in bond. Thus, in the work of Belaı¨d et al. with ribbed galvanized and black steel bars, the failure was by pullout. Therefore, the plane of failure was away from the interface. Thus, if the zinc resulted in retardation during the first 7 days, this would be reflected in the bond strength of the galvanized bars being lower than the black steel bars. This observation was precisely what Belaı¨d and her co-workers recorded at early age. It seems, however, that this retardation did not affect the further development of hydration products after 7 days.
8.7.2.5. The Practice of Adding Chromates Prevention of hydrogen generation on the surfaces of galvanized steel can be achieved by the application of a dilute chromate solution to those surfaces. This may be done by either applying the solutions directly to the galvanized bars or by adding chromates to the water of the concrete mixture [71]. The Concrete Institute of Australia, in its recommendations for the use of galvanized reinforcement in concrete [94], suggests that it is necessary to passivate the zinc in order to prevent the zinc– alkali reaction and that this can be achieved by dipping the galvanized reinforcement in a 0.2% sodium dichromate solution. It is important to maintain the correct solution strength to achieve passivation. Equally, however, it is important to remember that most of the steel treated in this way will not be immediately embedded in concrete and will likely be exposed to different periods and types of on-site storage prior to casting. Since the passive film does not survive on the surface for extended periods, especially in chloride-rich environments, the level of actual passivation cannot therefore be guaranteed or even be known to exist at the time of casting. This consideration gives credibility to the alternative practice of adding the chromates
Bond of Steel in Concrete and the Effect of Galvanizing
257
to the concrete mix. Chromates in the form of sodium or potassium dichromate may be added at the rate of 70 ppm, expressed as CrO3 by mass of cement. This is equivalent to 104 g/ton of cement of pure sodium dichromate dihydrate or 103 g/ton of cement of pure potassium dichromate [94]. Hofsoy and Gukild [88] showed that the chromate content in cement significantly affected the bond strength of galvanized bars. The retardation effect discussed above was also eliminated when the zinc coating was chromate treated [88]. Their tests on plain galvanized and non-galvanized bars showed that, when cement with water extractable chromate content in the range of 10 –20 ppm was used, the bond strength of galvanized bars was less than that of non-galvanized bars. The addition of potassium dichromate to the concrete eliminated this reduction in bond strength. In contrast, cement with chromate content between 30 and 40 ppm resulted in the galvanized bars achieving higher bond strength than non-galvanized bars. Their results for deformed bars showed similar trends with the reduction of bond strength in galvanized deformed bars in cement of low chromate being eliminated by the chromating of the bars before their immersion in concrete. Although these results are very interesting, the conclusions cannot be generalized because the tests were performed on 7-day-old specimens. It has been demonstrated that the age of the concrete has a larger effect on the bond of galvanized bar compared with non-galvanized bar. This occurs to the extent that, after the early age difference, the fully developed bond strengths of galvanized and non-galvanized bars were similar [67], and even plain galvanized bars were superior to non-galvanized bars in some instances [95]. Roberts et al. [96] conducted pullout tests on zinc reinforcing tendons and reported superiority in bond of galvanized over non-galvanized steel. They did not find a correlation between chromate addition and bond performance but reported that a low passivation of only 0.031% sodium dichromate was slightly superior to normally passivated (0.13% sodium dichromate solution) and highly passivated (0.395% sodium dichromate solution) galvanized steel tendons. It is to be noted, however, that their tests were also performed after only 7 days of curing and thus the effect of age referred to earlier has not been accounted for. Nevertheless, the indication here is that, when tendons which have special geometry were used, the reduction in bond observed in the case of 7-day-old specimens noticed by previous investigators was obscured and that chromate application lost its significance even at that early stage. The practice of adding chromate solutions to either the galvanized bars or the water of the concrete mixture is obviously time- and resource-consuming but, more importantly, it is a practice that has raised concerns from the health hazard point of view. Human exposure to chromium can produce allergic contact dermatitis, although there has been no evidence of increased risk of skin cancer [97]. Alternatively, the small quantities of chromates which most cements contain
258
O. Kayali
are sufficient to provide passivation to the galvanized surfaces, provided that these quantities produce at least 20 ppm of chromates in the final concrete mixture [78]. There is evidence, however, that the amount of naturally existing chromates in cement varies widely. Investigations carried out in Australia [98] and Europe [99] showed that the level of water-extractable chromates ranged between #1 and 21 ppm of cement. However, the amount of chromates extractable by sodium sulphates ranged from 47 to 110 ppm of cement and in two cases was 400– 510 ppm [98]. As cement contains sulphates mainly in the form of gypsum, it is expected that sulphate extractable chromates may become available in fresh cement paste. This may still be the case in spite of the expected reduction in water extractable chromates as part of recent mandatory health requirements in some countries. These require the addition of ferrous sulphate to cement to specifically reduce water extractable chromates [98]. The author believes, however, that future dependence on the presence of any amount of chromates in cement should be dismissed because of the impending extremely stringent regulations which impose a reduction of chromium ions in concrete to even less than 2 ppm [77]. Koch and Wohlfahrt [100] reported experiments on galvanized and nongalvanized ribbed bars in concrete with either low chromate content cement (3 ppm) or cement with normal chromate content in the vicinity of 30 ppm. They found that, in non-galvanized steel and bars with a thin zinc layer (69 mm), the bond strength was lower with higher chromate content while, for bars with thick zinc layers (220 mm), those embedded in the higher chromate content cement had superior bond strength. In a more recent publication, Fratesi et al. [77] reported the results of pullout tests on 2 months cured concrete samples reinforced with smooth bars of galvanized, black or chromate-treated galvanized steel. They showed that the bond strength of galvanized but non-chromate-treated bars was consistently and significantly higher than both the black bars and chromate-treated galvanized bars. Moreover, they found that galvanized bars with chromate treatment showed lower bond strength than black bars. The superiority of the non-chromate-treated galvanized bars is attributed to the formation of calcium hydroxyzincate crystals in a direction perpendicular to the galvanized surface [77]. A further finding in their work was that the bond strength of non-chromate-treated galvanized bars increased steadily with time when the samples were exposed to a chloride-rich corrosive environment. As the exposure to such environments is one major reason why galvanizing is utilized in the first instance, the improvement with time in bond performance of such reinforcement is of particular importance. Fratesi et al. attributed the continuously increasing bond strength of non-chromate-treated galvanized bars to the progressive densification of the interfacial zone by the penetration of nonexpansive zinc corrosion products, which fill the pores in the transition zone
Bond of Steel in Concrete and the Effect of Galvanizing
259
causing strengthening of the adhesion between the bar and the cement paste matrix. Chloride ions help dissolve zinc-corrosion products and thus facilitate their penetration and migration towards nearby cement paste voids. These results have been supported by Yeomans [92], who conducted SEM/XRD studies on samples of galvanized bars embedded in mortar cylinders immersed in chloride solution after 7 days of curing. Sakamoto and Iwasaki [101] also reported bond tests, which apparently did not include any chromate addition that indicated that the bond of deformed galvanized bars in concrete containing sodium chloride was superior to non-galvanized bars in the same condition. Andrade et al. [80] conducted a long-term investigation (to 10 years) on the bond of galvanized reinforcement in chloride-contaminated concretes. Their results indicated no loss in the bond of galvanized steel after such long-term exposure to chlorides. This is due to the non-expansive nature of the corrosion product formed when galvanized bars suffer chloride-initiated corrosion. In contrast, chloride-initiated corrosion of non-galvanized bars produces expansive products which impair bond. Moreover, their results did not show any significant effect of chromate addition when the bond of non-chromate-treated galvanized bars was compared with that of galvanized and chromate-treated bars. Andrade and her co-workers further indicated that the generation of hydrogen bubbles did not have a significant effect on bond strength. Nishi [102] reported that deformed galvanized and non-galvanized bars did not differ in their bond strength although the results varied with the type of deformation. Yeomans and Ellis [21] reported extensive pullout test results comparing black, galvanized and epoxy-coated plain bars. They found no statistically significant difference in the ultimate bond stress of plain black steel bars and galvanized bars. In comparison, the bond stress of epoxy-coated plain bars was some 26% less than that of the black steel or galvanized bars. The addition of chromates via the mixing water in the range 35– 150 ppm by weight of cement significantly increased the bond strength of galvanized bars with the largest increase (about 38%) for additions of 35 ppm. The results did indicate that increasing the level of chromate above 35 ppm caused a slight reduction in bond below this level, but this could not be statistically verified. It does support the view, however, that there is no advantage to be gained through increased levels of chromate once a threshold level, which was not determined but is presumably below 35 ppm, is reached. One reason for the lack of significance between the groups may lie in the fact that the tests were conducted on 14-day-old specimens. Cornet and Bresler [79] have pointed out the importance of age and its influence on masking the effects of chromate treatment. Yeomans and Ellis [21] also presented some limited data on the bond strength of deformed black, galvanized and epoxy-coated bars. Their results showed that there was no significant difference between the bond strength
260
O. Kayali
of the three types of deformed reinforcement. It should be noted, however, that all the deformed bar specimens failed by splitting, not pullout. In other work, Kayali and Yeomans [40] obtained results of a pullout test on plain black bars and galvanized bars, which were either untreated, or treated with solutions of sodium dichromate or other chemicals reported to have a passivating effect on zinc, namely sodium molybdate and sodium phosphate. These tests were all conducted after 28 days of fog curing. They reported that plain galvanized bars were 50% superior in bond strength than plain black bars and that sodium dichromate or sodium molybdate treatment prior to casting did not result in any significant difference when compared with untreated galvanized bars. Bars which were treated with sodium molybdate, however, resulted in a 20% reduction in bond strength.
8.7.2.6. Other Protective and Repair Methods The discussion in this chapter has been mainly confined to the bond of reinforcing steel with a corrosion-protective coating. Nevertheless, it should be remembered that any corrosion-protective measure, whether by applying an electric current or a physical barrier, may have an effect on the bond properties, either by virtue of the mechanical changes at the steel surface or the electro-chemical processes taking place at the interface. It is therefore necessary to establish the extent and significance of these possible effects. New methods such as electrochemical chloride removal and re-alkalization have started to gain acceptance in the repair and regular maintenance of reinforced concrete structures. These methods can be applied to elements in which coated reinforcement has already been employed. The effects of such methods on bond with concrete have only recently started to be addressed. Vennesland et al. [103] have used pullout tests to study the bond strength of epoxy-coated bars in concrete subjected to chloride removal. They found a significant effect on bond due to the applied current density as well as the duration of the treatment. With current density between 600 and 5000 Ah/m2, the bond strength suffered severely, while currents higher than 5000 Ah/m2 seemed to enhance bond. Ihekwaba et al. [104] reported alarming degradation of bond strength between high-strength steel and concrete after the application of electrochemical chloride extraction. They attributed this effect to the softening of the alkali ions on the cement silicate hydrates around the steel–concrete interface. In another, albeit limited, investigation, re-alkalization was found to increase the bond strength of smooth black bars because of the densification of the interfacial zone [105]. It is expected that these methods will have effects on the bond of galvanized steel because of the ionic activity taking place at the interfacial zone [77,92].
Bond of Steel in Concrete and the Effect of Galvanizing
261
8.8. Effect of Rib Geometry and the Coating In addition to the influence of the rib-face angle on the mode of failure, the relative area of the rib face with respect to the surface area of the bar between the ribs plays a large role in establishing the mode of failure as well as the bond strength of the bar. This effect was recognized quite early [106,107]. This criterion has been descriptively stated by Darwin and Graham [11] as follows: Rr ¼
projected rib area normal to bar axis nominal bar perimeter £ centre to centre rib spacing
A formula to determine the relative rib area had been given in 1982 [108] and restated more recently [12] in different notations as fr ¼ ðkAe sin fÞ=ðpdb sr Þ þ iah =j
ð18Þ
where fr is the relative rib area, k the number of transverse ribs around the bar perimeter, Ae the area of engagement of the rib above the bar core, f the angle of inclination of the rib to the bar axis, sr the spacing of transverse ribs, i the number of helical ribs (if present), ah is the height of a helical rib, and j is the pitch of a helical rib. Clark [106,107] recommended desirable relative rib areas in the range of 0.17– 0.20 and a minimum value of 0.10. Present practice does not follow these recommendations and in general the value of relative rib area is below the minimum recommended by Clark [11]. Hamad [66] recommended, for improved performance of coated and uncoated bars, that they be manufactured with a rib-face angle of 608, a rib spacing of 50% of db and a rib height of 10% of db ; resulting in a relative rib area of 0.2. Hot-dip galvanizing modifies the rib geometry [100,109]. Koch and Wohlfahrt [100] found that, for zinc-coating thicknesses between 50 and 150 mm, the bond strength was higher in galvanized than non-galvanized bars. Koch and Wohlfahrt also reported that, in spite of the reduction in the relative rib area, the bond strength of galvanized bars did not seem to suffer any reduction when compared with non-galvanized bars. Muller [109] indicated that galvanizing had a negligible effect on the rib height for bars greater than 16 mm diameter, although it is reduced by about 10% in smaller diameter bars. Moreover, the roughness of the galvanized bars was 20– 40% greater than that of the black bar. Muller concluded that, with a galvanized layer not less than 100 mm thick, the effect on bond strength was negligible. It is worth noting here that the minimum required thickness of zinc coating in hot-dip galvanized reinforcement according to ASTM A 767 [110] is 86 mm for a Class II coating and 150 mm for a Class I coating. Although it is not clear which situations apply for Class I or II according to ASTM, it has been suggested that
262
O. Kayali
Class I coatings are intended for non-structural elements while Class II is meant for load-bearing components [62]. The recommendations of the Concrete Institute of Australia [94], however, are a minimum thickness of 84 mm for bar reinforcement in general while a larger thickness is recommended for particularly severe conditions. The bond characteristics of hot-dip galvanized ribbed bars in masonry reinforcement have also been studied. Pullout tests were conducted horizontally on bars embedded in the mortar that binds masonry block walls. The results of Schissl and Schwarzkopf [111] revealed that there was no significant difference in the bond strength between galvanized and non-galvanized ribbed bars. They also found that there was no significant effect of the zinc-coating thickness, to about 350 mm, on the bond behaviour of galvanized bars embedded in mortar. This was despite the fact that galvanizing is known to alter the geometry of ribbed bars due to accumulation of zinc between the ribs [111]. In the case of epoxy coating, it has been found that the relative rib area decreased due to the presence of the coating [112]. However, it has also been reported that, with epoxy coating, the bond of bars larger than 16 mm diameter was relatively insensitive to coating thickness [113]. The tests of Schissl and Schwarzkopf showed that, in the case of galvanized reinforcement, bars as small as 8 mm were insensitive to the coating-thickness variation. It is to be recalled here that Slater et al. [67] observed that, once the full bond strength has developed, there was no significant difference in bond strength between galvanized and non-galvanized ribbed bars. Adding these pieces of evidence to the findings of Kayali and Yeomans [23,40], which reported no significant difference between ribbed galvanized and non-galvanized bars in pullout, beam and beam end tests, it can be reasonably concluded that galvanizing does not adversely affect the bond strength of ribbed reinforcing bars.
8.9. Design Standards and Bond As stated in Section 5 of this chapter, the development length required for deformed bars in tension is given by the American Concrete Institute [24] as ld =db ¼ ð3=40Þ £ ð fy =ð f 0c Þ0:5Þ £ ½abgl=ððc þ Ktr Þ=db Þ
ð16Þ
The ACI code assigns a value of 1.5 to the b factor for epoxy-coated bars or wires with cover less than 3db or clear spacing less than 6db ; while a value of 1.2 is assigned to all other epoxy-coated bars or wires. The value of 1.5 for b is in agreement with the results recently obtained by Kayali and Yeomans [40]. No stipulation is mentioned in the ACI code regarding galvanized reinforcement.
Bond of Steel in Concrete and the Effect of Galvanizing
263
As yet, neither epoxy-coated nor galvanized bars are recognized in the United Kingdom structural design codes. A committee within the European Code du Beton has been working towards establishing a set of design rules for coated bars in concrete [114]. Although galvanized reinforcement is not mentioned in the German code, it has received technical approval. Technical approvals are on the same acceptance level as codes and are given mainly for newly developed products [115]. This approval requires the mean thickness of the coating to be 85 mm and the maximum thickness not to exceed 200 mm. It is interesting that this technical approval permits the same admissible bond stresses to be applied for coated as well as uncoated bars [115]. Australian Standards specification [116] for development length for all deformed bars is given as Lsy:t ¼ ðk1 k2 fsy Ab Þ=½ð2a þ db Þð f 0c Þ0:5 $ 25k1 db
ð19Þ
where k1 is the factor reflecting the effect of bar position and is equal to 1.25 for a horizontal bar with more than 300 mm of concrete cast below the bar or 1.0 for all other bars; k2 the factor reflecting the spacing of the bars, the type of reinforced member and whether the bar is with fitment: this is equal to 1.7 for bars in slabs and walls if the clear distance between bars is not less than 150 mm, 2.2 for longitudinal bars in beams and columns with fitments and 2.4 for any other longitudinal bar; Ab the cross-sectional area of the reinforcing bar; and 2a is twice the cover to the deformed bar or the clear distance between adjacent parallel bars developing stress, whichever is less. It is notable that coated reinforcement is not mentioned in this Standard. However, the Concrete Institute of Australia, in its Practice Note 17 [95], has stated that the bond of concrete to deformed reinforcement should not suffer any significant loss provided the zinc –alkali reaction is prevented by appropriate passivation.
8.10. Concluding Remarks The evidence presented in this chapter points to the highly probable conclusion that the galvanizing of ribbed steel reinforcing bars has no adverse effect on bond strength and, if anything, would result in an improvement of bond. To date, however, this effect has not been shown to be statistically significant. The overall view from the results of the available research to date indicates that the practice of chromating the reinforcement, whether by applying it directly to the steel surfaces or to the concrete mixing water, is not at all necessary. Abandoning this practice would also conform to health and environmental requirements in
264
O. Kayali
many countries, as well as making the handling and use of galvanized steel much easier and possibly cheaper. Concerning the effects of remedial treatments to the concrete, preliminary data suggest that the electrochemical extraction method, for example, can severely reduce bond strength. Resorting to such new and sometimes very sophisticated, disruptive and probably costly methods may be avoided altogether if appropriate protection is applied to the reinforcement in the first instance. Repair and protection methods which are primarily aimed at dealing with the problem of corrosion of reinforcement must first be carefully assessed from the point of view of bond performance. Failure in bond, even if it were temporary, could result in disastrous consequences. Current and future developments in the concrete industry are becoming more oriented towards performance and durability, even though very high-strength concrete is available. The use of galvanized steel reinforcement in this context is quite appropriate. It is to be noted, however, that there is a large gap in many of the national design standards for reinforced concrete construction with regard to specifications on coated reinforcement in general and galvanized reinforcement in particular. This gap seems to parallel that on high-strength concrete. The author believes that these gaps have to be addressed in the context of the design for highperformance concrete structures.
References [1] Goto, Y. (1971). Cracks formed in concrete around deformed tension bars. ACI Journal, 68, 4, 244–251. [2] Park, R., & Paulay, T. (1975). Reinforced concrete structures. Wiley, New York. [3] ASTM C234. (1993). Standard test method for comparing concretes on the basis of the bond developed with reinforcing steel. American Society for Testing and Materials, Philadelphia, PA, USA. [4] Lutz, L. A., & Gergely, P. (1967). Mechanics of bond and slip of deformed bars in concrete. ACI Journal, 64, 711–721. [5] Rehm, G. (1958). The fundamental law of bond, Proceedings of Symposium on Bond and Crack Formation in Reinforced Concrete, Stockholm 1957, RILEM, Paris. Tekniska Hogskolans Rataprinttrychkeri, Stockholm, Sweden. [6] Tepfers, R. (1979). Cracking of concrete cover along anchored deformed bars. Magazine of Concrete Research, 31, 106, 3–12. [7] Tepfers, R. (1980). Bond stress along lapped reinforcing bars. Magazine of Concrete Research, 32, 112, 135–142. [8] Orangon, C. O., Jirsa, J. O., & Breen, J. E. (1977). Re-evaluation of test data on development length and splices. ACI Journal, 74, 3, 114–123. [9] Hansen, E. A., & Thorenfeldt, E. (1966). Bond properties of deformed reinforcement bars in high strength concrete, Fourth International Symposium on the Utilization of
Bond of Steel in Concrete and the Effect of Galvanizing
[10] [11] [12]
[13] [14]
[15] [16] [17] [18] [19] [20] [21]
[22] [23] [24] [25] [26] [27] [28] [29] [30]
265
High Strength/High Performance Concrete, May 1996, Laboratoire Central des Ponts et Chaussees, Proceedings, Vol. 3, pp. 1105-1114. Rehm, G. (1968). The basic principles of the bond between steel and concrete, Translation No. 134. Cement and Concrete Association, London, 66 pp. Darwin, D., & Graham, E. (1993). Effect of deformation height and spacing on bond strength of reinforcing bars. ACI Structural Journal, 90, 6, 646–665. Cairns, J., & Jones, K. (1995). Influence of rib geometry on strength of lapped joints: an experimental and analytical study. Magazine of Concrete Research, 47, 172, 253– 262. Soroushian, P., Mirza, M., & Alhozaimy, A. (1994). Bonding of confined steel fibre reinforced concrete to deformed bars. ACI Materials Journal, 91, 141–149. Hadje-Ghaffari, H., Choi, O. C., Darwin, D., & McCabe, S. L. (1994). Bond of epoxycoated reinforcement: cover, casting position, slump, and consolidation. ACI Structural Journal, 91, 1, 59– 68. Yeomans, S. R., & Kayali, O. A. (1998). Unpublished research data. Ferguson, P. (1981). Reinforced concrete fundamentals. Wiley, New York, 4th ed.. Losberg, A., & Olsson, P. A. (1979). Bond failure of deformed reinforcing bars based on the longitudinal splitting effect of the bars. ACI Journal, 76, 1, 5–17. Bazant, Z. P., & Sener, S. (1988). Size effect in pullout tests. ACI Materials Journal, 85, 5, 347–351. Abrishami, H. H., & Mitchell, D. (1992). Simulation of uniform bond stress. ACI Materials Journal, 89, 2, 161–168. Windisch, A. (1985). A modified pull-out test and new evaluation methods for a more real local bond-slip relationship. Materials and Structures, 18, 105, 181–184. Yeomans, S. R., & Ellis, D. R. (1992). Further studies of the bond strength and slip characteristics of galvanized and epoxy-coated steel reinforcement in concrete, Progress Report No. 5, ILZRO Project ZE-341. International Lead Zinc Research Organisation, Research Triangle Park, NC, USA. Mathey, R. G., & Watstein, D. (1961). Investigation of bond in beam and pullout specimens with high yield strength deformed bars. ACI Journal, 57, 9, 1071–1090. Kayali, O., & Yeomans, S. R. (1995). Bond and slip of coated reinforcement in concrete. Construction and Building Materials, 9, 4, 219–226. American Concrete Institute. (1995). ACI Building Code Requirements for Structural Concrete, ACI 318. American Concrete Institute, Farmington Hills, MI, USA. Ferguson, P. M., Breen, J. E., & Thompson, J. N. (1965). Pullout tests on high strength reinforcing bars. ACI Journal, 62, 8, 933–950. Chana, P. S. (1990). A test method to establish realistic bond stresses. Magazine of Concrete Research, 42, 151, 83– 90. Clarke, J. L., & Birjandi, F. K. (1993). Bond strength tests for ribbed bars in lightweight aggregate concrete. Magazine of Concrete Research, 45, 163, 79 –87. RILEM. (1970). Bond test for reinforcement steel: 1 — beam test, materials and structures, 3, 15, 169–174. Treece, R. A., & Jirsa, J. O. (1989). Bond strength of epoxy-coated reinforcing bars. ACI Materials Journal, 86, 2, 167–174. Choi, O. C., Hadje-Ghaffari, D., & McCabe, S. (1991). Bond of epoxy-coated reinforcement: bar parameters. ACI Materials Journal, 88, 2, 207–217.
266
O. Kayali
[31] Cleary, B. D., & Ramirez, J. A. (1991). Bond strength of epoxy-coated reinforcement. ACI Materials Journal, 88, 2, 146–149. [32] Darwin, D., McCane, S., Idun, E., & Schoenekase, S. P. (1992). Development length criteria: bars not confined by transverse reinforcement. ACI Structural Journal, 89, 6, 709–720. [33] de Lallard, F., Schaller, I., & Fuchs, J. (1993). Effect of bar diameter on the bond strength of passive reinforcement in high-performance concrete. ACI Materials Journal, 90, 4, 333–339. [34] Hamad, B. S., & Jirsa, J. O. (1993). Strength of epoxy-coated reinforcing bar splices confined with transverse reinforcement. ACI Structural Journal, 90, 1, 77 –88. [35] Hester, C. J., Salamizavaregh, S., Darwin, D., & McCabe, S. (1993). Bond of epoxycoated reinforcement: splices. ACI Structural Journal, 90, 1, 89 –102. [36] Azizinamini, A., Stark, M., Roller, J. J., & Ghosh, S. K. (1993). Bond performance of reinforcing bars embedded in high-strength concrete. ACI Structural Journal, 90, 5, 554–561. [37] Benmokrane, B., & Chaallal, O. (1996). Bond strength and load distribution of composite GFRP reinforcing bars in concrete. ACI Materials Journal, 93, 3, 246–253. [38] Hamza, A. M., & Naaman, A. E. (1996). Bond characteristics of deformed reinforcing steel bars embedded in sifcon. ACI Materials Journal, 93, 6, 578–588. [39] ASTM A 944. (1995). Standard test method for comparing bond strength of steel reinforcing bars to concrete using beam-end specimens. American Society for Testing and Materials, Philadelphia, PA, USA. [40] Kayali, O., & Yeomans, S. R. (2000). Bond of galvanized steel in concrete. Cement and Concrete Composites, 22, 6, 459–467. [41] Nielsen, C. V., Olesen, J. F., & Arup, B. K. (1996). Effect of fibres on the bond strength of high-strength concrete, Fourth International Symposium on the Utilization of High Strength/High Performance Concrete, May 1996. Laboratoire Central des Ponts et Chaussee´s, Paris, France. [42] Richard, P. (1996). Reactive powder concrete: a new ultra high strength cementitious material, Fourth International Symposium on the Utilization of High Strength/High Performance Concrete, May 1996. Laboratoire Central des Ponts et Chaussee´s, Paris, France. [43] Taerwe, L. (1996). Structural behaviour of HSC/HPC, Fourth International Symposium on the Utilization of High Strength/High Performance Concrete, May 1996. Laboratoire Central des Ponts et Chaussee´s, Paris, France, pp. 57–66. [44] Neville, A. M. (1995). Properties of concrete. Longman, London, 4th ed., p. 311. [45] Gjorv, O. E., Monteiro, P. J. M., & Mehta, P. K. (1990). Effect of condensed silica fume on the steel-concrete bond. ACI Materials Journal, 87, 6, 573–580. [46] Neville, A. M., & Brooks, J. J. (1987). Concrete technology. Longman Scientific and Technical, London, p. 20. [47] Price, W. H. (1951). Factors influencing concrete strength. Journal of the ACI, 47, 417–432. [48] de Almeida, I. R. (1996). Bond between reinforcing steel and high strength concrete, Fourth International Symposium on the Utilization of High Strength/High Performance Concrete, May 1996. Laboratoire Central des Ponts et Chaussee´s, Paris, France, pp. 1097–1104.
Bond of Steel in Concrete and the Effect of Galvanizing
267
[49] Hwang, S.-J., Lee, Y.-Y., & Lee, C.-S. (1994). Effect of silica fume on the splice strength of deformed bars of high performance concrete. ACI Structural Journal, 91, 294–302. [50] Khalaf, M. N. A., & Page, C. L. (1979). Steel mortar interfaces: microstructural features and mode of failure. Cement and Concrete Research, 9, 2, 197–208. [51] American Concrete Institute. (1987). Report of ACI Committee 213R, Guide for Structural Lightweight Aggregate Concrete, 27 p. [52] American Concrete Institute. (1993). Guide for structural lightweight aggregate concrete, ACI Manual of Concrete Practice, 1993, Part 1: Materials and General Properties of Concrete. [53] British Standards, BS 8110 (1985). Structural use of concrete: Part 2: code of practice for special circumstances. British Standards Institution, London. [54] Rossetti, V. A., Curcio, F., Galeota, D., & Giammatteo, M. M. (1995). Structural properties of lightweight concretes. Concrete 95 — International Conference, Concrete Institute of Australia and Federation Internationale de la Precontrainte. Conference Papers, vol. 1, pp. 187– 193. [55] Mor, A. (1992). Steel-concrete bond in high-strength lightweight concrete. ACI Materials Journal, 89, 76–82. [56] Goldman, A., & Bentur, A. (1989). Bond effects in high-strength silica-fume concretes. ACI Materials Journal, 86, 5, 440– 447. [57] Ezeldin, S., & Balaguru, P. N. (1989). Bond behaviour of normal and high strength fibre reinforced concrete. ACI Materials Journal, 86, 515–524. [58] Yeomans, S. R. (1995). Coated steel reinforcement for corrosion protection in concrete. Transactions, Hong Kong Institution of Engineers, 2, 2, 17– 28. [59] Robinson, R. C. (1972). Design of reinforced concrete structures for corrosive environments. Materials Performance, 11, 15 p. [60] Mathey, R. G., & Clifton, J. R. (1976). Bond of coated reinforcing bars in concrete. Journal of the Structural Division, ASCE, 102, ST1, 215–229. [61] ASTM A 775. (1991). Epoxy-coated reinforcing steel bars. American Society for Testing and Materials, Philadelphia, PA, USA. [62] Yeomans, S. R. (1993). Considerations of the characteristics and use of coated steel reinforcement in concrete, NISTIR 5211. Building and Fire Research Laboratory, United States Department of Commerce, National Institute of Standards and Technology, Gaithersburg, MD, USA, 41 p. [63] Hamad, B. S., Jirsa, J. O., & D’Abreu, N. I. (1993). Anchorage strength of epoxy-coated hooked bars. ACI Structural Journal, 90, 2, 210–217. [64] Swamy, R. N. (1992). Durability of rebars in concrete. Durability of Concrete, Proceedings of the GM Idorn International Symposium, ACI Publication SP-131, pp. 67 –98. [65] Cairns, J., & Abdullah, R. B. (1996). Bond of black and epoxy-coated reinforcement — a theoretical approach. ACI Materials Journal, 93, 4, 362–369. [66] Hamad, B. S. (1995). Comparative bond strength of coated and uncoated bars with different rib geometries. ACI Materials Journal, 92, 6, 579–590. [67] Slater, W. A., Richart, F. E., & Scofield, G. G. (1920). Tests on bond resistance between concrete and steel, Technical Paper No. 173. US National Bureau of Standards, pp. 9–33. [68] Robinson, K. E. (1956). The bond strength of galvanized reinforcement, Cement and Concrete Association Technical Report, TRA/220. Cement and Concrete Association, London, 7 p.
268
O. Kayali
[69] Bird, C. E. (1962). Bond of galvanized steel reinforcement in concrete. Nature, 194, 798 p. [70] Porter, F. C. (1994). Corrosion resistance of zinc and zinc alloys. Marcel Dekker, New York. [71] Yeomans, S. R. (1995). Galvanized steel reinforcement — a perspective view, Real World Concrete, Fifth CANMET/ACI International Conference on Fly Ash, Silica Fume, Slag and Natural Pozzolans in Concrete, Milwaukee, WI, USA. [72] Sanchez, G. (1994). Procede en continu de protection des aciers par revetement me´tallique. Annales de l’ITBTP, 523, 93 –154. [73] Galvanized Reinforcement for Concrete — II (1981). International Lead Zinc Research Organisation, Research Triangle Park, NC, USA, 280 p. [74] Koch, R., & Wohlfahrt, R. (1988). Effect of admixtures in concrete on the bond behaviour of galvanized reinforcing bars. Betonwerk þ Fertigteil-Technik, 54, 3. [75] Arliguie, G., & Longuet, P. (1979). Comportement du zinc dans les pates de ciment, ciments, betons. Platres Chaux, 79, 4, 201– 206. [76] Blanco, M. T., Andrade´, C., & Macias, M. (1984). SEM study of the corrosion products of galvanized reinforcements immersed in solutions in the pH range 12.6 to 13.6. British Corrosion Journal, 19, 1, 41 –48. [77] Fratesi, R., Moriconi, G., & Coppola, L. (1996). The influence of steel galvanisation on rebar behaviour in concrete. In: C. L. Page, P. B. Bamforth, & J. W. Figg (Eds), Corrosion of reinforcement in concrete construction (pp. 631–641). The Royal Society of Chemistry, UK. [78] Porter, F. C. (1991). Zinc handbook: properties, processing and use in design. Marcel Dekker, New York, 106 p. [79] Cornet, I., & Bresler, B. (1981). Galvanized steel in concrete: literature review and assessment of performance, Galvanized Reinforcement for Concrete — II. International Lead Zinc Research Organisation, Research Triangle Park, NC, USA, pp. 1–56. [80] Andrade, C., Arteaga, A., Hombrados, C. L., & Vazquez, A. (1997). Long-term results of bond of galvanized rebar and concrete submerged in natural sea water, private communication. [81] Belaı¨d, F., Arliguie, G., & Fanc¸ois, R. (2001). Effect of bar properties on bond strength of galvanized reinforcement. Journal of Materials in Civil Engineering, 13, 6, 454–458. [82] Arliguie, G., Grandet, J., & Ollivier, J. P. (1985). Orientation de la portlandite dans les mortiers et beton de ciment portland: Influence de la nature et de l’etat de surface du support de cristallisation. Materials and Structures, 18, 106, 263–267. [83] Tashiro, C., & Ueoka, K. (1981). Bond strength between C3A paste and iron, copper or zinc wire and microstructure of interface. Cement and Concrete Research, 11, 4, 619–624. [84] Mehta, P. K., & Monteiro, P. J. M. (1993). Concrete: structure, properties and materials. Prentice Hall, NJ, USA. [85] Mindess, S., & Young, J. F. (1981). Concrete. Prentice Hall, NJ, USA. [86] Page, C. L. (1975). Mechanism of corrosion protection in reinforced concrete marine structures. Nature, 258, 5535, 514–515. [87] Ramachandran, V. S., Feldman, R. F., & Beaudoin, J. J. (1981). Concrete science, Treatise on Current Research. Heyden, London. [88] Hofsoy, A., & Gukild, I. (1969). Bond studies on hot dip galvanized reinforcement in concrete. ACI Journal, 66, 174–184.
Bond of Steel in Concrete and the Effect of Galvanizing
269
[89] Arliguie, G., Ollivier, J. P., & Grandet, J. (1982). Etude de l’e´ffet retardateur du zinc sur l’hydration de la pate de ciment portland. Cement and Concrete Research, 12, 1, 79–86. [90] Tashiro, C. (1980). The effects of several heavy metal oxides on the hydration and microstructure of hardened mortar of C3S, Seventh EME Congress, International de la Chimie des Ciments, Paris, France. [91] Belaı¨d, F., Arliguie, G., & Franc¸ois, R. (2001). Porous structure of the ITZ around galvanized and ordinary steel reinforcements. Cement and Concrete Research, 31, 11, 1561 –1566. [92] Yeomans, S. R. (1997). Understanding the corrosion process for galvanized steel reinforcement in concrete, International Conference on Understanding Corrosion Mechanisms of Metals in Concrete. MIT, Boston, MA, USA, 7 p. [93] Hamad, B. S., & Mike, J. A. (2003). Experimental investigation of bond strength of hotdip galvanized reinforcement in normal- and high-strength concrete. ACI Structural Journal, 100, 4, 465–470. [94] Concrete Institute of Australia. (1984). The use of galvanized reinforcement in concrete, Current Practice Note 1. Concrete Institute of Australia, Sydney, 4 p. [95] Lewis, D. A. (1962). Some aspects of the corrosion of steel in concrete. CSIR, National Building Research Institute, RD, 24, XIII, 1, 547–555. [96] Roberts, A. W., Scott, O. J., & Leung, H. K. (1978). Bond characteristics of concrete reinforcing tendons coated with zinc, ILZRO Project ZE 222, Final report. International Lead Zinc Research Organization, Research Triangle Park, NC, USA. [97] Hamilton, J. W., & Wetterhahn, K. E. (1988).In: H. G. Seiler, & H. Sigel (Eds), Chromium: handbook on toxicity of inorganic compounds. Marcel Dekker, New York. [98] Ellis, V., & Freeman, S. (1986). Dermatitis due to chromate in cement: Part 1 — chromate content of cement in Australia. Australasian Journal of Dermatology, 27, 86, 86 –90. [99] Fregert, S., & Gruvberger, B. (1972).Chemical Properties of Cement, 20 –238. [100] Koch, R., & Wohlfahrt, R. W. (1989). Galvanized bars- rib geometry and bonding behaviour of hot-dip galvanized reinforcing steel bars. Concrete Precasting Plant and Technology (Betonwerk þ Fertigteil 2 Technik), 55, 2, 52 –57. [101] Sakamoto, N., & Iwasaki, N. (1982). Influence of sodium chloride on the concrete/steel and galvanized steel bond, Proceedings of the International Conference on Bond in Concrete, Paisley College of Technology, Scotland. Applied Science, Barking. [102] Nishi, T. (1974). Investigation on mechanical behaviour of galvanized steel reinforcement in concrete, ILZRO Project No. ZE-170, Final Report. International Lead Zinc Research Organization, Research Triangle Park, NC, USA. [103] Vennesland, O., Humstad, E. P., Gautefall, O., & Nustad, G. (1996). Electrochemical removal of chlorides from concrete — effect on bond strength and removal efficiency. In: C. L. Page, P. B. Bamforth, & J. W. Figg (Eds), Corrosion of reinforcement in concrete construction (pp. 448–455). The Royal Society of Chemistry, UK. [104] Ihekwaba, N. M., Hope, B. B., & Hansson, C. M. (1996). Pull-out and bond degradation of steel rebars in ECE concrete. Cement and Concrete Research, 26, 2, 267–282. [105] Al-Kadhimi, T. K. H., Banfill, P. F. G., Millard, S. G., & Bungey, J. H. (1996). An experimental investigation into the effects of electrochemical re-alkalisation on concrete. In: C. L. Page, P. B. Bamforth, & J. W. Figg (Eds), Corrosion of reinforcement in concrete construction (pp. 501– 511). The Royal Society of Chemistry, UK. [106] Clark, A. P. (1946). Comparative bond efficiency of deformed concrete reinforcing bars. ACI Journal, 43, 4, 381–400.
270
O. Kayali
[107] Clark, A. P. (1949). Bond of concrete reinforcing bars. ACI Journal, 46, 3, 161–184. [108] CEB Bulletin No 151 (1982). Bond action and behaviour of reinforcement. CEB, Paris. [109] Muller, H. H. (1993). Behaviour of galvanized rebars in concrete. In: S. Nagataki, T. Nireki, & F. Tomosawa (Eds), Durability of building materials and components (pp. 147–156). E & FN Spon/Chapman & Hall, London. [110] ASTM A767. (1990). Zinc-coated reinforcing steel bars. American Society for Testing and Materials, Philadelphia, PA, USA. [111] Schiessl, P., & Schwarzkopf, U. (1985). Bond behaviour of hot-dip galvanized ribbed reinforcing bars in masonry. Betonwerk þ Fertigteil 2 Technik, 51, 11, 735–740. [112] Cleary, D. B., & Ramirez, J. A. (1993). Epoxy-coated reinforcement under repeated loading. ACI Structural Journal, 90, 4, 451–458. [113] Cusens, A. R., & Yu, Z. (1992). Pullout tests of epoxy-coated reinforcement in concrete. Cement and Concrete Composites, 14, 269–276. [114] Cairns, J. (1997). Private communication. [115] Schiessl, P. (1998). Private communication. [116] AS 3600-2001, Concrete structures code. Standards Australia, Sydney, Australia.
Galvanized Steel Reinforcement in Concrete S.R. Yeomans (Editor) q 2004 Elsevier Ltd. All rights reserved.
Chapter 9
Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective John P. Broomfield Broomfield Consultants, UK
9.1. A Brief History It was as early as 1742 that Malouin, a French chemist, described a method of coating iron by hot dipping in zinc. It is not known if this was for the corrosion protection of iron or not, because this occurred some 40 years before the Italian biologist Galvani set frog legs twitching in the 1780s. Despite giving his name to galvanizing and the galvanic effect, Galvani little knew what train of events would follow and it was left to Volta to develop the first battery of alternating zinc and copper plates separated by acid impregnated cloths, independent of frog legs or any other animal matter. It was from this work that the ability of zinc to protect metals such as copper, iron and steel from corrosion was developed, particularly by Humphrey Davy, who invented galvanic cathodic protection in 1824. British and French patents for the cleaning and dipping of iron in molten zinc as a method of protection against corrosion were granted in 1837 and, by 1850, the British galvanizing industry was using 10,000 ton of zinc a year. In general operation, the galvanizing process has changed little since this time. Today, the galvanizing industry is recognized as a major sector in the structural steel fabrication field and galvanizing is extensively used to protect a vast array of steel structures in a wide range of environments. A thorough treatment of the operation of the wellknown galvanizing processes and description of the characteristics of galvanized coatings has been given in Chapter 4 and does not need to be discussed further here. Concrete has a considerably longer history than galvanizing. A concrete floor has been found dating back to 5600 BC and a form of concrete was used by the ancient Egyptians in the Great Pyramid at Giza in about 2500 BC. Lightweight concrete was used by the Romans in some of the arches of the Coliseum and, perhaps most impressively, in the dome of the Pantheon, built in 127 AD, which is
272
J.P. Broomfield
still standing and used as a place of worship in Rome even today. The Romans tried to reinforce their concretes with bronze, but the difference in thermal expansion lead to cracking and spalling. It is said that the bronze beams from the Parthenon were melted down and used to make the Papal throne in St Peter’s Basilica. It was after the invention of Portland cement by Joseph Aspdin, which was patented in 1824, that serious efforts were made to reinforce concrete with wires and rods of iron and steel. By the early 1900s, reinforced concrete structures were starting to appear. By the middle of the twentieth century, corrosion of reinforcement was being reported in the literature. Initially, it was put down to stray-current-induced corrosion from the earth return system of electric trams or trolleys. It was then discovered in the USA, however, that some bridges were suffering from reinforcement corrosion in mountainous regions of California where there was no source of DC current but the roads were being deiced with salt during the winter months. The problem was first noticed in the USA because, unlike most other western countries, the USA and Canada do not routinely put waterproofing membranes and asphalt wearing courses on their major highway bridge decks. As the reinforced concrete infrastructure grew older, reinforcement corrosion problems emerged more widely. Some problems were put down to poor construction, some to a misunderstanding of the environment and some to a misunderstanding of the chemical properties of concrete. The clearest examples of the latter are the use of seawater for mixing in concrete and the addition of calcium chloride as a set accelerator for cold winter concreting where it was assumed that the chlorides would be bound chemically during the setting process. However, experience has shown the binding process to be less than permanent and that corrosion frequently initiates, sometimes decades after construction. As previously noted in both Chapters 1 and 2, concrete is one of the most widely used man-made materials and damage due to reinforcement corrosion is one of the biggest durability issues currently being addressed in our steadily ageing infrastructure. The popularity of concrete is due to the fact that it is made from widely available materials that are comparatively easy to transport. Reinforced concrete is also highly versatile as the concrete can be molded to any shape. The concrete is strong in compression, while the steel is strong in tension. By good fortune, concrete and steel are extremely compatible materials; they have similar coefficients of thermal expansion (unlike the bronzes used by the Romans) and the hydration process in Portland cements shrinks the hardening concrete onto the steel where the high alkalinity in the cement paste forms a durable, selfmaintaining “passive” oxide layer on the steel surface that is highly resistant to corrosion. The passive layer is a dense, impermeable oxide which is stable,
Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective
273
growing on a logarithmic scale. It is stable roughly between pH 11.5 and 13.5, which comfortably fits with the pH of Portland cement-based hydraulic concretes.
9.2. The Corrosion of Steel in Concrete and Protection Systems One of the interesting properties of concrete is that, despite the porosity created by the escaping excess mix water during the hydration process, few substances can penetrate it without producing obvious damage to the concrete itself. Two of the substances that can penetrate concrete are oxygen and water. These substances are required to fuel the corrosion process on reinforcing steel. However, in the presence of the high pH (alkaline) pore water, the passive layer is formed instead of the high volume, hydrated ferric oxide that can crack and spall concrete. The dense, impenetrable passive oxide layer is self-repairing and could be thought of as the (corrosion) engineer’s dream protection system. Unfortunately, two other substances can penetrate the concrete as well as oxygen and water. One is the chloride ion and the other is the carbon dioxide molecule. Both of these substances lead to the break down of the passive layer and the subsequent corrosion of the steel. As noted through this book, reinforcement corrosion is a worldwide, multibillion dollar phenomenon. Some of the costs are discussed in more detail in the recent US study on the costs of corrosion [1]. In Chapter 1, Yeomans summarizes the corrosion problem of steel in concrete and points to the coating of reinforcing steel as a way of mitigating this huge worldwide problem. As stated earlier, one single problem has done more to drive research focused on the corrosion of steel in concrete than any other factor. This was the observation in the 1950s and 1960s that deicing salt ingress was causing corrosion damage of bare concrete highway bridges in the northern and mountainous regions of the USA. This observation, and the need to avoid problems inherent in the use of deicing salts, led to a significant number of intensive field and laboratory studies on how to measure, control and prevent the problem of reinforcement corrosion in concrete. Several of these studies and others specifically related to an understanding of the behavior of galvanized steel in concrete, are discussed by Yeomans in Chapter 6 and others elsewhere in this book. In July 1982, a committee empanelled by the Federal Highways Administration (FHWA), produced a three page panel report entitled Galvanized Rebar as a Long-Term Protective System [2]. Details of this report are given in Chapter 6 but, in summary, it concluded that “the service life of a bridge deck that is regularly salted during the winter months is approximately 10 to 15 years
274
J.P. Broomfield
before corrosion-induced distress develops. The use of galvanized reinforcement may extend the service life for a short period but by no more than 5 years.” There were a number of caveats in the panel report (refer Chapter 6) and the authors freely admitted that data available at the time were limited, contradictory, difficult to compare and inconclusive. However, they were charged with reaching a conclusion and did so on the evidence to hand in mid1982 and using their considerable collective experience. These conclusions contributed to the growth and development of fusionbonded epoxy-coated reinforcement (FBECR) as the protective system of choice for bridge deck protection during construction in the USA and Canada until around 1990. It was believed that the epoxy coating was extremely durable and non-sacrificial. Further, adhesion between the epoxy and steel was excellent and there was no undercutting of the coating at defects in it. Exposure tests suggested that bars with significant coating damage still gave decades of life extension compared to either black steel or galvanized steel [3]. However, in 1990, the Florida Department of Transportation investigated cracking and spalling of concrete in precast concrete piles and other elements of the Florida Keys bridges. Investigation showed that corrosion damage started in 1986 on a bridge built in 1981, with two other bridges built in 1982 and 1986 showing spalls in 1988 [4]. All the corroding reinforcing steel in these structures was fusion-bonded epoxy-coated. This land-mark investigation led to a major review of the performance of FBECR in the US and Canada. Many State Departments of Transportation found that their bridge decks were performing well. However, Florida continued to find problems on its substructures in the splash zone. Other states identified problems with the adhesion and damage level of the coating but few found active corrosion damage, with the notable exception of Ontario Ministry of Transportation. The situation in North America at this time was summarized in an excellent paper by Manning [4]. As well as initiating a new round of research projects on FBECR, the performance of alternative methods of enhancing corrosion resistance was reexamined and the possibility of new developments was heavily researched and promoted by interested parties. Alternative dielectric coatings, metallic coatings, stainless steel and stainless steel clad reinforcement were all researched, promoted and trialed [3], along with admixed corrosion inhibitors for concrete. Side by side trials of stainless steel, galvanized, epoxy-coated and plain steel reinforced concrete bridge decks were installed by the several US Departments of Transportation [5]. Reports of these trial installations and similar field investigations using galvanized reinforcement are detailed in Chapter 6.
Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective
275
At the present time (March 2004), FBECR is still widely used in North American bridge decks. It is also used in the Middle East. There are, however, no coating plants actively producing FBECR in Europe or Australasia where a wide range of systems are used to enhance the durability of reinforced concrete, particularly in severe chloride-exposure conditions.
9.3. Design for Durability When a reinforced concrete structure is designed, the issue of durability, life-cycle costing and overall life is frequently relegated to following the relevant codes and choosing the lowest bid from the tendering contractors. At present, most national design codes take a “deemed to satisfy” approach, whereby decisions on materials and fabrication details are determined from a table in which the aggressiveness of the proposed environment is compared to the required concrete compression strength and a minimum recommended cover to the reinforcement. For example, BS5328 requires that, for a marine exposure, a C50 (50 MPa) concrete with a minimum 0.45 water/cement ratio and cover to the reinforcement of 50 mm is required. A review of durability issues for reinforced concrete construction, dealing with these over-riding principles and issues related to design for galvanized reinforced concrete, has been presented by Swamy in Chapter 2 of this book. However, the technical literature is now increasingly demonstrating that, in aggressive environments, ordinary Portland cement concretes are not adequate in preventing the ingress of chlorides sufficient to initiate corrosion of the reinforcement for a reasonable life of the structure. This is particularly true for the splash and tidal zones of marine-exposed reinforced concrete structures and for multi-storey car parks in cold climate regions where deicing salts are widely used on road surfaces. With the range of developments in concrete technology and the availability of materials for use as reinforcement, when considering the durability of a structure, particularly when exposed to high levels of chloride, the designer has a number of options. These range from strategies to keep the aggressive species out of the concrete itself to slowing their progress through the concrete cover to the depth of the reinforcement. Alternatively, or in addition, the designer may choose to modify the surface of plain steel reinforcement or use a corrosion-resistant metal as an alternative. The following are some of the major methods of corrosion prevention for steel in concrete currently available: * *
increase in concrete cover to depth of reinforcement; control of concrete quality with low water-to-cementitious (w/c) ratio and high cement content;
276 *
* * *
* *
*
*
*
J.P. Broomfield
use of alternative cementitious materials such as ground granulated blast furnace slag and microsilica fly ash; use of controlled permeability formwork; incorporation of fiber admixtures; use of additives to the concrete mix such as corrosion inhibitors or water-proofing products; use of coatings and sealers to the concrete; use of membranes and impermeable barriers to prevent ingress of aggressive species; application of cathodic protection systems (both sacrificial and impressed current); use of alternative reinforcement including fiber-reinforced polymers, stainless steel bar (both solid or clad) and other corrosion-resistant metals; and use of coated steel, including epoxy-coated reinforcement and galvanized reinforcement.
This list is not exhaustive and variants of many of these protection systems are being developed. It is also to be noted that in most, if not all, reinforced concrete construction, combinations of many of these systems are routinely used. A summary of the major strategies associated with the use of the more common and widespread of these corrosion-protection systems, as compiled by this author, is given in Table 1. Included in this is an analysis of issues related to the use of galvanized reinforcement, the underlying principles for this largely being gleaned from the data presented in the various chapters of this book. Also of benefit and guidance to the designer, life-cycle costing models have been developed. These enable the designer, or other user, to select the environment that the particular structure or portion of a structure is exposed to and calculate the lifecycle cost and repair options when using any of the above strategies to achieve a given life. Repair and maintenance cycles are also given. As might be expected, the accuracy of such models depends on the reliability of the input data. They frequently require the extrapolation of short-term data to estimate long lives, limiting their absolute accuracy. However, they are useful comparative tools, giving relative changes in life for different combinations of strategies. One example has been developed by ACI Committee 365 and is available via the web [6]. It is interesting to note that zinc, in the form of galvanic anodes, is increasingly being used to address corrosion problems in traditionally reinforced existing structures containing chloride-contaminated concrete. Rather than zinc-coated bar, however, these are sheets, meshes or thermal-sprayed coatings of zinc applied to the concrete surface or blocks of zinc or zinc wire embedded in recesses cut into the concrete. They are designed to corrode sacrificially and protect the steel galvanically. For new structures where galvanized reinforcement is to be used, it is
Table 1: Methods for improving the durability of reinforced concrete (compiled by Broomfield). How it works
Advantages
Limitations
Where used
Increase cover over reinforcement
Increases time for chlorides or carbonation front to reach steel
Low cost
Requires very good quality assurance Increased risk of cracking giving Cl2 and CO2 more rather than less rapid ingress
Everywhere
Reduces concrete porosity to increase time for corrodents to reach steel
Low cost
Makes concrete less workable and more prone to poor consolidation Higher cement content increases heat of hydration
Everywhere
Makes concrete less workable and more prone to poor consolidation Lower cement content decreases heat of hydration
Marine structures
Decrease water/ cementitious ratio
Included in Codes
Included in Codes
Ground granulated blast furnace slag, microsilica, fly ash
Reduces concrete porosity to increase time for corrodents to reach steel
Modest cost
Included in Codes to achieve durable concrete in marine exposure
(continued)
Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective
Method
277
278
Table 1: Continued. How it works
Advantages
Limitations
Where used
Corrosion inhibitors
Allows formation of protective chemical layer on steel
Modest cost
Marine structures and car parks
Can be used with other methods
Needs to work after 20 years or more. Some can be washed out in marine environment Dosage must reflect exposure
Coatings and penetrating sealers
Increases time for chlorides or carbonation to reach steel
Low cost Ease of maintenance
Changes appearance Requires maintenance
Bridges decks and car parks
Membranes and barriers
Increases time for chlorides or carbonation to reach steel
Ideal for bridge decks and similar structures
Requires maintenance Can hide problems Changes profile and dead load
Bridge decks and car parks
Fiber reinforced polymer reinforcement
Corrosion resistant reinforcement
Theoretically very long life Easy handling Few if any codes exist
Needs new design codes Unknown deterioration rates and mechanisms over 50 year life or more
Experimental
Stainless steel reinforcement
Corrosion resistant reinforcement
Very long life if correct grade used Scrap (recycle) value
High first cost
High value long life structures or elements
J.P. Broomfield
Method
Makes steel cathodic and therefore corrosion resistant
Theoretically up to 100-year life of anodes
High first cost, high maintenance, technically complex
High value high chloride exposed structures mainly in Arabian Gulf
Epoxy coated reinforcement
Coating prevents chloride access to steel surface
Easy to use Low cost in North America
Durability problems in some highway structures in Florida and Ontario Handling damage during fabrication and placement
Bridge decks etc. in North America and Arabian Gulf
Galvanized reinforcement
Higher chloride tolerance and resistance to carbonation delays onset of corrosion
Widely available, ease of handling and installation
Moderate cost premium
Europe, UK, North America, Central America, India, Japan, Australasia, Hong Kong, Singapore, Taiwan
Extends life during dissolution of coating
Tolerates variations in concrete
Provides sacrificial protection
Design and use as per black steel
Performance advantage reduced in poor quality concrete and severe-to-extreme chloride exposure Risk of galvanic coupling if connected to black steel in high chloride conditions
279
Delays need for early repair
Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective
Cathodic protection
280
J.P. Broomfield
preferred that the galvanized steel would passivate in contact with the cement and the rate of zinc consumption to be as slow as possible. These matters have also been discussed by Yeomans in Chapter 1. Throughout the various chapters of this book, the authors emphasize the extent of the reinforcement-corrosion problem and the fact that its incidence is still costing the world economy billions of dollars per year. The engineer needs a “toolbox” of options to solve these problems and galvanized reinforcement is one such tool.
9.4. The Performance of Galvanized Reinforcement Issues of the properties of the galvanized bar, the bond between reinforcing steel and concrete and the cracking of the zinc coating and coating repair are variously dealt with in Chapters 2, 4 and 8 of this book. These important practical considerations cannot be overlooked but they can and have been dealt with, as seen from the range of data and discussion presented and from the number of structures in service containing galvanized reinforcement. If the galvanized steel is of appropriate quality and the reinforcement performs as required, the next question is: how much will zinc enhance the durability of a concrete structure? At the outset, it is important to note the wide availability of hot-dipped galvanized reinforcing bars, as evidenced by the number of countries in which it is used, the ease of handling and fabrication (i.e. bending) and its ease of use on site, particularly compared with FBECR. Although more expensive than black steel, galvanized bars are, for example, far less expensive than stainless steel and so less attractive to site pilfering. Overall, galvanized bars can be handled and treated in much the same way as black steel bars and this makes for an ease and simplicity of handling, especially when compared with FBEC bars.
9.4.1. Durability in Carbonated Concrete From the results discussed by Swamy in Chapter 2 and Andrade and Alonso in Chapter 5, it can be seen that galvanized steel is resistant to changes in the pH of concrete due to the effects of carbonation. It seems clear that zinc remains passive in carbonated concrete where the pH may be well below the level at which corrosion of black steel initiates (typically below pH 11.5) in the absence of chlorides. Because of differences in the rate of attack on the zinc coating due to variations in the chemistry and structure of the concrete and the nature of the galvanized coating, there are no precise details of consumption rates of the coating
Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective
281
in carbonated concrete. They are, however, quite low, usually somewhat less than a few microns per year.
9.4.2. Durability in Chloride-Contaminated Concrete In Chapter 1, Yeomans gives a comparison of models for the corrosion performance of black steel versus galvanized steel in concrete. For galvanized steel, there are three critical periods, namely: *
* *
the time to the initiation of corrosion of the zinc, i.e. the time for the chloride level to exceed the zinc corrosion threshold; the time for the zinc and zinc alloys in the coating to be consumed; and the deterioration rate of exposed steel due to active corrosion.
Obviously, to determine the time to initiation is a function of the diffusion rate of chlorides into concrete and the threshold chloride concentration for corrosion initiation for galvanized versus black steel. With black reinforcing steel, there is a consensus of the threshold for corrosion and the probability curve of chloride content versus corrosion initiation, as shown in Fig. 1 [7]. Such a model remains to be developed for galvanized reinforcement, noting, however, the complexity of doing this for galvanizing because of differences in the behavior of different forms and structure of the zinc-alloy coating.
Figure 1: The probability of corrosion versus chloride content for black steel (after Vassie, 7).
282
J.P. Broomfield
For black steel, the lowest threshold used is 0.2% chloride by mass of cement, equivalent to 0.028% by mass of concrete or 0.66 kg/m3 based on 14% cement content. This level is the one most widely use in North America, although a threshold of 0.4% is frequently used in the European literature and less than 0.1% is generally agreed as guaranteeing no corrosion. An overview of corrosion thresholds for black steel is given by Glass [8]. In Chapter 2, Swamy states that the chloride threshold is elevated several times with respect to black steel. He also states that the time to initiation is 4 or 5 times that of black steel and that there is a 25 –30 year life in a high-chloride environment. For example, a galvanized reinforced bus garage in Hamilton, Bermuda was demolished after 45 years exposure to a marine environment with no signs of corrosion. Similar conclusions are repeated many times in the literature, as cited in Chapter 6, where it is widely reported that galvanizing has a significantly higher chloride tolerance (to as much as 10 times) compared with black steel, resulting in an extension of life before active corrosion commences. The reports of many field surveys, also reported in Chapters 6 and 7, support these experimental determinations. Unfortunately, it is very difficult to compare results as they are for different exposure conditions, different concrete mixes and the results are quoted in different units, often in terms of chloride by mass of cement or chloride by mass of concrete but with no indication of the actual cement concrete in the mix. In Chapter 6, Yeomans enumerates various laboratory and field experiments involving galvanized reinforcement undertaken over more than a 40-year period. The review of the Bermuda installations by Allan in Chapter 7, which represents the single most complete and significant data set, extends this world-wide survey. It would appear that the corrosion threshold is in the region of 1–2% chloride by mass of cement. If the corrosion threshold of black steel is taken to be 0.4% chloride by mass of cement (a conservatively high value) and the threshold for galvanized steel to be 1% (a conservatively low value), then, using a typical diffusion coefficient of 1.4 £ 10212 m2/s, it can be calculated using Fick’s law of diffusion (see Chapter 2) that, for a concrete structure exposed to a marine environment (0.35% chloride by mass of concrete at the surface) and with 30 mm cover, black steel will start to corrode after 15 years. By comparison, galvanized steel will start to corrode after 44 years, indicating a life extension of about 30 years in actual time or a factor of about three times over black steel. This extension of life is consistent with that reported in several investigations cited in Chapter 6. As stated above, the chloride thresholds chosen for this example are conservatively high for black steel and low for galvanized steel, which has the effect of contracting the difference in time to the initiation of corrosion for the two types of steel. If, however, the equivalent
Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective
283
mid-point or low-point chloride threshold is used for both steels, then the case for galvanized steel is further strengthened. As far as the economics of galvanizing is concerned, Allan notes in Chapter 7 that galvanized reinforcement is 50% more expensive than black steel ex-factory, but only 10% more costly once shipping and handling are taken into account. For a typical building project, the cost reduces to 0.25–1.0% of the total capital cost compared with using black steel. A similar cost analysis is presented by Yeomans in Chapter 1. It should be noted, however, that the cost of galvanized steel varies widely from country to country, and also in-country, depending on the location and capacity of galvanizing plants and transportation and handling costs, as well as the market demand and competition from either black steel or FBECR in the region. From the diffusion analysis given above, it is easy to see that extending the time to first repair from 15 years to a minimum of 44 years (based on conservative chloride thresholds) is likely to save at least two cycles of concrete repair, which, even with life-cycle cost analysis, will easily save considerably more than 1% of the original construction cost. It would appear, therefore, that, if the corrosion threshold for galvanized bar is assumed to be 1% chloride by mass of cement or better, then galvanized reinforcing steel is likely to achieve a repair-free life in marine exposure for a Portland cement concrete structure with a planned 50-year life, since it is unlikely to need significant repairs only 6 years after corrosion initiated, if this were to occur. If it is necessary to guarantee a 100 þ year life, it may be better to specify other protection strategies such as stainless steel reinforcement, as has already been done for a number of structures where serious corrosion risks exist. In this case, however, the initial cost differential will be far higher compared with black steel and also galvanized steel. It should also be noted that there are a number of existing reinforced concrete structures containing galvanized bar that are at least 40 years old and which have provided a significant body of in-service performance data. There are no stainless steel reinforced concrete structures anything like 100 years old, so the issue of extrapolating data should be carefully considered, as discussed earlier. The designer should also consider whether a 100-year maintenance-free life is essential and whether future changes and upgrading might waste the initial investment.
9.4.3. Consequences of Installing Galvanized Reinforcement Once a structure has FBECR, stainless steel or galvanized reinforcement installed, there are consequences, not just for the durability of the structure but also for the ability to assess the structure during corrosion surveys and the like. Because
284
J.P. Broomfield
galvanized steel has a ferritic base which is magnetic, a coating of zinc will not upset a standard cover meter, unlike the situation with non-magnetic austenitic stainless steels, which need a specific type of cover meter to find and measure the reinforcement depth. Unlike FBECR, galvanizing will not affect electrical continuity if there is a need to use electrochemical techniques (e.g. half-cells) to assess or control corrosion. However, it will affect the interpretation of electrochemical test results because of the magnitude of the potential shift between steel and zinc. Once corrosion initiates, the potential of zinc is some 400 mV more negative than that of black steel and so it is important to understand how to interpret reference electrode (halfcell) potentials where the criteria suggested in ASTM C876 for the corrosion of black steel simply will not apply. This issue is discussed in some detail by Yeomans in Chapter 6 where it is noted that the potential of the zinc and zinc alloys in the coating rise steadily from about 2 1100 mV (vs CSE) to about 2600 mV as dissolution of the coating proceeds and the more iron-rich alloy layers in the coating are exposed. The interpretation of polarization resistance measurements may also be affected by the different corrosion processes occurring with galvanized bars compared with black steel bars. This problem will be exacerbated by the fact that, in some cases, zinc will be corroded and, in other areas, the steel itself will be corroding where the zinc has been consumed. The criteria for controlling cathodic protection systems may also be affected.
9.5. Conclusions Galvanizing is one of a range of techniques used to extend the life of reinforced concrete structures exposed to aggressive environments that promote reinforcement corrosion. The data show that it is very effective in combating the effects of carbonation-induced reinforcement corrosion. It increases the corrosion threshold from the range 0.2 – 1.0% chloride by mass of cement for black steel (see Fig. 1) to 1.0 –2.0% in the presence of the zinc coating. In a typical construction, this may extend the time to first repair from 15 years to at least 44 years and probably much longer. This is at a cost increase in overall construction of typically not more than about 1% (and often somewhat less than this), depending on local circumstances. For a structure with a life expectancy of 50 years, life-cycle costing shows this should be a cost-effective solution. It is to be noted, however, that no attempt has been made to compare such costs with all other corrosion-protection options listed in Table 1. Where very long life is required, say 75 – 120 years and more, complementary options such as 65% ground granulated blast furnace slag cement with high cover to the reinforcement could be used. Alternatively, the most expensive option,
Galvanized Steel Reinforcement in Concrete: A Consultant’s Perspective
285
the use of high-quality stainless steel reinforcement (such as the duplex grades) could be considered in conjunction with superior quality concretes. There are no absolute solutions to the problems of reinforcement corrosion, although the high chromium-bearing stainless steels are at present the best that can be achieved in the real world. What is clear is that there is a toolbox of techniques available with different performance, limitations and costs for each one. As shown in Table 1, galvanized reinforcement is one of these options and deserves wider appreciation and use than at present. Hopefully, this book will encourage more designers to consider its use in appropriate conditions and structures.
References [1] Koch, G. H., Brogers, P. H., Thompson, N., Virmani, Y. P., Payer & J. H. (2002). Corrosion cost and preventive strategies in the United States. Federal Highways Agency Report, FHWA-RD-01-156, March 2002, Washington, DC (Refer http://www. costofcorrosion.org). [2] Manning, D. G., Escalante, E. & Whiting, D. (1982). Galvanized rebar as a long-term protective system. Federal Highways Agency Report, FHWA-DTFH61-82-P-300-30041-2/3, July 1982, Washington, DC. [3] McDonald, D. B., Pfeifer, D. W., & Sherman, M. R. (1998). Corrosion evaluation of epoxycoated, metallic clad and solid metallic reinforcing bars in concrete. Federal Highways Agency Report, FHWA-RD-98-153, December 1998, Washington, DC. [4] Manning, D. G. (1996). Corrosion performance of epoxy-coated reinforcing steel: North American experience. Construction and Building Materials, 10, 5, 349–365. [5] Tonini, D. E. & Dean, S. W. (Eds) (1976). Chloride corrosion of steel in concrete. ASTM STP 626. American Society for Testing and Materials, Philadelphia, PA. [6] ACI 356.1R-00 Service life prediction — State of the art report (2000). American Concrete Institute (Refer http://www.masterbuilders.com/support/pageETools). [7] Vassie, P. R. (1987). The chloride concentration and resistivity of eight reinforced concrete bridge decks after 50 years service. Transport and Road Research Laboratory, Report 93. Department of Transport, Crowthorne, Berkshire, UK. [8] Glass, G. K., Reddy, B. & Buenfeld, N. R. (2000). Quantifying the inhibitive and aggressive nature of chloride contaminated concrete. Paper 10803, Eurocorr 2000 (published on CDROM). Institute of Materials, London.
This Page Intentionally Left Blank
Colour Plates
288
Colour Plates
Plate 1: Galvanized reinforcement concrete — general construction. (See Chapter 1, page 23)
Plate 2: Galvanized reinforcement concrete — buildings. (See Chapter 1, page 24)
Plate 3: Galvanized reinforcement concrete — bridges and highways. (See Chapter 1, page 25)
Colour Plates
289
Plate 4: Galvanized reinforcement concrete — coastal and marine. (See Chapter 1, page 26)
Plate 5: Galvanized steel products for use in concrete. (See Chapter 4, page 99)
290
Colour Plates
Plate 6: Aerial view of the islands of Bermuda. (See Chapter 7, page 200)
Plate 7: The New Watford Bridge. (See Chapter 7, page 212)
Colour Plates
291
Plate 8: Tynes Bay Waste-to-Energy plant foundations with 100% galvanized reinforcement. (See Chapter 7, page 219)
Plate 9: Tynes Bay construction showing a heavily reinforced ground beam. (See Chapter 7, page 219)
This Page Intentionally Left Blank
Index
abrasion 101 accelerated corrosion tests 188 –9 alkalinity 111– 18, 120– 3, 185 alloy layers 93– 8, 123– 4 anchorage 229– 31, 243 annealing 125 –7 applications 18 –26 beam tests 245 –6 bearing capacity 229 bending 99– 101, 217 Bermuda 199– 220, 222– 7 black steel 13 –14, 18 Bermuda 207 – 9 bond 251, 254 –6, 258 durability 50, 52, 281 – 3 research history 148, 150, 154, 165– 6, 173– 80, 191 bond 229 – 64 calcium hydroxyzincate 253 – 5 chloride ions 256– 7 chromates 256 – 60 design standards 262– 3 failure 235 –8, 240 –1, 264 fibre-reinforced concrete 248 flexural members 243 – 6 fusion-bonded epoxy-coating 249 –50 high strength concretes 246 –7, 249 hot-dip galvanizing 250 –62 hydrogen evolution 251 –2 pullout tests 235, 238 –42, 245, 254 repair methods 260 retardation effect of zinc 255 –6 rib geometry 261– 2
strength 218, 229, 238– 41, 253 stress-slip relationship 241– 3, 252 structural lightweight concretes 247 –9 bond-slip relationship, durability 46 –8, 64 brittleness 248 Building Research Establishment 167 calcium hydroxyzincate 113 –18, 123, 126, 130, 138– 41, 179– 82, 185, 190, 253 –5 carbonation 2, 27, 71, 83 Bermuda 210– 11, 213 corrosion 128 –31, 141 durability 48 –50, 280 –1 research history 153, 157, 164 –5 cathodic protection 279 chloride ions 2, 8 –11, 13 –14 addition during mixing 133 – 8, 141 barriers 79 Bermuda 201– 6, 208, 210– 11, 214, 224 –5 bond 256 – 7, 259 concentration threshold 73 – 4, 82 – 3 corrosion 71 –2, 132 –41 durability 39, 42, 48– 9, 51– 5, 60– 3, 281 –3 electrochemical removal 260, 264 hydrogen evolution 119 passivation 126 penetration from outside 138– 41 research history 150 – 5, 157, 161 – 3, 167 –80 tolerance 8– 11, 13– 14, 169, 279
294
Index
chromates 47, 53 bond 256– 60 hydrogen evolution 120 research history 153, 156, 175 – 6, 185 –7 chromium trioxide 149 –50 climatic conditions 32 –5, 51 –2, 275 coating mass 218 cobalt alloys 176 codes of practice 105– 7 cold climates 32, 35 cold-worked galvanization 43– 5 compression 239 concrete quality, research history 154, 191 concrete surface treatments 79 corner protection 101 corrosion 78 – 81, 273– 5 alkalinity 111 –18 atmospheric 35 Bermuda 201 –3, 208 –11 bond-slip relationship 46 – 8 carbon steel 80 – 1 carbonation 71, 128– 31, 141 chloride barriers 79 chloride ions 71 –2, 132 –41 concrete quality 78 – 9 concrete surface treatments 79 costs 2– 3 cover thickness 78 –9 cracking of galvanized coating 43– 5 current 116 –17, 121 –2, 129 – 31, 135 –40, 158, 163 durability 1, 78 –83 electrochemistry 111– 41 epoxy-coated reinforcement 81 –2 hydrogen evolution 118– 20 inhibitors 79 –80 initiation stage 16, 72 –5, 82 –3 integrated forecasting 76 –8 iron electrochemistry 71– 2 models 15 – 18, 72 – 6 permeability 4 potential 116 – 17, 119, 129, 131, 135 –7, 158
prevention 4 – 5 products 2, 9– 12, 163– 6, 191 propagation stage 17, 72 –3, 75 –6, 82– 3 protection stage 16– 17 research history 147 –8, 150, 153– 8, 163– 4 salinity 140– 1 cost analysis 14– 15, 27, 39 – 40, 215, 276 cover thickness 78– 9 cracking 2 Bermuda 206 –7 bond stress 233, 236 durability 38, 40, 52 –3, 56 –8 galvanized coating 43 –5, 98 –9 research history 148 – 150, 161, 169 –70 critical bond stress 244 critical slip 245 debonding 45 deemed to satisfy premise 33 deformations 229– 30, 233, 235 –6, 261 –2 delta layer 124 depassivation 9, 18 carbonation 49 –50 design specifications 33 –4, 41 –8 design standards 262– 3 dry climates 32, 35 ductility 44 durability 1, 275 bond-slip relationship 46– 8, 64 carbonation 48 –50, 280 – 1 chloride ions 39, 42, 48 – 9, 51– 5, 60 –3, 281– 3 climatic conditions 32 –5, 51 –2 concrete quality 36, 63– 4 corrosion strategies 78– 83 cracking of galvanized coating 43 –5 design 275 –80 design specifications 33– 4, 41– 8 economics 39 – 40 engineering implications 58– 60 exposure 34 –41, 61 geomorphology 36
Index hydration process 37 hydrogen evolution 46 –8 in situ performance 60 –4 long-term field tests 55– 7 mechanical properties 45– 6 replacement materials 40– 1 salinity 32, 35, 50, 53 structural integrity 38 –41 economics 14– 15, 27, 39– 40, 215, 276 ECR see epoxy-coated reinforcement edge protection 101 electrochemical chloride removal 260, 264 electrogalvanization 89 – 90, 178 epoxy-coated reinforcement (ECR) 81 –2 Bermuda 216 – 18 bond 259 –60 durability 276, 279 research history 160, 169, 173 see also fusion-bonded epoxy-coated reinforcement eta layer 124, 158, 160 European Code du Beton 263 evaporation rates 36 expansion forces 156 exposure 34– 41, 61 fabrication techniques 98 –104 bending 99– 101 coating damage 217 cracking of galvanized coating 98 –9 impact resistance 101 protection 101 repair 103– 4 welding 101– 3 failure in bond 235– 8, 240– 1, 264 fatigue strengths 45 –6, 56 –7 FBECR see fusion-bonded epoxy-coated reinforcement Federal Highways Administration (FHWA) 145 –6, 153 –6, 273 –4 ferrous sulphate 258 FHWA see Federal Highways Administration
295
fibre-reinforced concrete 248, 276, 278 field handling techniques 107 – 8 field performance 11 –14 field tests 55– 7 flexural members 243 –6 fluxing 92 fly ash 183, 276, 277 friability 11 fusion-bonded epoxy-coated reinforcement (FBECR) 4 – 6, 249 – 50, 274 –5, 280, 283 –4 galvanic series 88 gamma layer 124 geomorphology 36 Hamilton Harbor (Bermuda) 203 Hamilton Old Bus Depot (Bermuda) 207 handling damage 216– 17 high strength concretes 246– 7, 249 hot climates 32, 35 hot-dip galvanization 4, 6– 7, 88– 93 alloy layers 93– 8 Bermuda 200– 1, 216 bond 250 – 60, 261– 2 coating process 92 –3 coating structure 93 –5 codes of practice 105 – 7 durability 42 –3, 45 fluxing 92 operating conditions 125 –6 phosphorus content 95, 97 – 8 service life charts 89 silicon content 94 –6 standards 105 –7 surface preparation 91 –2 humidity 35, 157 hydration process 37 hydrogen evolution 46– 8, 118– 20, 149, 164 –5, 251 –2 ILZRO see International Lead Zinc Research Organization impact resistance 101
296
Index
in situ performance 60– 4 industrial pollutants 35 inhibitors 79– 80, 104– 5, 276, 278 initiation stage 16, 72 – 5, 82– 3 initiation –propagation model 72 – 6 inorganic silicate polymers 82 integrated corrosion forecasting 76– 8 International Lead Zinc Research Organization (ILZRO) 145, 150– 2, 155 –60, 170– 1, 182– 3, 187 –9, 201 –2 lightweight concretes 247 –9, 271 long-term exposure tests 55 –7, 171 –5, 184 –6 Longbird Bridge (Bermuda) 205 –7, 222 –7 mechanical properties 45– 6 membranes 276, 278 metallic coatings 82 metallized zinc coatings 89 – 90, 103 – 4 Ministry of Works and Engineering (MW&E) 199 –201, 203 –6 moment forces 229, 231– 4 MW&E see Ministry of Works and Engineering New Watford Bridge (Bermuda) 212 –15 nickel 126 Old Watford Bridge (Bermuda) 210 –11 OPC see ordinary Portland cement Ordinance Island Bridge (Bermuda) 208 –9 ordinary Portland cement (OPC) 120, 127, 275 paint systems 89 –90 passivation alkalinity 120 –3 chloride ions 126 electrochemistry 111– 18 galvanized coatings 123– 8
passivation current 116– 17, 121– 2, 129– 31, 135– 40, 158, 163 Penno’s Wharf (Bermuda) 204 permeability 4, 37, 51 petrographic examinations 222– 4 phosphorus content 95, 97 –8 pitting 167 pitting depth 55 – 7 pollutants 35 polyvinylchoride coatings 169 porosity 277 Portland cement 39 – 41, 158, 160, 272 –3 Poubaix diagrams 112 –14, 118 propagation stage 17, 72 – 3, 75 – 6, 82 – 3 protection stage 16 –17 pullout tests 235, 238 –42, 245, 254 radiation 35 re-alkalization 260 repair methods 260 repairing galvanized surfaces 103 –4 replacement materials 40– 1 retardation effect 255– 6 ribs 229 –30, 233, 235 –6, 261 –2 Royal Bermuda Yacht Club 204 rust formation 54 salinity Bermuda 200 corrosion 140– 1 durability 32, 35, 50, 53 research history 149, 165– 6, 172– 3, 177– 9, 185, 188 – 9 Sandelin effect 43, 96 –7 Sargasso Fish Processing Facility (Bermuda) 208 service life charts 89 serviceability 1 silicon content 94 –6 slip measurements 148 slippage 235– 6, 238– 9, 241 –3, 245 solders 103 specifications 33 –4, 41 –2 splitting 235– 41
Index sprayed zinc see metallized zinc coatings stainless steel 54, 80 – 1, 153, 278 Standard Beam Test 246 standardization 155 standards 105 –7, 262 –3 State of the Art Report on Coating Protection for Reinforcement 20, 22 steel 163 –6 storage 44, 216 stress loading 45– 6 stress-slip relationship 241– 3, 252 structural integrity 38 – 41 structural lightweight concretes 247 –9, 271 sulphur dioxide 35 surface preparation 91 –2 surface treatments 79 swelling 2 Technigalva process 97 tensile stress 233– 4, 239, 245– 6 thickness of zinc layer 45 transition zone 255 – 6 Tuutti model of corrosion 15– 18 Tynes Bay Waste-to-Energy Facility (Bermuda) 219 –20
297
ultimate bond stress 259 –60 ultimate strength 245 useable bond strength 244 volume expansion 273 Watford Bridge (Bermuda) 209 –15 wedge action 236 welding 101 –3 wind speed 36 yield points 235 zeta crystals 94– 5 zeta layer 124, 158, 160 zinc 163, 165– 6, 191 corrosion products 163, 165– 6, 191 electrochemistry 8, 88, 111 –18 inhibitors 104 –5 metallized coatings 89– 90, 103 –4 paint systems 89– 90 plating 89– 90 retardation effect 255 –6 solders 103 zinc hydroxychloride 156, 165, 176, 184 zinc-rich paint 103
This Page Intentionally Left Blank