The Hatfield Memorial Lectures Volume III

SEQ 0001 JOB IOM8627-000-013 PAGE-0001 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Hatfield Memorial Lectures Volume III

SEQ 0002 JOB IOM8627-000-013 PAGE-0002 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Dr William Herbert Hatfield FRS, 1882–1943. Courtesy of Sheffield Industrial Museums Trust

SEQ 0013 JOB IOM8627-000-016 PAGE-0003 PRELIMS REVISED 04AUG05 AT 00:07 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Hatfield Memorial Lectures Volume III Edited and Foreword by

Peter Beeley

Woodhead Publishing and Maney Publishing on behalf of The Institute of Materials, Minerals & Mining

CRC Press Boca Raton Boston New York Washington, DC

SEQ 0004 JOB IOM8627-000-013 PAGE-0004 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Woodhead Publishing Limited and Maney Publishing Limited on behalf of The Institute of Materials, Minerals & Mining Published by Woodhead Publishing Limited, Abington Hall, Abington, Cambridge CB1 6AH, England www.woodheadpublishing.com Published in North America by CRC Press LLC, 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487, USA First published 2005, Woodhead Publishing Limited  IoM Communications Ltd, 2005 The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the author and the publishers cannot assume responsibility for the validity of all materials. Neither the author nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Woodhead Publishing ISBN-13: 978-1-84569-101-1 (book) Woodhead Publishing ISBN-10: 1-84569-101-6 (book) Woodhead Publishing ISBN-13: 978-1-84569-114-1 (e-book) Woodhead Publishing ISBN-10: 1-84569-114-8 (eook) CRC Press ISBN-10: 0-8493-9242-X CRC Press order number: WP9242 The publishers’ policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acid-free and elementary chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Typeset in the UK by Dorwyn Ltd, Wells, Somerset Printed by TJ International Ltd, Padstow, Cornwall, England

SEQ 0005 JOB IOM8627-000-013 PAGE-0005 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Contents Foreword Printed Sources

vii ix

Process Technology and Research Basic Knowledge, Discovery, and Invention in the Birth of New Metallurgical Processes, F. D. Richardson, 1964 Twenty-Five Years On, H. M. Finniston, 1968 Electroslag Remelting – a Modern Tool in Metallurgy, E. Pl¨ockinger, 1972 Eurosteelresearch, R. S. Barnes, 1973 From Invention to Industrial Development, L. Coche, 1978 Net Shape Solidification Processing of Steel, 1945–1995, M. C. Flemings, 1989 The New World of Steel, J. Edington, 1995 Technology–Driving Steel Forward, M. J. Pettifor, 2001 Metallurgical Modelling of Thermomechanical Processing, C. M. Sellars, 2002 Economics and Developments in the Iron and Steel Industry Development in the Iron and Steel Industry in Great Britain during the last Twenty-Five Years, T. P. Colclough, 1954 The Place of Mini-Steelworks in the World, W. F. Cartwright, 1971 Materials and Malthus, Sir Alan Cottrell, 1974 European Steel: What Future? Sir Robert Scholey, 1987 Iron, the Hidden Element – the Role of Iron and Steel in the Twentieth Century, R. Boom, 1999 Author Index Subject Index

1 3 25 45 63 89 97 101 123 143

165 167 183 201 207 227

247 249

v

SEQ 0006 JOB IOM8627-000-013 PAGE-0006 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0007 JOB IOM8627-000-013 PAGE-0007 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Foreword This third volume of Hatfield Memorial Lectures comprises two groups. In the first of these the central theme is process metallurgy, the lectures encompassing technical aspects of plant design and operation, including associated research and development. Lectures in the second group address the organization of the iron and steel industry, including its general structure and economic circumstances. Although there is occasional overlap, these two broad themes complement the three contained in the two earlier volumes, to cover five fields as contemplated at the outset: Volume I is devoted to Properties and Behaviour of Materials and to Applications, and Volume II to the single theme of Metallography and the Structure of Iron and Steel. A similar system to that used in the earlier volumes has been adopted in the present case: the lectures are arranged in date sequence within each group, providing pictures of the respective subject areas over the whole period since the outset in 1946, when the first lecture was delivered by Dr Hatfield’s contemporary Dr George B. Waterhouse. This dealt with Hatfield’s own work and formed the introduction to the first volume, which also contained biographical notes on Hatfield himself; further aspects of his life and character were portrayed in Professor A. G. Quarrell’s personal recollections, which provided the introduction to Volume II. William Herbert Hatfield had been born in 1883 and had died in 1943. The Memorial Lecture series had been instituted by the University of Sheffield in 1944 as a joint project with the Royal Society and the Iron and Steel Institute. The fiftieth lecture, the most recent and last to be included in the present series of volumes, was delivered by Professor C. M. Sellars of the same University in December 2002. The annual Hatfield Lecture continues to draw large audiences from far afield and reflects the ongoing interest in the major metallurgical advances of our time. Many of the chosen topics have significance over the wider field of materials and across diverse branches of engineering. It is hoped that supplementary volumes will add to the record of these occasions at suitable intervals in future years, and will be arranged on a similar thematic basis. Peter Beeley

P. R. Beeley DMet is a Life Fellow and former Senior Lecturer in metallurgy in the University of Leeds

vii

SEQ 0008 JOB IOM8627-000-013 PAGE-0008 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

A shot at leisure. A picture of W. H. Hatfield from the Firth Brown Photographic Collection. Courtesy of Sheffield Industrial Museums Trust (copyright)

SEQ 0009 JOB IOM8627-000-013 PAGE-0009 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Printed Sources Listed below are the lecture numbers, titles and authors of each of the papers appearing in this volume. The original place and date of publication are also given. Seventh Lecture: Development in the Iron and Steel Industry in Great Britain during the last Twenty-Five years, by T. P. Colclough J. Iron Steel Inst., July 1954, 297 Sixteenth Lecture: Basic Knowledge, Discovery, and Invention in the Birth of New metallurgical Processes, by F. D. Richardson J. Iron Steel Inst., March 1965, 217 Nineteenth Lecture: Twenty-Five Years On, by H. M. Finniston J. Iron Steel Inst., February 1969, 145 Twenty-second Lecture: The Place of Mini-Steelworks in the World, by W. F. Cartwright J. Iron Steel Inst., April 1972, 221 Twenty-third lecture: Electroslag Remelting a Modern Tool in Metallurgy, by E. Pl¨ockinger J. Iron Steel Inst., August 1973, 533 Twenty-fourth Lecture: Eurosteelresearch, by R. S. Barnes Ironmaking Steelmaking, 1974, 1 (2), 61 Twenty-fifth Lecture: Materials and Malthus, by Sir Alan Cottrell Met. Mater., March 1975, 32 Twenty-eighth Lecture: From Invention to Industrial Development, by L. Coche Met. Mater., February 1979, 25 Thirty-fifth Lecture: European steel: what future? By Sir Robert Scholey Ironmaking Steelmaking 1987, 14, 267 Thirty-seventh Lecture: Net Shape Solidification Processing of Steel, 1945–1995, by M. C. Flemings Report from Cast Metals, 1990, 2 (4), 231 Forty-third Lecture: The New World of Steel, by J. Edington Ironmaking Steelmaking, 1997, 24 (1), 19 Forty-seventh Lecture: Iron, the hidden element: the role of iron and steel in the twentieth century, by R. Boom Steel World, 2000, 5, 1, 88 Forty-ninth Lecture: Technology: driving steel forward, by M. J. Pettifor Steel World, 2002, 7, 11 Fiftieth Lecture: Metallurgical Modelling of Thermomechanical Processing, by C. M. Sellars Previously unpublished

ix

SEQ 0010 JOB IOM8627-000-013 PAGE-0010 PRELIMS REVISED 31JUL05 AT 22:03 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0003 JOB IOM8627-001-009 PAGE-0003 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE SIXTEENTH HATFIELD MEMORIAL LECTURE

Basic Knowledge, Discovery, and Invention in the Birth of New Metallurgical Processes F. D. Richardson At the time the lecture was given, Professor Richardson was Professor of Extraction Metallurgy at the Imperial College of Science and Technology. The lecture was presented at Church House, Great Smith Street, London, in the evening of Wednesday 25 November 1964.

I am greatly honoured that you should have invited me to give this 16th Hatfield Memorial Lecture. Honoured because of the importance and standing throughout the metallurgical world of our Iron and Steel Institute, and honoured because of the distinction of the man to whom these lectures are a memorial. I was not privileged to know Dr Hatfield, but from all I have read and heard of his life, I realise that I missed knowing a truly remarkable man. Many remember him for his obvious achievements – his work on stainless and heat resisting steels, his post as Director of Firth Brown, his Chairmanship of three very important ISI Committees, his Fellowship of the Royal Society and his Bessemer Medal. But to those who worked personally with him there remains a lasting impression of enterprise, energy and openmindedness: an awareness of scientific ability in others, great intuition, and style in all things.

When I came to consider a subject for this lecture I looked at Dr Hatfield’s Campbell Memorial Lecture1 given in the USA in 1928, and I was struck by his sharp interest in new ideas and new developments: by his recognition of the importance in this connexion of the growth of purely scientific knowledge and of the need for industrial enterprise and the state to ‘cast its bread upon the waters’ and help those who work, particularly in the universities, to develop the fundamental aspects of their subjects. I therefore came to thinking of the way new processes come about. I suppose that today people are searching as never before for new ideas which show promise of becoming new processes. At the same time most of us are unaware of how the new processes which surround us came to be developed. Although full details of existing processes are often available, the exact ways in which they have come about are rarely disclosed. I have always found the histories of new developments extraordinarily interesting and I think we can learn from them. I know that Halevy wrote ‘We learn from history that we do not learn from history’, and this pessimistic view may be true as regards politics and international affairs. But

3

SEQ 0004 JOB IOM8627-001-009 PAGE-0004 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

4

Hatfield Memorial Lectures Vol. III

when it comes to scientific development I believe some greater optimism is justified. I have therefore decided to look at a number of new processes and to examine how they came about, and at what points basic knowledge, discovery, and invention had their own particular impacts. I have chosen extractive processes because they are my special interest. I shall include both ferrous and non-ferrous topics, because much of interest has been happening in the non-ferrous world, and because I consider we lose a great deal if we restrict our thoughts, or our reading, to one field or the other. It is difficult to acquire the detailed historic knowledge with which I am here concerned for it is not to be found in books. I am therefore personally indebted to those who have so kindly given me the information I have sought. But before I come to discuss those modern processes I wish to dwell shortly on the character of the men who bring about new processes – or what may be called the ‘flesh and blood’ aspect of development.

THE INVENTIVE SPIRIT For a few minutes therefore let us go back to 1870 and to a young man of twenty, who was the first to look seriously at the chemistry of slags, and what more entrancing subject could you have than that! I refer to Sidney Gilchrist Thomas who in 1878 brought forth the basic steelmaking process. A lively, unfinished sketch of him is shown in Fig. 1. despite his frail physique, Thomas possessed exceptional enterprise, energy and intuition, and he coupled these with that singleness of purpose, which is usually needed for a great invention. At the age of twenty he had a mind uncluttered with irrelevant experience and he was able to think along lines which were closed to experienced metallurgists such as Percy and Lothian Bell. Thomas epitomises the human aspect of creating a new process. He had the character of an inventor; and was always looking for unsolved problems. We see him for example insisting, thirty five years before Haber’s synthesis, that ammonia should be produced from air and water because the elements hydrogen, nitrogen and oxygen were to be had for nothing from these sources. In 1870 Thomas was the clerk in the Thames police court, working to support his widowed mother and family. Chemistry was his hobby and he attended a course of lectures at the Birkbeck Institution, delivered by George Challenor. At this juncture Challenor justified all the lectures he ever gave, for, with Thomas in his class, he had occasion to say that the man who eliminated phosphorus from iron by means of the Bessemer converter would make his fortune. This challenge sank deeply into Thomas’s mind and inspired him to wrestle with the idea for eight years, excited as much by the commercial prospect as by the scientific challenge. One cannot but be reminded of the young Hall who was similarly inspired by Professor Jewetts when studying chemistry at Oberlin, Michigan. From then on Hall’s ambition was to find a way of producing aluminium cheaply and he invented the electrolytic process in 1886, simultaneously with H´eroult in France but quite independently. I hope we university professors are as inspiring today as we were at the end of the

SEQ 0731 JOB IOM8627-001-013 PAGE-0005 16TH LECTURE REVISED 06FEB05 AT 19:27 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

Fig. 1

5

Portrait of Sydney Gilchrist Thomas.

last century! Sometimes I think that in teaching our students we lay too much stress on the unsolved theoretical problems and too little on the unsolved practical ones. We have a great opportunity to invent by proxy and we ought not to miss it! The problem of phosphorus elimination existed because there were great quantities of phosphoric ores; the orebody was providing the impulse for development as is so often the case in extractive metallurgy. The problem was solved in principle in the space of five years; during this time Thomas pieced together in an original way the available chemistry of phosphoric acid and silica, and of the slag–metal reactions which occur in the acid Bessemer and the old puddling processes. The problem was solved in practice after a further two or three years of dedicated work, during which Thomas obtained the help of his cousin Percy Gilchrist at Blaenavon steelworks in Wales. In this period Thomas ran, at first secretly, his tests with hot metal and developed the basic linings and the method of making lime additions. Most of the work was done at weekends, which started with a night’s journey from London to Blaenavon, and finished with a dash from the train home so Thomas could be in court on Monday morning. As Thomas wrote laconically in a letter to his cousin2 ‘I have not time enough to do. I only go home to sleep and eat. Most unsatisfactory’.

SEQ 0006 JOB IOM8627-001-009 PAGE-0006 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

6

Hatfield Memorial Lectures Vol. III

Ultimately this hard struggle undermined Thomas’s health and he died in 1895. But by then he had reaped a great harvest of fame and had shown himself to be a man of social as well as technological imagination. His success depended on brilliant scientific extrapolation, and on invention in making the linings. A similar theoretical approach to this problem was made simultaneously by George Snelus, who was a chemist at Dowlais Steelworks and some thirteen years older than Thomas. He had likewise concluded that a basic lining would be necessary if phosphorus were to be removed in the converter process. Snelus appears to have been the only other man to stand on the verge of Thomas’s invention. He was also a remarkable man, later a Fellow of The Royal Society and Vice-President of The Iron and Steel Institute, but his way of life was utterly different from Thomas’s. He was an expert shot, an enthusiastic horticulturalist, and a staunch conservative, in much demand for public meetings, at least according to the Proceedings of the Royal Society! With all these interests he had not that singleness of purpose which is required to turn a good idea into a workable process. Today we seem far away in time from 1878, and our industrial scene has vastly changed. But the reactions we exploit in steelmaking are much the same and some of our current thoughts on the utilisation of slag are just the kind that Thomas would have had. For Thomas saw that the mountains of phosphoric slag that accrued from his basic process contained great quantities of phosphorus which could be of value in agriculture. He devoted the last years of his life to this problem and to buying tons of slag in anticipation of success! It was at first thought that the valuable phosphorus should be extracted chemically from the slag, but it later turned out that for agricultural use, it was only necessary to grind the slag to a sufficiently fine powder. As often happens, the ultimate solution to the problem was simpler than anyone had dared to hope. Today we have again become concerned with making use of slags, those blast furnace slags which we have regarded all too complacently in the past, and there is hope that inspired and tenacious work will lead to the production of useful microcrystalline bodies from this material.

BASIC KNOWLEDGE AND THE STRUCTURE OF METALLURGY Since Thomas’s time there have been enormous developments in our understanding of scientific and engineering principles. Metallurgy has developed in three main directions, first on the physical side, then the chemical and now most lately on the process engineering side. I referred to this matter in detail in my Howe Memorial Lecture to the AIME last February.3 Tonight I merely wish to stress that these three basic sectors are equally important and they provide the springboard from which the metallurgist must take his leaps into the unknown, it he is to make the best use of all the basic knowledge which has been accumulated before him. I find it convenient to represent the metallurgy of today by two triangles which will help to make clear the terms I use later in this lecture. These are shown in Fig. 2. On the left are the three basic sectors of chemical metallurgy, physical metallurgy and process

SEQ 0007 JOB IOM8627-001-009 PAGE-0007 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

7

engineering metallurgy. Most of the phenomena of metallurgy can be considered on the micro scale, in terms of chemical and physical metallurgy. But as soon as one turns to the behaviour of metallurgical materials in bulk or to any form of processing, it is essential to bring in process engineering metallurgy. This is concerned with fluid and plastic flow and heat and mass transfer, in all kinds of metallurgical situations. It covers the principles involved in chemical and physical processing ranging from ore upgrading to metal treatment, and is especially concerned with phenomena, in which the chemical engineer has not so far displayed any marked interest. I call this sector process engineering though it might equally well be termed process science. It is a metallurgical responsibility and I do not believe that we can wait for the chemist to come up with the answers in chemical metallurgy. We have only to ask ourselves how much would be known about the chemistry of steelmaking, if it had not been for the work of Chipman, Korber, Oelsen, Herty, Schenck, and other distinguished scientists working principally in the metallurgy departments of universities or in the steel industry.

Fig. 2

The anatomy of metallurgy: (a) the three basic sections; (b) the three applied sections.

Founded on the three basic sectors of metallurgy are three applied sectors as shown on the right hand side of Fig. 2. These are extractive metallurgy based primarily on chemical and process engineering metallurgy, mechanical metallurgy (or metal treatment) based primarily on physical and process engineering metallurgy, and materials science (the design and behaviour of new materials and alloys) based primarily on chemical and physical metallurgy. The shaded overlap between these three sectors is meant to stress that none of them is based so completely on two of the three basic sections, that it is entirely independent of the third. The relative importance of the three basic sectors of metallurgy in any new development depends on the precise nature of the new steps. Broadly speaking there are three

SEQ 0008 JOB IOM8627-001-009 PAGE-0008 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

8

Hatfield Memorial Lectures Vol. III

kinds of new development; processes based on new reactions, in which both chemical and process engineering factors are important: processes in which old reactions are exploited in new ways, and which therefore depend primarily on new process engineering: new routes from ore to metal, in which existing processes are integrated in new ways, and which depend on critical path analysis and optimisation via systems engineering. If we look at this group in an iron and steel context we see that so far as oxide routes from ore to metal are concerned, there is little or no likelihood of discovering important new reactions, apart possibly from some vaporisation reactions, which might conceivably become important in vacuum refining. New oxide processes must therefore come from conducting old reactions in new ways, and such developments hinge on finding new methods of obtaining the necessary heat and mass transfer, either more conveniently or at greater rates. New iron and steelmaking reactions may, however, be found in halide routes; we have already seen the novel work of Reeve in this direction, and now the Peace River scheme in Canada (see below) looks to be imaginative and promising. New reactions may also be found in routes involving aqueous and organic solutions, and in this area I feel much awaits to be discovered. Reactions in such media may be uneconomic today but the progressive cheapening of heavy organic and inorganic chemicals may make them worthwhile in the future. Such routes will always involve relatively expensive chemicals, and these, especially the halogens and the organic solvents, must be recovered and recirculated, so that they do their chemical work again and again. For new aqueous and organic processes to become competitive for steel, I believe they must supplant not only the steps of iron and steelmaking but also much of the route from ingot to finished or semi-finished product. One way of doing this is to make powders and to rolll them into strip. This latest step is the weakest link in the chain of operations, and those interested in such routes need to concentrate much of their work on this stage.

THE ZINC-LEAD BLAST FURNACE A splendid example of a new process in which old and well known reactions have been exploited in new ways is the zinc–lead blast furnace,4 developed by the Imperial Smelting Corporation, with furnaces working or building in the UK, Australia and Europe. The blast furnace which is often considered ‘old hat’ in the iron and steel industry is very much ‘new hat’ in the zinc world. These zinc furnaces are rectangular in section; they have about 100 ft2 of hearth area and produce about 160 tons of zinc and 80 tons of lead per day. The process is represented in broad outline in Fig. 3. Our story starts in the late 1930’s when Stanley Robson and Kenneth Morgan of the Imperial Smelting Corporation considered it important to free the production of zinc from the small scale imposed upon it by continuous retorting. Their idea was to reduce a zinc oxide sinter to zinc vapour in a blast furnace and to recover the metal from the blast

SEQ 0297 JOB IOM8627-001-011 PAGE-0009 16TH LECTURE REVISED 21DEC04 AT 01:50 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

9

Fig. 3 Broad outline of the zinc–lead blast furnace process.

furnace gases. They saw that by this means they might also be able to treat lower grade concentrations than the retort process could accept. At the time they were much stimulated by a very good and imaginative review by Maier5 of the US Bureau of Mines, entitled ‘Zinc smelting from a chemical and thermodynamic standpoint’. This report in no way foreshadowed the zinc blast furnace. Rather the reverse, for it showed how the old idea of making liquid zinc in a blast furnace under high pressure was inherently unsound. Nevertheless Maier unwittingly contributed to the new process, for his report laid the basis for much of the thermodynamic reasoning done in the course of the development. This needs stressing for the contributions which reviews of this type make to industrial progress is all to often forgotten. Robson and Morgan realised that on grounds of fuel economy, it would be essential to convert substantial amounts of carbon to carbon dioxide and to carry this in the top gas. This brought with it a great risk from the back reaction Zn + CO2 = ZnO + CO which might occur so rapidly on cooling the top gases that the yields of metal would be poor. There was a second intrinsic chemical risk: that it might be impossible to avoid serious losses of zinc oxide in the slag, if the oxygen potential at the bottom of the furnace were kept high enough to keep iron oxide in the slag and prevent the production of solid iron. There were virtually no activity data on slags at that time (we have much more precise ideas today) but in the event this did not turn out to be a difficulty.

SEQ 0010 JOB IOM8627-001-009 PAGE-0010 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

10

Hatfield Memorial Lectures Vol. III

At the time the back reaction looked to be the most likely cause of failure; but the investigators had noted that in the existing process for fuming zinc from molten slag and recovering zinc oxide, reoxidation of the metal vapour was not so swift or complete as to preclude all hope of success. Shortly before the war the Corporation’s Research Department made some small-scale experiments on shock cooling gas mixtures of Zn + CO + CO2 by means of water cooled tubes and the results gave some encouragement. During the period 1945 onwards, a thorough thermodynamic analysis, together with heat balances, was made for the process by Stephen Woods and John Lumsden. It became clear that it should be possible to retain iron in the slag provided one could obtain a satisfactory control on the oxygen potential at the bottom of the furnace, and mass transfer in this region did not let one down. It also became clear that a good balance between Zn, CO, and CO2 in the top gas would be about 6% of metal, with a ratio of CO/CO2 of about 2/1 The problems then centred around the suppression of the back reaction and collection of the zinc. A small experimental brick furnace was built and fed with hot briquettes and cold coke. The zinc laden gases were drawn away at about 900°C via a flue situated about half way up the furnace. They were then passed through a heated column of coke to raise their temperature to 1000°C and eliminate some of the potentially dangerous carbon dioxide. Finally the gas was shock cooled by bubbling through liquid zinc. But the suction required to operate this condenser was undesirably high and the amount of reoxidation was unacceptably large. A highly original proposal was then made by Jack Derham. His idea was to bubble the gases through cool liquid lead, so that very rapid cooling and absorption of the zinc by the lead would be simultaneously attained. The idea of shock cooling a metal vapour in this way had come to him some years before when thinking about the production of magnesium. In this new context it seemed more promising. If lead were used as coolant it would dissolve the zinc and this might help in suppressing oxidation. The solubilities were known from the lead zinc phase diagram and there appeared to be practical advantages. Liquid lead was known to be much less corrosive to steel than liquid zinc, and it fortunately retains this advantage even when saturated with zinc. At the time however it did not look so promising, and the line of criticism was that if the zinc could be condensed as easily as that, somebody would surely have done it before! But Derham went ahead with his experiments and these led to the highly successful lead splash condenser. Today liquid lead saturated with zinc (2.15%) at 450°C is pumped into the condenser and broken into a spray by paddles. It cools the gas rapidly from 1000°C, absorbs 0.25% more zinc and leaves at 560°C. On cooling the lead back to 460°C, liquid zinc separates at the surface where it can be removed by a system of weirs and channels. This idea was undoubtedly the major inventive step, and on it hinged the success of all that had gone before. The rest of the development was carried out on a larger furnace of 7 ft2 cross-section and 10 ft high. There followed simplifications in treating the top gas before entering the condenser. It was found not to be necessary to pass the emergent gasses through heated

SEQ 0011 JOB IOM8627-001-009 PAGE-0011 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

11

coke, for the back reaction was not serious except on the walls, even when the zinc pressure was three times greater than that in equilibrium with the CO and CO2, and even this surface reaction could be avoided, if the walls were kept hotter than the gas. It then turned out that provided the gases were raised from 900 to 1000°C after leaving the furnace, the zinc could be collected in the splash condenser with virtually no back reaction. The temperature is now raised by admitting a little air, a very neat thermodynamic trick. The equilibrium constant for the reoxidation shifts so rapidly from zinc oxide to zinc with rising temperature, and the heat evolved in the combustion of some of the CO by the added air is so great, that one can shift the equilibrium in favour of zinc vapour by the unexpectedly simple procedure of admitting air! Only a group well versed in the thermodynamic approach would have spotted this elegant solution. For it is not normal to admit an oxidising agent to prevent oxidation occurring! But the trick could never have been conceived, without the basic thermodynamic data for CO, CO2, Zn, and ZnO, which had been worked out by previously careful experimenters. With lead drosses inevitably being produced in the process of recovering the zinc, it was obviously sensible to smelt these in the furnace and collect the lead at the bottom. It was immediately apparent that this would impose no thermal, and virtually no chemical, demand on the furnace. Once this had been found to work, it was clear that a zinc/lead oxide sinter could be charged, and both zinc and lead produced simultaneously. In 1950, over ten years from the inception of the project, two pilot furnaces producing 20–30 tons of zinc per day were built: these now have capacities of 30–40 tons per day and production furnaces are making 160 tons per day. This development, now so very successful, took a very long time to complete. It was much delayed by the war and at more than one point it came near to being abandoned. Its very slowness would have daunted lesser teams and lesser companies, and it demanded as much tenacity as skill from all concerned, and especially from Kenneth Morgan.

LD AND LDAC The development of the LD provides another good example of a new process in which old reactions are exploited in new ways. Its origins are worth considering in detail because important point emerge, and because the fact that the process was not invented in Britain, is sometimes cited as criticism of the research effort in steel in this country. The idea of blowing oxygen into hot metal from above is mentioned in Bessmer’s Patent of 1856, but at that time there was no tonnage oxygen and the possibilities of making steel in this way could not be explored. It is evident from a drawing in his patent (Fig. 4a) that Bessemer had no conception of the hydrodynamics of jetting oxygen into hot metal. As we all now know, what actually happens in this system, so violently dispersed by the combined effects of oxygen jet and CO evolution, is more closely represented by Fig. 4b. When oxygen became available on the tonnage scale it was natural that its use should appeal to steelmakers on the Continent. The steel industry there was based on air blown

SEQ 0296 JOB IOM8627-001-010 PAGE-0012 16TH LECTURE REVISED 21DEC04 AT 01:45 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

12

Hatfield Memorial Lectures Vol. III

Fig. 4

Top blowing of hot metal with oxygen (a) drawing from Bessemer’s patent of 1856; (b) the top blown converter today.

converters, and it had always been troubled by the nitrogen which got into the metal at the end of the blow. In particular the Austrians at the end of World War II had no satisfactory means of making steel from their irons, which were of intermediate phosphorus content and unsuitable for the basic Bessemer. They did not want to build open hearths, because they were very short of scrap and they wished to avoid the high capital costs associated with these furnaces. They therefore turned to basic coverters using enriched air, and finally Suess of V¨oest produced the top blown oxygen process.6 In this he was much influenced by the work of Durrer and Hellbrugge7 who had produced steel successfully by blowing an inclined jet of oxygen on a bath of pig iron in a two ton converter. Suess started his experiments in 1949 and his company began regular production in 25–30 ton converters in late 1952. Today in some countries converters of 200 tons capacity are in continuous operation. Suess was the first to study in a systematic, though empirical, manner what I regard as the process engineering of top blowing; that is the relationships between lance height, and diameter, the rate of oxygen flow, and the course of the refining reactions. There was no directly relevant experience on which he could draw and he made his first trials with two ton melts. But he wisely passed to the 15 ton scale without waiting to get everything working smoothly with his little converter. In this he was exceptionally wise, for heat, mass, and momentum transfer and the course of the reactions in a dynamic and violently dispersed system such as this, cannot be reliably simulated on the small scale. To my mind the importance of this move cannot be overestimated. It was the key to rapid success and it underscores the necessity for those aiming to develop new processes or to increase our knowledge in process engineering, to have adequate large scale facilities at their disposal.

SEQ 0013 JOB IOM8627-001-009 PAGE-0013 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

13

It was fortunate that only low phosphorus Austrian iron was available for the first stage of the work, for later attempts to use high phosphorus iron ran into difficulties. These were associated with getting much more lime into the slag and fluxing it at high speed, so that the phosphorus could be removed at the same time as the carbon. This problem was solved simultaneously by the Benelux CNRM and the French IRSID, who were also working for steel industries which were converter based. The CNRM had previously worked at the engineering of injecting powdered lime into flowing gases, for they had tried this on bottom blown converters and abandoned it. Thus it came about that when people tried to adapt the LD process for high phosphorus irons, Metz of the CNRM8 saw that the lime stood a far better chance of being fluxed immediately if it were injected into the oxygen gas and carried directly to the very hot zone where gas and metal meet. Experiments were then made at ARBED on the 26 ton scale and quick success fully justified this action. Marc Allard and his colleagues at IRSID thought along precisely the same lines, for they likewise had gained previous experience of injecting lime into flowing gases from their work on desulphurising hot metal with powdered lime carried in nitrogen. I referred to the industrial reasons why the LD did not evoke much response in Britain, but there was another reason for this lack of interest. There was a rational scepticism concerning the suitability of the LD process for British irons containing some 1%P. It arose from what I call a ‘hypertheoretical’ attitude. Basic knowledge of the [P]–[O]– (P2O5) equilibrium indicated that phosphorus could only be removed to low levels in the presence of substantial amounts of P2O5 in slag, if the oxygen activity in the metal was very high. This was perfectly correct. But it was then supposed on the basis of previous experience, that sufficiently high oxygen activities could not be attained in a pneumatic process without a correspondingly low carbon. At that time little or no basic knowledge existed concerning the kinetics of slag–metal and gas–metal reactions, although it was generally thought that they were transport controlled. Unfortunately thermodynamic data on slag and metal solutions were then too scarce to show that in steelmaking, equilibria betwen slag and metal of the type [Mn] + (FeO) = (MnO) + [Fe] 2[P] + 5(FeO) = (P2O5) + 5[Fe] could be quite closely approached (within 15%), although the equilibrium between metal and gas (the [C] [O]–CO equlibrium which is approached at a different interface) might be very far from attainment. This fact makes the removal of phosphorus in the LD possible. The reason lies in surface areas, concentrations, and mass transfer coefficients.9 Kinetics is played off against thermodynamics, not chemical kinetics but the kinetics of mass transfer in the vicinity of two different phase boundaries. Had our knowledge of process engineering metallurgy been as developed in 1949 as it is today, the first British theoretical estimates of the potentialities of the LD would have been much more favourable than they were. The tendency for the knowledgeable, good theoretician to put too much faith in his predictions, has resulted in highly competent and indeed imaginative persons failing to

SEQ 0014 JOB IOM8627-001-009 PAGE-0014 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

14

Hatfield Memorial Lectures Vol. III

make important discoveries. This was how Larmor in 1897 missed discovering the Zeeman effect. Larmor’s theory of radiation, before the era of the electron, suggested that if a source of radiation were plunged into a magnetic field the lines of the spectrum ought to be broadened. Larmor then calculated the magnitude of the effect to be expected with a reasonable field: he reckoned it would be too small to observe and that the experiment was not worth attempting. But Larmor’s treatment was not correct for he thought the atom would be radiating. Had he known of the electron he would have come to quite a different conclusion; for his theory predicted that the magnitude of the effect depended on the ratio of change to mass for the radiating particle. Zeeman however made the experiment with sodium flame in a strong magnetic field, and obtained a small but observable effect, which now carries his name. Looking back we can see that Larmor should have pressed on regardless. We can also see that whenever we make theoretical prodictions, it is wise to remember, that any experiment that has not been done before is worth attempting at least once; or if we believe, as I do, in the German saying ‘Ein Versuch ist kein Versuch’, at least two or three times! One must not, of course, over stress the caution of the theoretician, for more often than not it is the experienced practical man who has the lesser vision.

THERMAL MAGNESIUM In contrast with this last story, stands the development of the thermal process for magnesium production.10 Shortly before World War II, Professor Lloyd Pidgeon, who was then in the National Research Council Laboratory in Ottawa, was asked to study the production of magnesium for defence purposes in Canada. After considering the dry chlorination of brucite (Mg(OH)2) followed by electrolysis, he turned to direct reduction processes involving carbon and silicon. He considered that calcined dolomite, where the magnesium is present as oxide, might be reduced in a metal retort by ferrosilicon, the magnesium being distilled away from the solids and suitably condensed. The reaction is now known to be 2CaO + 2MgO + Si = 2Mg(g) + 2CaO.SiO2 but it was not then realised that the orthosilicate of calcium would be formed. Calculations were made on the supposition that the reaction gave only magnesium and silica. They were based on the then existing heat of formation of silica, which is now known to have been greatly in error. These calculations suggested that substantial equilibrium pressures of magnesium and hence reasonable rates of production, could only be achieved at 1300–1400°C. At these temperatures great problems would have arisen with the retorts so in the light of knowledge then available the theoretical assessment was most unpromising. Pidgeon, however, pushed ahead with experiments using Ni–Cr retorts, hoping that it might be possible to get the reaction to go reasonably fast under vacuum at 1100°C. The

SEQ 0015 JOB IOM8627-001-009 PAGE-0015 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

15

results exceeded expectations for relatively high pressures of magnesium, of the order of 20 mmHg are in fact produced by the dolomite reaction under equilibrium conditions at 1100°C. But two further difficulties had to be overcome. The first tests had been made with a simple retort heated and cooled in a batch cycle, but this treatment produced cracks in the retort after some 20 experiments. The retorts therefore had to be maintained at the operating temperature and arrangements made to open the equipment to the air and disconnect the condenser with the retorts at temperature. Trouble then arose from the pyrophoric nature of the condensed magnesium. It was observed that the deposit was feathery and friable at the cooler end of the condenser and that ignition was initiated at this end when the vacuum was broken. This was also the end where the alkali metals were concentrated. Further experiments with sodium carbonate introduced into the dolomite, confirmed the connexion between the alkali metals and the feathery nature of the deposit and the difficulty was overcome by the introduction of a fractionating condenser. With this the deposit of magnesium was massive and crystalline: it could be exposed to air even at the condensing temperature of 500–600°C, and when melted it formed an ingot without significant losses from oxidation. This process is fully competitive with the electrolytic route and is used in the magnesium plant which has just been built in Britain at Hopton.

THE SUBHALIDE PROCESS FOR ALUMINIUM Now I would like to turn to a process which springs primarily from the discovery of a new reaction, and has a fair chance of slowly outsting the classic electrolytic process for making aluminium. This is the subhalide process which involves only thermal reactions.11 During World War II Phillip Gross, working in England, had occasion to advise on attempts to make magnesium by reducing salt water magnesia with aluminium scrap, a process very similar in principle to the Pidgeon process I have just described. The aluminium and magnesia were heated together with some 2% of calcium fluoride which was used as a flux. The flux was there because metallurgists quite rightly believe in fluxes, much as ordinary people believe in luck; it was actually unnecessary on this occasion but its presence turned out to be singularly fortunate. When the mixture was heated at 1200°C, magnesium was distilled away and the spinel MgO.Al2O3 was rather wastefully produced. In his work to improve this process, Gross observed that the magnesium collected always contained a percent or so of aluminium. The operators thought this came over as dust or vapour, but Gross reckoned that so much could not be accounted for in this way. Gross was interested and puzzled and searched the literature for papers relating to the vapour pressure and evaporation of aluminium. It became clear that the aluminium was not vaporising directly in significant amounts. Gross then found Willmore’s 1939 US Patent12 for purifying aluminium from iron and silicon by heating it at 1000–1200°C

SEQ 0016 JOB IOM8627-001-009 PAGE-0016 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

16

Hatfield Memorial Lectures Vol. III

under reduced pressure with Alf3, MgF2, CaF2 or Na3Alf6. Under these conditions Willmore had found that the aluminium vaporised together with the halide, but without the impurities, and that both metal and halide could be recovered intimately mixed together in a condenser. They could then be separated from one another by fusion, an operation that in practice is exceedingly difficult, so the process is of no account. Willmore had supposed that ‘some low boiling mixture or compound was formed’ between the salts and the metal and that this explained the simultaneous vaporisation of the two components. Gross therefore produced some magnesium from salt water magnesia without flux, and found no aluminium in the metal produced. Recollecting that although the aluminium atom has three electrons in the M shell, two are S electrons and one is a P electron, he thought it possible that the metal could be monovalent as well as trivalent. He then thought that Willmore’s reaction and the one occurring in the magnesium distillation might be 2Al + CaF2 = 2AlF(g) + Ca(g) So he searched the literature for spectroscopic data and found that Miescher had studied the band spectrum of AlCl3, had allocated some bands to AlCl and had estimated its dissociation energy. From this information and the known thermochemical properties of AlCl3, Gross estimated the free energy changes for the reaction 2Al + AlCl3(g) = 3AlCl(g) and concluded that this could be made to proceed at temperatures of 1000–1200°C and reversed at low. Gross immediately conceived the idea of exploiting this reaction to make aluminium by a thermal route. He could see that an iron–aluminium alloy could be made from bauxite in arc furnaces in the same way as ferrosilicon was made for the Pidgeon process, and that the aluminium could then be vaporised from the alloy as AlCl by treating it with AlCl3, the iron remaining behind. He therefore conducted experiments in which he led unsaturated AlCl3 vapour through finely divided aluminium supported on alumina at 1000–1200°C, and successfully condensed aluminium from the emergent gas stream. The existence of the monohalide was thus fully confirmed together with his view that in Willmore’s experiments the emergent vapours had consisted of AlF and Mg, Ca, or Na which reverted to Al and MgF2, etc. on cooling: similarly the simultaneous volatilisation of AlF and Ca had been the cause of the carryover of aluminium in Gross’s magnesium distillation. The next move was to determine step by step the thermodynamics of the reactions,13 and these established the correctness of his estimates of the properties of AlCl. The thermodynamics of the aluminium–iron alloys that might be conveniently produced from bauxite had also to be worked out, together with the heat and mass transfer problems arising in the treatment of briquettes of Fe–Al alloy moving down a tower up which AlCl3 gas passes. The process is being tried by Alcan, and a plant with an estimated

SEQ 0017 JOB IOM8627-001-009 PAGE-0017 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

17

output of 8000 tons per year is being built. There are problems still to be resolved, and the process will have a hard time competing with the enormous electrolytic plants which now exist, but it seems to have a good chance of succeeding in the long run. So far as the conception of the process is concerned, the important points are the way Gross saw significance in a chance piece of information which others consistently neglected, how he deduced the existence of AlCl from the most basic of chemical principles, and how he then applied this in an utterly different context.

THE SHERRITT-GORDON PROCESS There is an extremely important hydrometallurgical process that springs directly from the discovery of new reactions. This is the Sherritt-Gordon process for the recovery of nickel, cobalt, and copper from the Lynn Lake ores of Saskatchewan. After mineral dressing, these provide a sulphide concentrate which contains some 12%Ni, 0.5%Co, 1.5%Cu and 30%Fe. The first part of the process was discovered by Professor Frank Forward, when he was Head of the Department of Metallurgy at the University of British Columbia. He had been interested in nickel ever since 1923 when he was an operator in a nickel refinery in Quebec. He had subsequently worked on the production of nickel in Japan and in 1945 he was working on the recovery of nickel from its oxide in Cuban laterites by the ammonia–ammonium carbonate leaching process. This involved prereduction of the concentrate to give nickel metal, followed by leaching in open vessels and boiling the leach solution to drive off ammonia and so precipitate basic nickel carbonate. In 1946 Forward heard of the Sherritt-Gordon discovery at Lynn Lake and he thought out a way of treating the concentrates, which was rather similar to the procedure for the Cuban ores. His idea was to roast to remove sulphur, then reduce the nickel, copper, and cobalt to metal, and leach with a solution of ammonia and ammonium carbonate. The copper had then to be separated, and a basic nickel carbonate produced, this finally being reduced to metal. He was strongly encouraged by the President of Sherritt-Gordon, Eldon Brown, to do everything possible to find a route different from established smelting and refining, and he produced a small amount of metal by the roasting–leaching route. The process was feasible but not economically attractive. Forward substantially gave up hope of success and most people would have abandoned the project at this point. But Forward was reluctant to give up work on ammonia leaching completely, so he pressed on hoping that it might somehow be useful for extracting metallic nickel from a high grade matte (such as that produced at Falconbridge), leaving the sulphides undissolved. In his next experiment therefore 5–10 g of powdered matte were placed in a Winchester with 500 mL of ammonia–ammonium carbonate solution. The air was displaced with oxygen, and the bottle sealed and rotated for two days at room temperature. At the end of that time it was found that the metallic nickel had dissolved as expected, but to everybody’s surprise, the nickel and copper sulphides had oxidised and

SEQ 0018 JOB IOM8627-001-009 PAGE-0018 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

18

Hatfield Memorial Lectures Vol. III

also gone into solution. The experiment was immediately repeated with the Lynn Lake concentrates; these behaved similarly; the nickel, copper, and cobalt went into solution, and the iron was left behind as a hydrated oxide. A most important new reaction had been discovered: as first conducted it was too slow for large scale use, but it was a reasonably simple step to increase the rate by raising the temperature to 83°C in the presence of an air pressure of 100 lb/in2. At the same time it was found that the ammonium carbonate was not required, because of the ammonium sulphate produced autogeneously during the leach, as shown in the equations below. It took some time to find the detailed course of the reactions and the kinds of ions produced. The most important steps may be represented stoichiometrically as follows14 2NiS + 8FeS + 14O2 + 20NH3 + 8H2O → 2Ni(NH3)6SO4 + 4Fe2O3.H2O + 4(NH4)2S2O3 2(NH4)2S2O3 + 2O2 → (NH4)2S3O6 + (NH4)2SO4 (NH4)2S3O6 + 2O2 + 4NH3 + H2O → NH4SO3NH2 + 2(NH4)2SO4 These discoveries opened the door to others and stimulated research into a whole new field of hydrometallurgy. V. N. Mackiw, Director of Research for Sherritt-Gordon, took up the work. He found that when the leach solution was boiled to remove some of the ammonia, considerable amounts of copper sulphide were precipitated. He then discovered that if sufficient thionate were present, it decomposed, as ammonia was removed, to give sulphide ions, and these precipitated most of the copper as sulphide. On this reaction the separation of copper in the full scale process is now based.15 The next question was how to recover the dissolved nickel and cobalt: the first idea was to produce crystalline nickel ammonium sulphate, to convert this to oxide and then reduce it to metal. At this stage the Chemical Construction Corporation was called in to discuss engineering aspects of the leaching process, because of their experience with ammonia plants. It turned out that they had been working on the deposition of nickel from ammonia solutions by hydrogen reduction. It had long been known that hydrogen under pressure would reduce metals from solution. There were, for example, the experiments of B´ek´etoff16 with silver salts in the Sorbonne in 1859; and the researches reported by Ipatiew17 from 1909 onwards. Sporadic interest in this type of reaction had continued, and in the 1930s I. G. Farben produced a patent for the production of nickel catalysts on silica by hydrogen reduction of ammoniacal nickel solutions. Schaufelberger,8 working for the Chemical Construction Company, had successfully produced nickel metal by using hydrogen to reduce nickel ammines dissolved in ammonia solutions. But he had difficulties with nucleating the new phase and unless he introduced powdered seed material, he obtained his metal as deposits on the walls of his reaction vessels. Such reductions may be represented by the general equation M2+ + H2 = M + 2H+

SEQ 0253 JOB IOM8627-001-009 PAGE-0019 16TH LECTURE REVISED 20DEC04 AT 21:50 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

19

This equilibrium is exceptionally amenable to manipulation: the activity coefficients of the metal ions may be varied by complexing (e.g. with ammonia), and both hydrogen pressure and pH may be adjusted over wide ranges. By this means it is possible to get separations of one metal from another which are extremely selective and impossible to obtain by other methods. With the chance of a large scale aqueous route for nickel opening up, Schaufelberger’s work became very important, and it was quickly followed by more intensive studies by Mackiw19 and his colleagues in Sherritt-Gordon. They worked out the chemistry in greater detail, but still used fine nickel powder to seed the reduction and make it go at reasonable rates. It was then discovered that some solutions would reduce satisfactorily without seeding. Careful analytical investigations showed that these solutions had low ammonium sulphate contents and were contaminated with iron. This further discovery led to the use of ferrous sulphate as a reduction catalyst, so that a very fine and active metal powder could be produced without seeding. It appears that under the conditions obtaining in these solutions, the ferrous sulphate gives a very finely divided ferrous hydroxide, which is responsible for the catalysis, and that this dissolves at a later stage as the concentration of ammonium sulphate rises. This third discovery was the result of careful experiment and observation. It could hardly have come in any other way, in view of the present state of theory concerning nucleation from solution. The hydrogen reduction process, which is operated at about 170°C and 500 lb/in2, now consists of producing very finely divided nickel in the presence of ferrous sulphate and using this as seed on which to grow nickel reduced from batches of solutions in the absence of ferrous ion. This latter step is termed densification and is conducted in the presence of stearic acid, which prevents the particles from agglomerating and coating the walls of the autoclave. Some 12 to 15 densifications, averaging 60 min each, are typically used. Two samples of nickel are shown in Fig. 5a and b. The reduction of cobalt is conducted in a similar manner, although different catalysts and anti-agglomerating agents are used. Some beautiful cobalt platelets obtained in this

Fig. 5 Particles of metal powder about 10–2 cm across, produced by hydrogen reduction, all from Mackiw et al.20 (a) nickel after 40 successive depositions from an ammine solution; (b) nickel of almost theoretical density produced as in (a) but with small amounts of anthroquinae present; (c) platelets of cobalt from an ammine solution.

SEQ 0020 JOB IOM8627-001-009 PAGE-0020 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

20

Hatfield Memorial Lectures Vol. III

way are in Fig. 5c. These powders can be roll compacted, sintered and then cold rolled to produce a satisfactory strip. The possibilities are not of course limited to simple metal powders, for ‘tailor-made’ alloys may be produced by coating one metal with another,20 and some of these could not possibly be made by a high temperature process: furthermore metals or alloys may be precipitated upon different kinds of nuclei which could act as dispersion hardeners. Particles of iron and phosphorus coated with nickel, are shown in Fig. 6a and b.

Fig. 6

Particles of (a) iron and (b) phosphorus coated with nickel deposited by hydrogen reduction.

Aqueous processes for treating ores or concentrates culminate in electrodeposition, hydrogen reduction, or precipitation of a pure compound. If this compound is subsequently reduced at low temperatures, a metal powder must result. Such powder may become an important starting material for making strip. The work of compacting powder may be much less than that of rolling down ingots. Furthermore alloy strip of novel compositions, which could not be prepared by melting, might be made by such routes. In the world of iron and steel, it could be that aqueous processes can only become competitive if the end product is a powder whose physical properties are exploited to the full in making strip or special powder metallurgical products. This is the principle that seems to be in the minds of G. R. Heffernan, President of the Peace River Mining and Smelting Co. and C. P. Gravenor21 of the Research Council of Alberta. Here, as in the Sherrit-Gordon process, the desire to exploit a new orebody has been the impulse for devising a new route from ore to iron powder and developing new steps within this route. No entirely new reactions had to be discovered, so that process engineering has presumably dominated the project. The Clear Hills iron deposits near the Peace River consist of quartz fragments surrounded by layers of goethite and green iron aluminosilicate. The iron content is some 34% with 27%SiO2, 5%Al2O3 and 0.6%P. The process hinges on partially reducing the ore and leaching with hydrochloric acid to give ferrous chloride solution. Most of the ferrous chloride is crystallised in pure form by partial evaporation of the solution, and then converted to FeCl2.2H2O. This is next reduced by hydrogen to give iron powder and a simultaneous recovery of the HCl. But the valuable halogen must also be recovered from the solution left after separating the

SEQ 0021 JOB IOM8627-001-009 PAGE-0021 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

21

ferrous chloride crystals. The existing Nordac-Aman process can be used for this purpose. It was originally designed for converting metal chlorides to oxides and for the recovery of hydrochloric acid from spent pickle liquor. The solution left after crystallisation is sprayed through oil or gas burners: the water evaporates and the solid chloride particles react at about 1000°C with steam. Hydrochloric acid gas is produced together with oxides of iron, aluminium, magnesium, and calcium which are all extracted in greater or less extent when the ore is leached. The reduction of ferrous chloride by hydrogen is conducted at 650°C in a packed bed or kiln. The reaction is strongly exothermic and half the heat required for the entire process is used in this step. The energy demands for the reducing roast, the crystallisation and reduction of the ferrous chloride and the hydrochloric acid recovery from the waste liquor, are so great that the process may not be economic except in a location where fuel is cheap and where the ore can be cheaply mined. These conditions are satisfied in Alberta where there is cheap natural gas and where these very large ore deposits can be worked by strip mining. The iron powder contains some 99.2%Fe, 0.05%P, 0.005%S, 0.03%C, 0.17%SiO2 and 0.75%Al2O3, and it can be rolled directly into strip. Thus a series of relatively low temperature steps supplants not only smelting and refining, but also casting and much of the mechanical working involved in the classical route from ore to strip. A pilot plant to produce some 15 tons of powder per day is now being built to test the process as a whole and the marketability of the product.

CONCLUSIONS How far is there any pattern in the origins of these diverse processes? Everything starts from the impulse to solve a problem, and in extraction metallurgy that impulse has most commonly come from the financial incentive to exploit a new orebody. We have seen this in four of the processes considered. But the impulse may also come from the desire to improve some existing and well established method of making metal. I call this the self-imposed problem, and the zinc– lead blast furnace is an example. For tackling this type of problem more enthusiasm and tenacity is required on the part of both the individuals and the companies involved. It is never so easy to pick worthwhile projects of this type, and the advantages which may come from success are never easy to assess. The opposition is usually great because it is so difficult to upset the complacent view that ‘we are getting along perfectly well as we are, thank you’. There are always financial reasons for not making changes, for this inevitably involves scrapping costly plant, and upsetting a smoothly working organisation. All the intertia of habit has to be overcome, and the new thoughts have to emerge without any spur from outside. In the development itself there are important principles which should be kept in mind. Some of them are almost platitudes, but these principles soon become overlooked once

SEQ 0022 JOB IOM8627-001-009 PAGE-0022 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

22

Hatfield Memorial Lectures Vol. III

the broad strategy of a development has been decided, and the detailed planning and execution is started. One must first exploit available basic knowledge to the full without putting undue faith in the data or the theories available. One must do experiments that have a chance of leading to discovery, and in this connexion any relevant experiment not previously attempted is worth doing once: one must make the most of all kinds of information and remember that acute observation can draw discovery out of accident or out of work and trials conducted for quite different ends. Success depends on invention when the route forward is blocked, and the logical extrapolation of existing knowledge can lead no further. Then one has to break out of current modes of thought and seek inspiration in quite different areas of science, technology or art. It is at this stage that weight of experience becomes a drawback and the newcomer to the problem has the great advantage of ignorance, the advantage of a mind uncluttered with objections and without a sense of impossibility. Tennyson wrote in ‘Ulysses’: I am part of all that I have met Yet all experience is an arch where thro’ Gleams that untravelled world whose margin fades For ever and for ever as I move. From this archway of experience we get a much-needed perspective, but from it we also get a restricted view. And when we have a great deal of experience, we run the risk of viewing the untravelled world as through the end of a long tunnel. A tunnel so long that sometimes the most experienced among us begin to think that everything has been done before and no untravelled world exists. The value of being able to bring ideas across from other fields of science and technology cannot be overstressed. This is where a multi-disciplinary training has a real advantage over narrower specialisations. This is one reason why I am so keen to see metallurgy taught in breadth. The metallurgist should have at his disposal as wide a range of basic ‘tools of thought’ as he can get. After university it is extremely difficult to acquire these. There is never sufficient time and most industrial and research experience tends to force the really successful man to become more and more specialised. For this reason I feel strongly that there are great advantages in a metallurgical training which recognises the anatomy of the subject as represented by my two triangles and gives the student a sound grounding in all three basic sectors. We should remember this need to bring new ideas across from other fields when we plan the activities of our metallurgical institutes. In terms of ideas it is a grave disadvantage to have our ferrous publications separated from non-ferrous. It is bad for students and bad, I suggest, for the old and experienced. I know there are strong organisational reasons for not making a change. But we could make positive attempts to offset the bad effects of segregation by organising our discussions more imaginatively. Instead of confining, for instance, our own Institute discussions to a group of papers published in our own Journal, we could bring together for discussion the authors of a group of papers which have some common interest, although scattered through all three British metallurgical journals, and

SEQ 0023 JOB IOM8627-001-009 PAGE-0023 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Basic Knowledge, Discovery, and Invention

23

some of the chemical engineering and chemical publications also. I think, we would have all the advantages of a symposium without the tiresome business of pressing authors for special papers produced against the clock, and I am sure we could have some much livelier discussions. But there are other ways in which positive action could be taken to pave the way for new developments, and to make discovery and new ideas more likely. New chemical reactions involving metals in molten, aqueous, and organic phases should be sought and studied irrespective of any application. In the field of iron and steel work of a speculative nature needs to be undertaken to test the potentialities of halide, aqueous, and powder routes. One cannot tell where such work may lead, but there is a worthwhile chance that useful new products and processes may emerge. Basic knowledge in chemical and process engineering metallurgy needs to be built up where it is scanty, and where more detail looks like leading to interesting theory or to possible applications. Work in the field of process engineering needs to be supported by large scale experiments. Some national facilities are needed to which academic investigators should have access. The rates of processes, the patterns of heat and mass transfer, and sometimes the order in which reactions proceed in a process, are dependent on scale. Process engineering problems cannot be studied solely in the laboratory or on the very small scale which is adequate for investigating the thermodynamics and chemical kinetics of metallurgical reactions. Small scale tests may frequently be irrelevant and the results sometimes quite misleading. Much of the work that I have mentioned can be done in or organised by universities: it can probably be done most cheaply in this way. But when we look at research in British universities in the field of metallurgy, what do we see? Less than 10% is going into areas related to metal production (that is chemical and process engineering metallurgy), and over 90% is devoted to physical metallurgy. This is a gross imbalance. It is an imbalance that has, I am glad to say, now been recognised by the DSIR who are striving to redress it. Much can be done by industry and government to bring about a change. Contacts can be improved between the metals industries and the universities; academic posts, fellowships, bursaries and research grants, can be set up specifically for work in this field, for no scientist or engineer can work effectively in isolation. I regard four experienced persons as the minimum for a viable group, and the interests of these four need to be spread over both chemical and process engineering fields. Tonight I have attempted to look primarily at the interplay of the abstractions basic knowledge, discovery, and invention in the origins of processes. But I have not forgotten that the exciting game of developing processes is a ‘flesh and blood’ business. It requires forceful people with enthusiasm almost to the point of obsession. The kind of enthusiasm that was displayed by Sydney Thomas, that most lively inventor, and which shines forth even today in the message which he wrote at the end of a letter to his sister2 in 1883: By the way, dear child, you have still got to learn some chemistry and work with me. I am absolutely brimming over with things that demand investigation; the lines are

SEQ 0024 JOB IOM8627-001-009 PAGE-0024 16TH LECTURE REVISED 19DEC04 AT 20:11 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

24

Hatfield Memorial Lectures Vol. III

already laid down and they must be investigated, I shall never have time by myself and you must help; you can’t tell what a glorious, entrancing, delightful occupation it will be, with rewards of the most magnificent description in reputation, work, benefits and lucre.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.

W. H. HATFIELD: Application of Science to the Steel Industry, 1928, ASTM. R. W. BURNIE: Memoirs and Letters of Sidney Gilchrist Thomas, Inventor, Murray, London 1891. F. D. RICHARDSON: Trans. AIME, 1964, 230, 1213. S. W. K. MORGAN: Trans. IMM, (London), 1956–7, 66, 553–565. C. G. MAIER: US Bur. Mines Bull., 1930, 324. T. E. SUESS: V¨oest Jahrbuch, 1950/51, 21. H. HELLBRUGGE: Stahl Eisen, 1950, 70, 1208–1211. P. METZ: Pat. 551., 946, 19.10.1956. F. D. RICHARDSON: Iron and Coal Trades Review, 1961, 1105–1116. L. M. PIDGEON and W. A. ALEXANDER: Trans. AIME, 1944, 159, 315–352. P. GROSS: US Pat. 2, 470, 306 (17.5.1949). C. B. WILLMORE: US Pat. 2, 184, 705 (26.12.1939). P. GROSS: Disc. Faraday Soc., 1948, 4, 206–215. F. A. FORWARD: Can. Min. Met. Bull., 1953, 56, 363–370. V. N. MACKIW et al.: Chem. Eng. Prog., 1958, 54, 79–85. M. N. BEKETOFF: Compt. Rend., 1859, 48, 442–444. W. N. IPATIEW and W. WERCHOWSKY: Berichte, 1909, 42, 2078–2808; 1911, 44, 1755– 1758. F. A. SCHAUFELBERGER: Min. Eng., 1956, 8, (5), 539–548. V. N. MACKIW et al.: J Met., 1957, 789–793. V. N. MACKIW et al.: ‘Practice and potential of pressure hydrometallurgy’, IUPAC Congress, Montreal, 1961. C. P. GRAVENOR et al.: Can. Min. Met. Bull., 1964, 421–428.

SEQ 0025 JOB IOM8627-002-003 PAGE-0025 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE NINETEENTH HATFIELD MEMORIAL LECTURE

Twenty-Five Years On H. M. Finniston At the time the lecture was given Dr Finniston was Deputy Chairman-Technical of the British Steel Corporation. The lecture was given at Sheffield University on 11 December 1968.

The title of this lecture is intentionally unspecific, even ambiguous, because when an author has to concern himself with the content much ahead of delivery, he is uncertain of the scope and even more of the detail of what may ultimately emerge. The options thus left open in the title through indecision, however, have the advantage of possibly attracting a larger audience by leaving each to put his or her own interpretation on what is intended, although one suspects that the success attending this temporary device is more probably a direct tribute to the man to whom this event is a permanent memorial. ‘Twenty-five years on’ is really compounded of two different eras. Twenty-five years have passed since William Herbert Hatfield died in October 1943 and the contrast between the iron and steel industry of his era and the present day is therefore appropriately caught in the title; the ambiguity concerns itself with an ill defined 25 years from now and the iron and steel industry which might face the Hatfield lecturer of the year 2000, who today is either just completing his postgraduate studies or just entering the metallurgical industry.

RESEARCH THEN . . . What would Hatfield think about the industry he served so well were he to return today to assume the prominent place he occupied in the late 30s and early 40s? Being oriented towards research (but not constrained by it, however) it would be to the changes in the research structure of the industry and its future thinking he would look first. Although the industry acknowledged in principle the importance of research to its continued well being, its laboratories were mainly concerned with trouble shooting and work directed towards short term quality improvement and control; laboratories which consciously and recognisably devoted a significant part of their effort to research in the broader sense were limited to very few companies. One of these was Hatfield’s own: the Princess Street Laboratory of Thomas Firth & John Brown Ltd opened in 1908; another was the United Steel Companies Ltd Central Research Laboratories which, under Dr Thomas Swinden, started at Stocksbridge in 1934. Research workers in the steel industry

25

SEQ 0026 JOB IOM8627-002-003 PAGE-0026 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

26

Hatfield Memorial Lectures Vol. III

outside the privileged few had to seek facilities in routine test laboratories or permission from management to do experiments on the shop floor. Hatfield, though an independent, recognised that in a technologically burgeoning industry he could not go it alone. He was, therefore, faute de mieux a strong proponent of collaborative research. The main channel for this form of enterprise was the Iron and Steel Industrial Research Council set up in 1929, whose Secretary, Edgar C. Evans, and a staff of half a dozen or so, also doubled as the Technical Department of the British Iron and Steel Federation. This organisation serviced a range of projects with a total budget of £65,000 at the time of Hatfield’s death; at today’s money values this would be equivalent to about £140,000. The Council had no laboratories directly responsible to it, but relied upon access to occasional facilities for construction and calibration of equipment for works trials at one of the universities to which part of the programme was contracted. The actual expenditure on research, however, was many times this budget, because this was the period of programme delegation by committee, which confused accountancy. The famous Heterogeneity of Steel Ingots Committee (the pioneers of them all), the Alloy Steel Research Committee, the Corrosion Committee, and the Steel Castings Research Committee had been joined by the Blast Furnace and Open Hearth Committees, all with their many sub-committees steering programmes entered into cooperatively and with specifically defined parts apportioned for action among the industrial members. Hatfield was a master at chairing and (less frequently) serving on these committees. A complex series of chemical analyses, a large amount of polishing and etching, or that supremely expensive business of sectioning an ingot or two: such tasks were distributed with an authority which left little room for dissent. The prosecution of experimental work is not an end in itself, however. The conclusions to be drawn, the lessons to be learned, and the implementation of the new knowledge in practice are the raisons d’etre for industry engaging in research development. Hatfield was as concerned with these aftermaths of the experimental work generated by him as in the initiation of the programme, and he showed the same authoritarian attitude towards gaining as much information out of the work as possible. As his obituary notice in the Royal Society archives reports: ‘He would occasionally argue obstinately against the acceptance of conclusions on which most of the members (of his committees) were agreed’, but he did this to improve the work rather than his reputation. It would be rash to assert that the value to the industry of results achieved through the cooperative practices of Hatfield’s time has been any less than are the much larger sums spent today in individual laboratories on self-centred projects. The pattern of laboratories and of the practices of Hatfield’s time can still be recognised today, even if some radical changes have been superimposed. The Iron and Steel Industrial Research Council was replaced by the British Iron and Steel Research Association in 1944, and in 1967 BISRA became the British Steel Corporation’s Inter-Group research organisation with four major laboratories, specialising in different fields of research and with an annual budget of £2m. With the British Independent Steel Producers

SEQ 0027 JOB IOM8627-002-003 PAGE-0027 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On

27

Association, however, the Corporation retained close contacts for continuing collaboration, using BISRA as its principal. New research laboratories costing £700,000 were opened in 1967 by Firth Brown; the United Steel Companies Stocksbridge Laboratories, although still in use, were replaced in 1950 by the extensive laboratories in Rotherham commemorating Swinden’s name; but as important as these are, there has been since Hatfield’s day a commendable increase in the numbers of research laboratories built by companies other than these two and now with considerable and established reputations. With nationalisation of the industry, many of these have become part of the assets of the Corporation, which is now faced with the problems of coordinating the research activities of 14 companies, some with research laboratories each of which spent more than a million pounds a year and others with much smaller investments in this area. Hatfield may not have approved of nationalisation but he would have welcomed the opportunity thus afforded for encouraging collaboration in research! In Hatfield’s day there was also limited support of a few favoured universities (mainly in the steelmaking districts) for fundamental research on ferrous metallurgy, notable examples being the work of Hay at the Royal Technical College, Glasgow (now the University of Strathclyde), of Edwards and of Luther Phillips at Swansea, and of Andrew at Sheffield. Of 22 universities in the UK in Hatfield’s time, nine boasted Chairs of Metallurgy and departments offering metallurgy as a first degree subject; now, among 44 universities, there are 22 Chairs of Metallurgy with some departments carrying two (or more) professors. Notwithstanding this growth, there may be some moral to be drawn from the fact that there are still today more professors of theology that there are of metallurgy.

. . . AND FOR THE FUTURE Depending upon how it is measured, research expenditure by the industry is running at a level of about £12m a year; the Science Research Council alone contributes some £500,000/year to ferrous metallurgy research in the universities. Will there be a continuing increase in this total expenditure? In some quarters, a figure of 1% of turnover has been suggested as a target for the research expenditure of the steel industry, and such a target would leave room for some expansion of present effort. It would be surprising if, during the next few years, the fields of our researches had not to be extended in keeping with increasing scientific growth within and outside the industry. This may well be achieved, however, without a corresponding increase in cost through a more critical attitude to research programmes and the freeing of existing resources to tackle new fields. In industry 90% of research expenditure has to be looked at with a close eye to what is really needed and can be justified rather than with an eye peering through rose coloured spectacles at some arbitrary target. The remaining 10% of funds has to be invested with faith in people who have the qualities of imagination, foresight, resourcefulness, and research ability; these are much fewer in number than self-opinionation would have us believe.

SEQ 0028 JOB IOM8627-002-003 PAGE-0028 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

28

Hatfield Memorial Lectures Vol. III TOOLS OF THE TRADE

Hatfield was a man and metallurgist of parts, interested in the properties of steels whether these were mechanical, physical, or chemical and whether viewed from the scientific, technical, or practical aspects. In these areas there are three features which Hatfield would now find different. Firstly, were he with us today, he would revel in the new techniques and facilities which could be applied to solving and understanding the problems of iron and steel and to assisting in the further development of the industry. Much research depended, and still does depend, on large numbers of analyses. In Hatfield’s day this work was limited by the sheer physical capacity of laboratories to carry these out by the conventional wet methods of the time. Now the automatic recording spectrometer can deal with many of the elements of conventional analysis in an astonishingly short time, with remarkable ease, accuracy and directness of presentation, and sophisticated instruments for oxygen analysis are threatening the vacuum fusion apparatus to which Hatfield’s laboratory made important contributions. Inclusion counting and discrimination, which used to be a wearisome business involving subjective assessment by optical microscope and confirmation by laborious extraction methods, is now performed by the automatic counter and the electron probe microanalyser. The optical microscope was then the basic tool of the physical metallurgist but had not yet been developed to provide phase contrast, ultra-violet and dark field illumination and interference techniques. Structure analysis by X-ray was well known in Hatfield’s time (the pioneering work of Jay at Stocksbridge in the early 30s is still of classical dimensions) but the equipment is now much more refined, versatile, automated, and quicker in its provision of results. The electron microscope was being developed in Hatfield’s day in the USA, and the Journal of The Iron and Steel Institute of the period has a number of abstracts of American papers describing early work with this then revolutionary device. Today, to back up these techniques, there are developments in emission and mass spectroscopy, X-ray fluoresence, activation and atomic absorption analysis, emission, scanning and transmission electron microscopy, electron and neutron diffraction and field ion microscopy. All these techniques and the instruments associated with them have been in the direction of greater refinement in understanding what steel is: from the macrostructural characteristics as explored by Hatfield’s Committees to the microstructural substructures shown up by the electron microscope and the pin-pointing of individual atoms by the field ion microscope. So powerfully, for example, has electron microscopy developed as a technique for metallurgists that the British Steel Corporation in association with the Science Research Council is investing in a 1 MeV electron microscope to be housed in the Corporation’s laboratories at Sheffield for use by the steel industry in concert with the Universities of Sheffield, Newcastle, Manchester, Leeds, and Liverpool. The lesson to be learned for the future is that today’s research investments are tomorrow’s tools of the trade. This general refinement of experimental method applies to works trials as well as laboratory practices. Some of my present colleagues remember with pleasure, mainly retrospective, the works trials in Hatfield’s day. Typical of these were the waiting through

SEQ 0029 JOB IOM8627-002-003 PAGE-0029 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On

29

the small hours beside an ingot mould stuffed with thermocouples for a tap which was delayed; the probing of an open hearth furnace uptake with a water cooled pitot tube, whose apparently erratic behaviour gave the first clues towards the gas flow phenomena later investigated so thoroughly by Dr Chesters (your President) and his school; and the measuring of cold blast temperature hour by hour by uncorking a hole drilled in the cold blast main and pushing in a mercury thermometer against the blast pressure because funds did not run to a recorder. Today, the technical parameters of works trials are no longer measured by laborious manual recordings of U tube or thermocouple readings, or results analysed by applying a planimeter to a fuzzy line drawn by a flowmeter. Data loggers are now plugged in and the subsequent analytical calculations performed on-line by the ubiquitous computer. Although he might be somewhat critical about the way in which research workers can now get by with much less toil in the way of developing instruments, Hatfield would be persistent in ‘arguing obstinately’ for money towards still further improvement in the facilities available for iron and steel research. A second feature of the climate of Hatfield’s time was that the mechanical properties of steel were confined to a determination of the proof stress, maximum stress, elongation, reduction of area and impact value, creep (generally stress rupture or by Hatfield by his own time yield) and occasionally fatigue. Today, to these would have to be added such concepts as transition temperature (or its variations, isothermal crack arrest temperature or nil ductility transition) and the whole new field of fracture mechanics. The latter augments the research for improved elastic and plastic behaviour by seeking to limit disposition to fracture. The approach of fracture mechanics relates an experimentally determined parameter of fracture toughness (Ke for high strength, K1e for lower strength materials) to the limits of size and shape of crack or defect so that propagation will not occur, at the chosen design stress. An important implication of this approach is that steel (or any other metal) is not considered or treated as an ideal material free from defects; on the contrary it is recognised that even the best steels (and the best of other metals and alloys) have defects, e.g. fine cracks, inclusions, etc., which can be lived with to a greater or lesser degree depending upon the characteristics of the defects and the stresses and environment in which the steel finds service. Hatfield never acknowledged that steels made by his firm contained even the finest defects. There is a story which I hope is not apocryphal that during an epidemic of hairline cracking in alloy steels, Hatfield was asked by a research worker from outside the firm for a sample containing such cracks. Hatfield replied that he could not provide the sample because the steels made by his firm were entirely free from them and recommended that the enquirer direct his request to a competing company (needless to say now in the public sector). Through the science of fracture mechanics Hatfield today could acknowledge that while defects did occur in steels made by his firm he could still retain the firm’s good name through the claim that these defects and inclusions were fewer and smaller than those in steel made by his competitors. With interest now moving towards understanding and controlling fracture through theory and experiment, a new science of relating the detailed topography of fractured

SEQ 0030 JOB IOM8627-002-003 PAGE-0030 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

30 Hatfield Memorial Lectures Vol. III surfaces to the structure of the material and its fracture toughness should make ground. Figure 1a shows the detail of a tensile testpiece fracture surface as photographed through an optical microscope compared with the startlingly greater visual information shown by stereoscan electron microscopy. The differences in appearance (and hence propagation) of ductile and brittle fractures are to be seen by comparing 13Fig. 1b and c. For the future, low energy electron diffraction combined with Auger emission spectroscopy or field emission microscopy associated with field ionisation time of flight mass spectrometry might well be developed to determine the composition of single atomic layers on these fracture surfaces.

Fig. 1 (a) Fracture of tensile testpiece, optical magnification (×20); (b) stereoscan micrograph of same fracture showing ductile dimpled fracture (×2000); (c) stereoscan micrograph of brittle fracture showing cleavage facets (×600).

Thirdly, relationship between mechanical properties and microstructure has been recognised since the earliest use of the optical microscope by metallurgists in the late 19th century; microstructural features, such as martensite, twinning, etc., constituted pattern recognitions associated with qualitative assessments of strength and ductility. These patterns have since been analysed in greater detail optically and by electron microscope. Illustrative of the finer detail to be gained from microstructural examination by modern microscopy is the comparison in Fig. 2 of a quenched and tempered Cr–Mo–V steel

SEQ 0031 JOB IOM8627-002-003 PAGE-0031 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On

31

Fig. 2 (a) Quenched and tempered 12%Cr–Mo–V steel optically magnified (×750); (b) electron magnified by 30,000 showing larger alloy carbide precipitates at grain boundaries and smaller carbides within the grains.

structure photographed optically at × 750 and the same structure at × 30,000; the latter shows molybdenum carbide at the martensite plate boundaries and smaller carbides within the grains where optical microscopy had failed to detect either feature. Surprisingly, Hatfield made no reference to dislocations, although these had been reported upon in the literature since 1934. The concept and properties of this submicrostructural condition and the influence of dislocation patterns on the strength and ductility of steel and development of the microstructure, including recrystallisation, transformation, precipitation, etc., on heat treatment and on the deformation characteristics in service either escaped Hatfield or was considered by him as an unnecessary complication of limited significance. Today’s literature might convince him otherwise.

MAKING RESEARCH PAY While recognising the industrial value of the work done in Hatfield’s time, the fields which were tackled outside the laboratory were confined to improvement of then

SEQ 0032 JOB IOM8627-002-003 PAGE-0032 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

32

Hatfield Memorial Lectures Vol. III

existing processes rather than the invention of new processes. The measurement of simple blast furnace conditions gave some clues leading to the appreciation that good blast furnace performances depended on an evenly sized burden and the elimination of fines; studies of gas flow, air/fuel ratio, and the first example of true automatic control using feedback, i.e. automatic draught control, were yielding marked improvement in performance of open hearth furnaces; and the theory of metal deformation in rolling was being worked out. The development of major new production processes such as spray steelmaking, compact annealing, or the various new methods for applying metallic or organic powders to a steel base, the manufacture of plastic coated steel or the ‘Flox’ process for making pure oxide from lean ore were unsuited to the period, not because technical competence was lacking but because the scale of new developments was determined by the physical scale of the standard laboratory. This absence of pilot plant scale research was one of the less satisfactory aspects of Hatfield’s time and to a lesser extent still is. The excuse usually given was that in prosperous times the industry was too busy to engage in pilot plant work, and in less prosperous times it could not afford to do so. In recent years, BISRA/IGL and some of the more researchminded companies have extended the scale of pilot plant researches considerably, and any significant increase in the research and development expenditure of the British Steel Corporation is likely to be in this direction. The difficulties of organising this scale of experiment as a lead in to commercial production are well recognised; in many respects the approach to date from the development phase to industrial practice has been too gradual or restricted by the precedence given to the immediate needs of existing plants. The demands of modern competitiveness may well require new approaches to this conflict between present and future. To avoid the competing claims of technologists and management of production units, one approach might be the treatment of pilot plants as potential profit centres with high risk rather than as exercises to eliminate technical uncertainties or establish procedures.

INFORMING THE PUBLIC As one in charge of an important research laboratory and with a sense of public relations and its value, Hatfield might be less staggered today at the mass of scientific papers published, the plethora of technical conferences to attend, and the invitations he would be receiving to speak. He himself wrote about 150 scientific and technical papers (some in joint authorship) and would have regarded with distaste any suggestion of arbitrarily rationing publication. His obituary notice says he took ‘great pains in the preparation of [his] papers and the tables accompanying them’ and he would have been one of the first to protest at the amount of what can charitably be called ill prepared information which, along with papers of undoubted merit, reaches the printed page today. The recently announced intention of The Iron and Steel Institute to charge on a page basis for printing may discourage premature publication and encourage the brevity Hatfield sought.

SEQ 0033 JOB IOM8627-002-003 PAGE-0033 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On

33

It is interesting to compare the Journal of The Iron and Steel Institute for the first part of 1944 (which, because of the time lag in publishing will outline the research actually in progress at the time of Hatfield’s death) with the first six months of the Journal for 1968. In such areas as metallography, physical metallurgy, general works practice and analytical subjects, the pattern of papers is similar. However, there are noteworthy differences. In the early 1940s there was still active work on the sectioning of large ingots, and the examination of their macrostructure and segregation pattern. This type of investigation is very infrequently engaged in these days; this is partly because the earlier work established the general features of segregation and what could and could not be done by various measures such as changing ingot shape, and partly because useful information can now be got in cheaper ways by the examination of blooms and billet ends or by corner sampling. Much attention was being paid to various ways of measuring liquid steel temperature, the immersion thermocouple now universally used in various forms not having been established to the same extent. Steelmaking practice was concentrated on the OH furnace, and there were papers on slag control, on the use of mixed gas for heating, and on thermal efficiency. The chemistry of rimming steel was the subject of several papers. Few papers in these fields now appear in the literature. On the other hand, this year’s Journal contains numerous papers on topics which were not heard of in Hatfield’s day. Of some ten papers on practical steelmaking which appeared in the first six months of this year, only two were largely concerned with OH furnaces; the rest related to basic oxygen furnaces and in one case the fuel–oxygen–scrap (FOS) process. Continuing casting and automation, not in the subject list for 1944, are reported upon. The 1968 Journal has several papers with a bias towards economics and costs, matters condemned as far too dangerous subjects for open discussion by metallurgists 25 years ago. The present benefit/cost analysis approach to major research and development programmes and the closer integration of research with the market needs of the future suggest that these aspects will gain greater space in the journals of the latter part of the century.

THE STEEL INDUSTRY: STATIC OR DYNAMIC? As a director of a manufacturing company, Hatfield was a business man who had a concern for production and production techniques. How would he view today’s achievement in these areas? Figure 3 shows how both world and UK steel production have grown in the last 25 years. The figures speak for themselves, but two features might be pointed out. Firstly, world production has continued on an upward trend since World War II at roughly 5%/annum compound. World steel consumption per head, however, is now running at about 150 kg per head of population whereas in the most advanced industrial countries it is around 650 kg per head. The upward trend of world steel consumption is therefore likely to continue for many years yet. Extrapolated to 25 years from now, the

SEQ 0034 JOB IOM8627-002-003 PAGE-0034 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

34

Hatfield Memorial Lectures Vol. III

Fig. 3 UK and world steel production 1943–1967.

world will have an output of 1800 m tons/annum which in relation to then world population would amount to 300 kg/head. The second point is that the UK share of world steel production has halved in the same period. It may be thought that we should not have expected to preserve our 10% share of the early post-war years as the developing countries (particularly Japan) increased their own steel production; that our share has halved, however, does surely point a warning finger to our competitiveness which we should be very foolish to ignore. One of the interesting developments since Hatfield’s day has been the change in pattern of steel product demand. Table 1 compares the pattern in 1943 with that of 1967. Rails, forgings and castings have declined and flat rolled products (sheet, tinplate, etc.) have boomed with the growth of the canned food and motor car industries. Hatfield Table 1

Changes in pattern of steel products made in 1943 and 1967.

Product

1943 ’000 tons

1967 ’000 tons

Ratio 1967–1943

Crude steel Rails, sleepers, and fishplates Plates (3 mm thick and over) Sections Arches and accessories Hot-rolled strip Cold-rolled strip Bright steel bars Sheets, uncoated Tinplate, terneplate, and blackplate Forgings (excluding drop forgings) Steel castings Alloy steels

13,031 354.0 1,836.3 1,244.8 163.7 588.8 230.5 412.7 993.3 532.1 233.9 365.1 245.5*

23,895 280.8 2,997.9 2,471.9 349.4 1,568.2 503.6 512.2 3,067.0 1,237.1 156.5 245.8 1,041.9

1.83 0.79 1.63 1.98 2.13 2.66 2.18 1.24 3.09 2.32 0.67 0.67 4.25

*1946: first peacetime year

SEQ 0035 JOB IOM8627-002-003 PAGE-0035 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On

35

would have been particularly interested to see the growth of alloy steel, since his reputation (and that of his firm) will long rest on association with stainless steel developments and to a lesser extent on his own developments of the 12Ni–5Mn–4Cr non-magnetic aero engine valve steels and the 5Ni–4Cr–3Mo for hot shears, and other applications for which thermal shock resistance is needed. The examination of how prices have moved since Hatfield’s time is sobering. Table 2 shows how the prices of some of the main resources used in making steel have increased, compared with the price of steel and the cost of living generally. The prices of most of the resources required for iron- and steelmaking have increased by between 3 and 4 times. A marked exception is imported ore where the availability of plentiful new deposits and lower shipping freights have helped keep prices down. The price of steel has increased by less than three times, which is most creditable when one considers the improved quality and performance one expects from similar articles of 25 years ago. Nevertheless, steel prices have increased by more than the general level of prices as reflected in the purchasing power of the pound, suggesting that if the steel industry is to maintain its position it must become even more efficient. Table 2

Typical increases in price of the main resources used in making steels. 1967–68 price as multiple of 1943–44 price

Home ore Imported ore Inland carriage of ore Coking coal Coke Scrap Wages Capital costs Ingot costs Board of Trade index of price of steel Reciprocal of purchasing power of pound

3.6 1.0 3.0 3.8 3.6 3.1 4.3 3.2 3.6 2.8 2.1

IRONMAKING Hatfield, although not directly concerned with ironmaking, was nevertheless interested in the whole spectrum of technology. In other places I have referred to the dramatic improvements in blast furnace practice and performance which have occurred since the war, but when we go back to Hatfield’s period the changes in ironmaking standards are even more marked. Arthur Dorman, in his Presidential Address to The Institute in 1944, said ‘the average weekly make per furnace [which at that time was about 1,300 tons] should be 3,000 tons at least, or considerably more if more adequate supplies of rich foreign ore were available, or some schemes of treating lean home ores became practicable’. Today, a good blast

SEQ 0036 JOB IOM8627-002-003 PAGE-0036 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

36

Hatfield Memorial Lectures Vol. III

furnace should produce 5,000 tons a day, some 12 times Arthur Dorman’s modest target of 1944. While much of the improvement has been due to the application of sound chemical engineering knowledge and will continue from this, it is only fair to point out that at the time of Arthur Dorman’s prophecy, rich imported ore was scarce, and it was thought this country would be much more dependent on lean home ores than has proved to be the case. Nevertheless, other shibboleths have had to be broken. For example, we are now all convinced that ore must be 100% prepared, if not sintered, but in a paper in June 1943 on ‘Blast furnace and steel plant’ it is stated categorically that no significant advantages accrue from going beyond 45% sinter! Many other defined limitations of the blast furnace have gone and are going. Table 3 shows the capability and performance characteristics of blast furnaces in the UK scene then, now, and (at a guess) in the future. The last column of Table 3 has been put in to answer a likely question by Hatfield ‘How far can we go?’ While it is dangerous to prophesy about blast furnaces, Table 3 does show what a long way technically we can still go in improving performance and, when one considers the enormous gap which exists between the average and the best, this branch of our technology affords one of the greatest challenges to the industry. Our average daily output is 830 tons (and the US about 1,600 tons) compared with the world best approaching 6,000 tons. Although size differences play a part, there is still room for improvement since our biggest (31 ft) furnaces record output (over a month) is only 2,700 tons/day. If our blast furnace practice were brought up to the standards of present day best world performance, some 80% of our existing furnaces could be dispensed with. To replace all 82 blast furnaces, which in 1967 provided for a steel output of 24 m tons by 14 modern blast furnaces to make iron for 36–38m tons of steel would mean a capital investment of some £300m. Although some ‘direct reduction’ processes were known in Hatfield’s time, the issue was not as live as it is today. Despite the great surge of activity in this field during the last 25 years in this country, the blast furnace has remained one economic jump ahead of Table 3

Blast furnace technology of UK industry. 1943

1967

No. of furnaces capable of being blown

141

82

Largest furnace in UK (hearth diameter), ft (metres)

22 (6.7) 7.2 66,900 3.10 10.6 1.22 .. 1.22

31 (9.5) 15.2 275,100 1.94 63.8 0.66 0.02 0.68

Production of iron, m tons Production per furnace year, tons Average burden weight, ton/ton Sinter or pellets, % Coke consumption, ton/ton Coke equivalent of injected fuel, ton/ton Total coke equivalent, ton/ton *assuming no change in basic technology

Sometime in next 25 years* 14 (would suffice for projected output below) 33–39 (10–12) 22 1,600,000 1.5–1.6 100.0 0.45 0.11 0.56

SEQ 0037 JOB IOM8627-002-003 PAGE-0037 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On

37

direct reduction. At today’s prices, coke constitutes some 40% of the cost of iron and iron 25% of the cost of finished steel. In the UK the price of coke in relation to the price of other fuels remains reasonably favourable to coke. Therefore, unlike the position of some other countries where alternative fuels (natural gas and oil) are very cheap, the influence of direct reduction on our practice is likely to be limited in the next few years to the import of some pre-reduced iron sponge or pellets in lieu of conventional scrap for electric arc furnaces, or to boost blast furnace output over short periods when iron may be scarce at a particular plant, e.g. during a blast furnace reline. In overseas countries where natural gas or oil has economic advantages direct reduction may be attractive economically and could take the place of the blast furnace.

STEELMAKING Twenty-five years ago, with the exception of the Bessemer plants at Corby, Ebbw Vale and Workington, the OH furnace had a practical monopoly of the manufacture of common steels. OHs accounted for 91% of the steelmake compared with 67% today, and electric furnaces were restricted to melting of the more expensive alloy steels which could carry the cost of the power. The decline in OH furnaces is due to their partial substitution by oxygen converters in hot metal shops and by electric arc furnaces in cold metal shops. Table 4 summarises the marked reduction in the number of OH furnaces and converters in the last 25 years and how there could be an even more dramatic reduction in future. OHs and converters are grouped together as non-electric steel furnaces to allow the picture then and now to be more easily summarised. (Electric furnaces have been excluded because the smaller requirements of special and alloy steel manufacture will require the retention of such furnaces in comparatively small sizes and correspondingly larger numbers.) The most remarkable change in steelmaking since Hatfield’s time is the arrival of the oxygen furnace which now accounts for about 22% of our total steel production and could make at anticipated rates of replacement and investment up to 70% of our steel in 1980. This is such familiar ground that one need not dwell on this subject except to say Table 4 Steel furnaces (other than electric) of UK industry. Sometime in next 25 years*

1943

1968

Number in active existence

452

253

Total non-electric steel production, m tons

12

Average production per furnace in active existence, tons/year Largest capacity unit Highest average, tons/h (est.)

27,650

22 (est.) 87,000

14 (would suffice) 28 (est.) 2,000,000

300 (OH) 25 (OH)

400 (OH) 120 (LD)

300 (LD) 400 (LD)

*assuming no change in basic technology

SEQ 0038 JOB IOM8627-002-003 PAGE-0038 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

38

Hatfield Memorial Lectures Vol. III

that the LD converter will prove able to make the full range of common and lowalloy steels required, and that this will be the only type of converter likely to be built in the immediate future. Assuming a wand could be waved over the UK steel industry then, of a total of 38m tons of crude ingot steel in, say, 1980, about one-quarter will be made in electric arc furnaces (to take advantage of scrap availability) and the remainder in 300 ton converters. Only 14 such furnaces (of which nine would be working at one time) would be needed compared with 253 active OH furnaces and oxygen converters in use today. Another 22 or so 200 ton electric furnaces would be needed for cold metal works. Thus the whole of say 33m tons of tonnage or non-special steels could be made in less than 40 furnaces. There is a risk in suggesting that oxygen furnaces will not get significantly bigger than 300 tons, but the savings in operating costs to be realised by increasing oxygen furnace size must over this seem small at present in relation to the inconveniences of operating larger units. There has been limited improvement in performance of OH practice compared with that of 25 years ago. The best OH furnaces in the UK today average about 38 tons/h output compared with 15 tons/h in Hatfield’s time. However, fuel consumptions have improved only from a ‘best practice’ of 45 therms/ton to one of around 40 therms/ton, ignoring the special case of the Ajax furnaces at Scunthorpe which, using 100% of hot metal, are more akin to oxygen furnaces than to the conventional OH. OHs are now driven harder, as the increase in furnace output shows, and hard driving leading to higher waste gas temperature is not conducive to thermal efficiency. Further, more low carbon steels requiring a higher tapping temperature are now made. Less spectacular than the advance of the basic oxygen furnace, but no less important, is the development of electric arc furnaces from exclusive use for special steels and a maximum size of about 30 tons to their use in large tonnage steel melting shops at sizes of up to 140 tons, and probably in future up to 200 tons. This is likely to be the main cold metal process of the future. It is sometimes said that the growing ability of LD converters to accept more scrap, up to, say, 40%, by preheating or by the introduction of fuel burners, will slow up the growth of electric arc furnace steelmaking. However, the tendency to concentrate cold metal practice in scrap producing districts may work against the transport of scrap to LD plants. How about spray steelmaking and the fuel–oxygen–scrap (FOS) processes? Like most new developments, original unqualified enthusiasm is now being tempered by the inevitable difficulties encountered in bringing them to full scale commercial production. Remembering the time it took, however, to bring other familiar steelmaking processes, such as LD or going further back the Bessemer and Thomas processes, to fruition, it may be some years yet before we see the large scale introduction of spray steelmaking or FOS, but my successor in 1993 with a lecture perhaps entitled ‘Fifty years on’ may be pointing out how spray steelmaking has replaced basic oxygen converters and FOS taken a large part of the manufacture in the electric arc furnace. Although the ‘undiluted’ FOS process, in which electricity was not used at all, has proved economically disappointing in its first large scale trial, the principle of oxy fuel

SEQ 0039 JOB IOM8627-002-003 PAGE-0039 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On

39

may yet be used in arc-furnace practice to preheat scrap, or to assist melting, because of the high cost of electricity in this country. Indeed, unless the latter is remedied, and the price advantage our large and constant load deserves is obtained, serious consideration will have to be given to the generation of our own power, maybe even using a nuclear reactor or coal fired station linked exclusively to the works. Indeed, as nuclear technology develops, frequent studies are made of the possibility of applying nuclear energy to the industry. So far, the maximum temperatures achievable in the heat exchange gases from gas cooled reactors are not high enough to be of much interest to conventional steelmaking, and the only immediate practical application would be the installation of a nuclear electric generating station to supply a steelworks. However, the more advanced reactors using ceramic fuel elements and carbon monoxide as coolant give promise of providing temperatures of the order of 1000°C, approaching those adequate for some of the direct reduction processes. Twenty-five years from now, direct application of nuclear heat may be widespread. Vacuum melting, i.e. carrying out the complete operation of induction melting and casting inside an evacuated vessel, was practised in Hatfield’s time but has not expanded much since, largely because it is expensive and the size of melting unit which can be used has been limited so far to about 50 tons. There are several furnaces of 10–15 tons in the world but the largest unit in the UK is the new Brown Firth 5 ton furnace capable on increase to 10 tons. Many of the desirable properties of vacuum melted material can be achieved by the relatively new processes of consumable arc vacuum melting and the more economical electroslag refining technique. The latter particularly is making great headway. There are now 47 plants in production or under construction in the world (outside the USSR) and ingots, including slab ingots, of up to 5 tons in weight have been produced in this way. Indeed, by the drop-bottom technique analogous to continuous casting there would seem to be no limit to the length of slab which could be produced in this way. Another dramatic change in the 25 years since Hatfield’s death is the recent rapid development of continuous casting, which is becoming a standard commercial method of production. Today, there are 172 plants of this type, ranging from small experimental units to major plants of over 2m tons a year capacity, and more revolutionary techniques, such as the casting of plates in the Hazelett type of machines, still await us. Hatfield would have called for the sectioning of many specimens produced in this way to examine the as cast structure of billet, slab or section with its tendency to axial porosity. Despite the amounts of commercial steel made by continuous casting, the possibility of developments in techniques are still in their infancy. These include sequence casting, the casting of shapes, and sequence rolling. Sequence casting could improve the yield of steel still more by eliminating a large amount of tundish skull and starter bar ends. The problems are largely those of securing refractories capable of extending tundish and nozzle life. Sequence rolling is already anticipated to some extent by the practice of imposing a reduction at the exit of the machine to remove central porosity, and to give a greater flexibility in size without the need to change moulds. The problems of finish

SEQ 0040 JOB IOM8627-002-003 PAGE-0040 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

40

Hatfield Memorial Lectures Vol. III

rolling in line with continuous casting are largely those of matching the relatively high rolling speed of a finishing mill to the slow output of a single strand, but by installing a transfer table at the discharge of a battery of strands and inserting a soaking or wash heating furnace for temperature equalisation, it should be possible to roll in sequence with the continuous casting plant. This is one more step towards continuous steelmaking. In recent years a number of proposals for continuous steelmaking have been put forward and the work of Thring in this country and of Howard Worner in Australia will be familiar. Both these inventors base their plants on counterflow slag movement in an OH type operation. IRSID in France are working on a continuous oxygen furnace and spray steelmaking could in theory be continuous, given a continuous supply of hot metal. The economic gains through continuous steelmaking have still to be established. One of the problems receiving much attention in Hatfield’s time was that of gases and inclusions in steel. It may seem commonsense to assume that applying a vacuum to liquid steel will at least encourage the removal of gases, but although this suggestion was made on more than one occasion to the Ingots Committee it was always dismissed on the grounds of high cost. Maybe Hatfield, had he lived, would have pursued this, but in the end it was left to the Germans to show that vacuum degassing was not hopelessly expensive, and it is now common practice for a range of steels. The first trials of vacuum degassing had as their objective the removal of hydrogen; it has since found a more generally applicable use for deoxidation. The chemistry is simple, namely that reducing the partial pressure of carbon monoxide in the system enables carbon to play its full part as a deoxidiser, so producing a gaseous deoxidation product and cleaner steel. The cost of processing is only a few percent of the cost of the steel, so considerable growth of vaccum degassing in future steelmaking operations can be anticipated. At present there is vacuum degassing capacity for only about 3% of the Corporation’s steel production.

STEELWORKING Rolling mills have not shown the novelty of the newer steelmaking and casting processes. Improvements have been more in the detail of engineering design, in better drives, and more powerful motoring; in Hatfield’s time, for example, there were many steam driven mills. In the section mill field the modern universal beam mill had been anticipated some time before by the Gray mill on Teesside. Plate mills were all two high, mainly single stand mills, compared with the modern plate mill with four high stands. There were two continuous hot strip mills (at Ebbw Vale and Shotton) but sheet was made mainly on the old fashioned hand mill, work at which was both heavy and dangerous. A considerable degree of automation ranging from programme control by punch card or peg board to the more sophisticated systems on wide strip mills is now in being and will be extended in future. Hatfield worked in a company which specialised in forging, casting and engineering

SEQ 0041 JOB IOM8627-002-003 PAGE-0041 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On 41 rather than in rolling. He might, therefore, show only mild interest in these high tonnage developments, apart perhaps from seeing stainless steel sheet rolled in a wide strip mill. Very marked improvements in welding as a form of fabrication have been developed in the past 25 years. The forged boiler drum has disappeared and the large monobloc rotor and the large casting are under competition from the welded article. Thus, the tendency to ever increasing sizes of ingot at the very large end is now being halted, although for billets, bar, section, and flat rolled products in the common steels, the average size of ingot is tending to increase in size because of the economic advantages of yield and output. We shall see less of a practice which was always regarded as something of a tour-de-force in Hatfield’s time: the tapping of the contents of several steel furnaces into one mould. The largest ingots cast by English Steel Corporation (ESC) were 310 tons from four acid OH furnaces and a 220 ton ingot from five electric arc furnaces; at Firth Brown 219 tons were cast from four acid OHs. The largest casting, also made by ESC, was 260 tons from four acid OHs. Continuous casting of 300 tons from LD furnaces, however, may become commonplace. Forging, particularly in its lighter varieties has moved a long way from Hatfield’s time in the direction of improved heating practice and particularly of automatic controls permitting much faster press operations. Since many of the problems of extrusion are akin to forging. Hatfield would have been interested in the various possibilities of direct production of gear blanks, twist drills, etc. by hydrostatic extrusion, particularly the elegant use of back pressure to prevent breaking up of the metal on the downstream side. He would also have become involved with the growing developments of powder metallurgy, not only for making conventional pressings as alternatives to drop forgings or castings, but for compacting into strip and other basic forms. This is one of the developments the steel industry in this country must enter into energetically to tap a rapidly growing market.

MAKING STEELS BETTER . . . The development of new steels was energetically pursued in Hatfield’s time, as it is now. This subject is so extensive that it requires a lecture to itself, but it is hardly necessary to point out the great advances which have been made in the last 25 years. As an example, however, may be quoted the improvement in mechanical properties obtaining in common classes of steel since 1943. Shipbuilding plates for hulls had a yield stress around 13 tons/in2, and the specified impact value of 20 ft lb could be measured at 20°C; now hull plates for critically stressed parts have to have a yield stress of 17–22 tons/in2 and the 20 lb impact value held at –40°C to –50°C. At the same time the steel has to be weldable. In 1943 steam piping for electric generating plant was usually mild steel for service at 450°C and had to show a 100,000 h rupture stress of 6 tons/in2; now steels for service at 600°C and 100,000 h rupture stress of 8 tons/in2 are used.

SEQ 0042 JOB IOM8627-002-003 PAGE-0042 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

42

Hatfield Memorial Lectures Vol. III

To these may be added the steels for special application, as for instance the high temperature steels for gas turbine engines and the maraging steels where strength and weldability need to be combined with lightness; the creep resisting steels to meet the continued increase (to its present plateau) in the temperature of operation of power stations and for some of the process stages of making gas from oil; and the life of our motor car engines between top overhauls has been largely improved through valve steel development. Hatfield would have been interested in the various thermomechanical-cumtransformation processes, such as controlled rolling, maraging, ausforming, and isoforming, which can give greatly improved properties over a much smaller range of alloy steel composition than had hitherto been the case. The extent to which these developments will be commercially adopted must be largely one of economics, the main extra cost being that of the reduced output enforced by operating at lower rolling temperatures. Progress in this field certainly is a continuing extrapolation of today’s knowledge, punctuated at indeterminate intervals by discontinuous changes due to new discoveries or inventions. The next 25 years should see an improvement of some 25% on present-day mechanical properties by gradualness and even greater advance if some new feature of solid state or metalmaking practice, e.g. fibre technology, is practised. After all, practical realised strengths are only 1/100 of what they could be ideally.

. . . AND SPECIAL Corrosion is still one of the great enemies of iron and steel, and the work of the Corrosion Committee set up 25 years ago still continues in laboratories throughout the country. The resistance to corrosion in service applications was tackled in a number of familiar ways (by using non-corrodible compositions such as stainless steel, by tinning or galvanising, or by painting) all of which were and still are the subject of much research. The market for stainless steel has grown enormously, largely in the form of flat rolled products for the chemical and food industries, and for domestic and building use, and, by further developments yielding economies of scale, stainless sheet can become more competitive in a wider range of markets. In the structural and building industries, low alloy corrosion resistance steels are now available, of which unpainted Cor-ten steel in structures, e.g. bridges, may be quoted. As for coated steels, the familiar tinplate and galvanised sheet are still much in demand, in a much greater variety of coating types and thicknesses made possible by the trend, almost complete in the case of tinplate, towards electrolytic instead of hot dip coating. In the past 25 years the whole process of manufacture of tinplate has made a dramatic change from the crude packmill to the modern continuous electrolytic plant of today. There is now rapid development of entirely new coatings, not significant in Hatfield’s time, for developing industries: the soft drinks industry, canned foods, etc., for which the chromate (Hi-Top) or chromium (Cansuper) substitutes for tinplate are being made and the various plastic laminate coatings or ‘colour

SEQ 0043 JOB IOM8627-002-003 PAGE-0043 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Twenty-Five Years On

43

coated’ steel made by a roller coating process for domestic application. A still wider range of coatings is being developed on a laboratory or pilot plant scale, based on vacuum deposition or the powder coating principle, and we now have the possibility of applying to steel almost any substance we wish.

CONCLUSION Were Hatfield able to return to the industry today there is little he would regret and much of which he would approve. He might regret the obsolescence of acid OH steel (thought by many in Hatfield’s day to be the best steel of all), the possible decreasing importance of the very large forging, and the standard of some of today’s literature. He would certainly approve the growth of the industry in size and range, the recognition of research as vital to its wellbeing, the powerful tools, theoretical and experimental, now available for metallurgical research, and some of the highly modern and efficient processes and plants for bulk production of all steels and particularly alloy and special steels. What he would think about the structure of the industry following nationalisation would be surmise, but that he would delight in the technical challenge to steel in the laboratories and in the shops and mills is undoubted.

SEQ 0044 JOB IOM8627-002-003 PAGE-0044 NINETEENTH LEC REVISED 19DEC04 AT 20:17 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0045 JOB IOM8627-003-004 PAGE-0045 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE TWENTY-THIRD HATFIELD MEMORIAL LECTURE

Electroslag Remelting – a Modern Tool in Metallurgy E. Pl¨ockinger At the time the lecture was given Professor Erwin Pl¨ockinger was Director of Research, Gebr. B¨ohler and Co., Vienna, Austria. The lecture was given at Ranmoor House, University of Sheffield, on 9 January 1973.

The electroslag remelting process is one of the most important of the many new processes which have been developed for the production of special steels and alloys. The main reasons for this predominance are: the remelting under a slag layer of high reactivity removes undesirable elements as well as non-metallic inclusions without altering the desired chemical composition. The chemical composition can be kept within close tolerances in the remelted material. The controlled solidification of relatively small volumes of liquid steel results in a very homogeneous ingot without internal defects. These characteristics of the remelted ingot are the essential basis for the production of wrought steels having superior cleanliness and mechanical properties which cannot be achieved by other processes. Another advantage is that almost any ingot section and size can be produced by ESR. This even includes hollow ingots for producing seamless tubings and rings and preforms of different shapes in cross-section. The ESR process can be easily automatically controlled thus giving close tolerances in the chemical composition and the structure of the whole remelted ingot. From a discussion of the metallurgical reactions involved in the ESR process it can be concluded, that there are still many possibilities for major improvements, e.g. the method of electrolytic refining and the melting under controlled atmosphere. Practical experience proves that this process is especially suited for high grade nickel and cobalt base super alloys and special steels, such as high strength steels, tool steels, and heat and corrosion resistant steels, which have to meet extreme requirements of the consumer industries. ESR ingots could successfully be used for the production of isotropic materials having the same properties in all directions. A very important field of application is the production of heavy forgings for the power industry. Despite the additional costs involved, ESR in many cases is economically competitive and sometimes even superior to conventional processes. This is mainly due to its reliability and the very good overall yield. Because of its metallurgical and economic advantages ESR has already found a wide use as a modern and necessary tool for the production of high grade steels and alloys.

45

SEQ 0046 JOB IOM8627-003-004 PAGE-0046 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

46

Hatfield Memorial Lectures Vol. III

The recent progress in the field of steelmaking processes has been characterised by many important developments which have found a worldwide use in a relatively short period; for instance, the techniques of oxygen blowing and vacuum metallurgy. Besides a more economic production of steels and alloys, the new processes have made it possible to meet the continuously increasing demands of the consumer industries on the quality of these materials. At present this especially applies to the electroslag remelting process for the production of high grade materials.1,2,3 It is no exaggeration to say that electroslag remelting has already established new measures for special steels and therefore there seems a justification to review the reasons for its fast and general acceptance, and to describe its widespread field of application already existing. The technical and metallurgical possibilities of the electroslag remelting process are today by no means fully utilised. However, their representation should be helpful in predicting future developments and in describing further fields of application. Such a contribution to the philosophy of electroslag re-melting necessitates some simplifications and the renunciation of metallurgical and technological details in order to emphasise clearly the essential items. For this reason I shall try to limit my paper to the importance of the ESR process within the entire field of modern processes for the production of extremely high quality steels and alloys. Although the ESR process can be used for the refining of many ferrous and nonferrous metals and alloys, I should like to speak only about the manufacture of steel ingots, as this is the main application of the new process today where the technical progress achieved can be easily recognised. A careful examination of all steps in steel production shows that the final technological properties of the wrought steel do not only depend on the proper melting processes, the proper chemical composition and an adequate hot working and thermal treatment, but especially on the properties of the ingot. These properties of the ingot influence remarkably the hot workability as well as the quality of the final product. Therefore, in steel production it is not sufficient only to adjust to the required chemical composition and aim for the utmost cleanness in the molten steel. The decisive step is the solidification of the molten metal into an ingot free from internal defects.4 Such an ideal ingot should have the following properties: (i) uniform chemical composition, i.e. no segregation of any kind (ii) absence of undesirable and harmful elements and non-metallic inclusions (iii) uniform structure throughout the whole ingot without microporosity and blowholes. With conventionally cast ingots, as is generally known, these requirements can be achieved only to a limited extent. The problems arising from this fact will increase with increasing size and weight of the ingots. The major reason for this is not due to an improper technology in ingot production, but mainly to the physical laws controlling the crystallisation process. In conventional casting it is not possible to influence sufficiently

SEQ 0047 JOB IOM8627-003-004 PAGE-0047 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Electroslag Remelting – a Modern Tool in Metallurgy

47

the solidification behaviour of industrial scale ingots under the given pouring conditions. To overcome all the disadvantages of the conventional solidification process in the ingot mould a fundamental change in the crystallisation of the liquid steel has to be achieved. That is possible only with the continuous solidification of small amounts of molten metal as it takes place in the remelting processes. Figure 1 shows schematically the difference in the structure of both a conventionally cast and a remelted ingot. The solidification of a large volume of liquid steel in the ingot mould leads to ingot segregation. The extent of such segregation depends upon the crystallisation behaviour of the solidifying multiphase system. Local accumulations of non-metallic inclusions and the formation of microporosity and shrinking holes cannot be avoided. The structure of the ingot shows dendritic and globular solidification patterns with varying orientation.

Fig. 1 Comparison of the structure of a conventionally cast and a re-melted ingot.4

In the remelted ingot a relatively small amount of liquid metal solidifies continuously according to the melting rate of the electrode. The solidification progresses into a shallow liquid-metal pool. Thus macrosegregation, accumulation of non-metallic inclusions and microporosity will not occur, and the remaining small amounts of non-metallic inclusions are evenly distributed. The ingot structure shows columnar crystals which are slightly inclined toward the longitudinal axis of the ingot. Therefore the remelted ingot nearly meets all the previously mentioned conditions for an ideal ingot. However, a certain microsegregation cannot be prevented even by this process. The first technical realisation of this principle took place in the vacuum arc furnace with consumable electrodes. The vacuum remelting process allows the degassing of the metal and the elimination of non-metallic inclusions, thus giving a remelted steel of high purity and the desired structure. However, no other metallurgical reactions are possible. Therefore the electrodes must have the same chemical composition as the remelted ingot, but possible losses of elements of high volatility must be taken into consideration. Some of the removed impurities and evaporated elements are deposited in the shell zone of the remelted ingot which generally has to be removed in the following manufacturing processes. If, on the other hand, the remelting of a consumable electrode is carried out in

SEQ 0048 JOB IOM8627-003-004 PAGE-0048 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

48

Hatfield Memorial Lectures Vol. III

a reactive slag as in the ESR process, further metallurgical and technological advantages can be achieved. These advantages have proved to be decisive for the rapid introduction of this new process. The principle of the electroslag remelting process is shown in Fig. 2. It is a continuous process in which melting, metallurgical reactions and solidification of the steel occur simultaneously. As the aim of the process is the production of ingots with superior quality, the rate of crystallisation determines the production rate. This is because the heat conduction of the solidified metal is limited and does not allow an increase of the rate of crystallisation at random.

Fig. 2 The principle of the ESR process.

Electroslag remelting can be carried out in different ways as shown in Table 1. The metallic charge usually is a cast electrode of suitable cross-section. Both forged and rolled electrodes can be used. The refining of liquid metal in the same way has also been suggested. This, however, would require additional heating of the slag with, for instance, auxiliary electrodes. Metallic powders together with a strip electrode can also be used for the production of remelted ingots. The melting heat is usually produced by resistance heating the slag bath. On the other hand, melting can also be carried out by means of either a permanent electrode or a plasma arc. The use of collar type moulds instead of fixed long moulds gives the possibility of producing ingots in any length required. By electroslag remelting not only round ingots, but also ingots of practically any desired shape such as square, rectangular or hollow, can be produced. It is possible to manufacture preforms even with varying cross-sections. This method is applied for example in the production of rolls.

SEQ 0049 JOB IOM8627-003-004 PAGE-0049 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Electroslag Remelting – a Modern Tool in Metallurgy Table 1

Possible features in electroslag remelting.

Metallic charge

Power supply

Mould design

Ingot shape

1. Cast electrodes

1. Resistance heating a) alternating current b) direct current c) ac superimposed dc

1. Fixed long mould

1. Round

2. Collar type mould a) movable mould b) moveable base plate

2. Square

2. Rolled or forged electrodes 3. Strip electrodes and metal powder 4. Liquid metal

49

3. Rectangular

2. Non-consumable electrode

4. Hollow

3. Plasma arc

5. Preform

The metallurgical reactions involved in the ESR process enable the metallurgist to adjust the refining process to a widespread range of requirements.2 I should like to show some details in the following examples in which only the standard process will be discussed. In this process a solid electrode is melted in a reactive slag and the liquid metal solidifies in a water cooled mould. Any metallurgical reaction in the ESR process has to be considered as a three phase reaction in which the molten metal, the slag and the gas phase take part. The reactions are characterised by high reaction rates due to a high surface to volume ratio between the liquid metal film at the electrode tip and the highly superheated slag, which furthermore is intensively moved by both thermal and electrical forces. The composition of the slag can be varied to quite an extent and thus be adapted for different metallurgical requirements. The slag consists mainly of fluorspar and varying contents of alumina and lime. Normally it should have a desulphurising effect on the metal and should be capable of absorbing the non-metallic inclusions. Other slags, however, may be adjusted to retain sulphur in the metal if desired, for example, in resulphurised tool steels. In any case, the composition of the slag should be such that the following conditions are met: (i) the electric specific resistivity should be between 0.20 and 0.50 ohm cm (ii) oxidation of alloying elements has to be avoided (iii) pick-up of undesirable elements such as P, Si, Al from the slag by the liquid metal should not occur (iv) the melting point of the slag should be lower than that of the metal to be remelted (v) the slag should be of low viscosity in order to produce just a thin slag layer on the ingot surface (vi) an adequate interface tension between slag and liquid metal must be achieved. The slag–metal reactions in the ESR process will follow the well known thermodynamic principles. Of course the high reaction temperatures, in the order of 1800°C, have to be taken into account. Figure 3, showing the Si–O equilibrium in electroslag remelting may be taken as a typical example.1 From the strong dependence of the silicon deoxidation constant on the slag basicity, or in other words its silica activity, it can be concluded that low oxygen contents in the remelted steel can be achieved only when slags with low silica

SEQ 0050 JOB IOM8627-003-004 PAGE-0050 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

50

Hatfield Memorial Lectures Vol. III

Fig. 3 The de-oxidation constant of silicon K'Si in electroslag remelting.1,2

contents are used. Therefore the oxygen content of the ESR ingot depends only on the physico-chemical interactions between slag and metal during remelting, thus being independent of the oxygen content in the electrode. The operating slag not only interacts with the liquid metal, but also with the gas phase above the slag pool. With open melting the slag may absorb hydrogen and oxygen from the air. Oxygen pick-up leads to the formation of SO2 which escapes to the atmosphere, thus resulting in a constant desulphurisation during the entire remelting operation. This reaction, however, causes oxygen to pass through the slag into the melt which can lead to the oxidation of certain alloys, for example Al, Ti and Mg in high alloy steels and Ni base alloys and to a higher oxygen content in the steel. Figure 4 shows the influence of the partial pressure of oxygen in the gas phase above the slag pool on the oxygen content of the remelted steel for various slags.2 An undesirable oxygen pick-up can thus be prevented by using a gas atmosphere with sufficiently low oxygen contents. Another method is the application of a continuous deoxidation of the slag. Both methods are used for the production of metals and alloys with an extremely low oxygen content. Hydrogen pick-up from the atmosphere can be even more harmful, as a transfer of hydrogen to the liquid metal may occur. Such pick-up will be more pronounced the higher the humidity of the atmosphere and the larger the slag surface exposed to the

SEQ 0051 JOB IOM8627-003-004 PAGE-0051 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Electroslag Remelting – a Modern Tool in Metallurgy

Fig. 4

51

Influence of the partial pressure of oxygen in the gas atmosphere on the oxygen content of AISI304 remelted under various slags.2

atmosphere. By remelting under a dry protective gas, hydrogen pick-up can be eliminated and even a reduction of the hydrogen content of the metal can be attained. As the slags used for remelting at high temperatures represent ionic solutions, the mode of current will influence the metallurgical reactions to an extent which should not be neglected. Varying results of the desulphurisation and oxygen content of the ESR ingot in dependence on mode of electric current are shown schematically in Fig. 5.5,6 When direct current is used, the metallurgical results obtained are strongly dependent on the current polarity and current density at the surface of the liquid metal pool. With increasing ingot diameter the current density will decrease and above a certain diameter, only the polarity of the consumable electrode will be responsible for the extent of the desulphurisation and deoxidation reactions. If alternating current is applied, the direction of these reactions will vary according to the frequency of the current. This in turn will result in an oscillation around the thermodynamic equilibrium which forces a rapid approach to the theoretical state of equilibrium. A combination of the metallurgical and economic advantages can be obtained if a superimposed dc is applied in ac remelting. As an example the effect of superimposed dc on the reduction of hydrogen during remelting is shown in Fig. 6. This effect is observed independent of the composition of the slag.7 Disadvantages of ac as compared to dc operation are increased resistance and inductive losses. These may be overcome by utilising a low frequency ac which is of special

SEQ 0052 JOB IOM8627-003-004 PAGE-0052 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

52

Hatfield Memorial Lectures Vol. III

Fig. 5

Oxygen content and desulphurisation in ESR ingots independent of ingot diameter and current mode (schematic).5,6

Fig. 6 The change of hydrogen content in steel lime content of the slag and the mode of current.7

SEQ 0053 JOB IOM8627-003-004 PAGE-0053 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Electroslag Remelting – a Modern Tool in Metallurgy

53

importance in the operation of large-scale ESR units. If solid-state equipment is used for the conversion to low frequency ac an alternating current with varying phase lengths can be obtained as shown in Fig. 7. This possibility could be used to influence further the direction of the desired metallurgical reactions.5

Fig. 7

Possible variations of negative and/or positive phase length with low frequency ac.5

In order to obtain ESR ingots having optimum structures it is important to adjust the melting rate and the power input to the slag pool so that a constant shape of the solidification fronts and unidirectional solidification are obtained. This requires a shallow liquid metal pool, which can be achieved by adequate selection of the operating parameters such as current and voltage under otherwise constant conditions. In Fig. 8 the influence of operating conditions on the primary structure of stainless steel ingots is shown.2 With proper conditions, as shown on the left, a uniform ingot structure with columnar crystallites slightly inclined to the longitudinal axis of the ingot is obtained. On the contrary, the use of operating conditions resulting in a deep metal pool leads to an ingot structure very similar to a conventional cast ingot with internal defects, as shown on the right. Before dealing with the material properties of remelted steels and alloys it should be mentioned that the ESR process is better suited for automation than any other melting process. More than that, it demands automation if highest uniformity in the production is to be maintained. Today the basic knowledge required is available, such as detailed experience in metallurgy, as well as the influence of operating conditions on the ingot structure. During remelting constant operating conditions can be maintained only by means of automatic control devices. Today such equipment is installed in all modern ESR plants, enabling not only automatic operation during the actual remelting but also automatic control of starting, hot topping and electrode change operation.

SEQ 0054 JOB IOM8627-003-004 PAGE-0054 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

54

Hatfield Memorial Lectures Vol. III

Fig. 8 Influence of operating conditions on the macrostructure of ESR ingots:2 (a) 48 V, 2000 A; (b) 42 V, 5000 A.

In ESR the consumable electrode will not only be converted into perfectly desulphurised and deoxidised metal but furthermore into an ingot with a more or less perfect structure as discussed before. In any case the ingot structure is of such paramount importance that even higher contents of equally distributed non-metallic inclusions are less harmful to the properties of the alloy than in conventional cast ingots. Having discussed the ESR process in general, an attempt will be made to demonstrate the superior quality of the ESR material with regard to workability and properties of the final product.8–10 The ingot structure, without internal defects and ingot segregation, leads to improved hot workability of the remelted material, especially for high alloy steels difficult to hot work. In Fig. 9 a hot torsion test diagram of the AISI329 Cr–Ni–Mo stainless steel is shown as an example.11 The distinctly improved hot plasticity of this material means that it can be successfully used for the manufacture of complicated drop forgings. Such forgings manufactured from the as cast ESR ingot by a single drop forging operation are shown in Fig. 10. Even new alloys such as a 3%C–12%Cr steel not suited for hot working from the conventional ingot can now be rolled or forged without major problems from the ESR ingot. The superior quality of even large diameter ESR ingots especially with regard to ingot segregation, microporosity, and accumulation of non-metallic inclusions, represents a definite advantage in the production of heavy forgings. For these reasons it was to be expected that high quality forgings could be obtained from comparatively small ESR ingots as less hot working should be required. In order to determine the minimum reduction required by forging, a step down forging, as shown in Fig. 11, has been made from a 22 t 1000 mm diameter Cr–Ni–Mo

SEQ 0055 JOB IOM8627-003-004 PAGE-0055 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Electroslag Remelting – a Modern Tool in Metallurgy

Fig. 9 Hot torsion test diagram of AISI329 from conventional and ESR ingots.11

Fig. 10 Upset forged parts directly produced from ESR ingot steel AISI329.2

55

SEQ 0056 JOB IOM8627-003-004 PAGE-0056 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

56

Hatfield Memorial Lectures Vol. III

Fig. 11 Influence of hot working ratio on the reduction of area, step down forging from a 1000 mm diameter Ni–Cr–Mo steel ESR ingot12 (0.25%C, 3.5%Ni, 1.25%Cr, 0.5%Mo).

alloy steel ESR ingot.12 This forging, which was forged without upsetting, was heat treated and samples were taken from all steps. The tensile strength varied betweem 830 and 900 N/mm2 and showed no dependence on the direction or the location of the samples within the forging. Therefore it appeared more appropriate to use the toughness values in order to differentiate the mechanical properties. Figure 11 shows the influence of the hot working ratio on the reduction of area. It can be seen that excellent toughness properties were obtained both for the longitudinal and for the transverse direction in the periphery, mid-radius and centre position. This is true even for a hot working ratio as low as 2.5. Similar results were obtained for the notch toughness. The superiority of the electroslag remelted material over conventionally produced forging ingots is also proved by the results on generator shafts and heavy turbine discs. Two generator shafts each weighing 8.9 t were produced from a Ni–Mo steel. The first shaft was made by stretch and upset forging from a vacuum degassed 22 t forging ingot having 1200 mm diameter, and the second shaft was stretch forged only from an ESR ingot having a weight of 14 t and 1000 mm diameter. It can be seen from Fig. 12 that after heat treating to the same tensile strength (680 N/mm2) the shaft made from the ESR ingot had improved toughness properties. This is especially true for the transition temperature which was lowered with electroslag remelting by 20°C. A statistical evaluation of the mechanical properties of large turbine discs having a diameter of 1700 mm and a height of 600 mm also showed the superiority of the electroslag remelted ingot. These forgings were made from Cr–Ni–Mo steels. Figure 13 shows the toughness properties of two 9.2 t discs produced from a conventionally cast ingot by stretch and upset forging or from an ESR ingot by upset forging only, respectively. Even in larger turbine discs (2000 mm diameter, 16 t weight) which were upset forged from an 18 t ESR ingot having 1000 mm diameter no fibrous stucture or segregation could be observed. The mechanical properties were independent of the direction.13

SEQ 0057 JOB IOM8627-003-004 PAGE-0057 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Electroslag Remelting – a Modern Tool in Metallurgy

Fig. 12

Comparison of the mechanical properties of two 8.9 t generator shafts of a low alloy Ni–Mo steel made from arc furnace vacuum cast and ESR ingots.12

Fig. 13

Comparison of the toughness properties of two 9.3 t turbine discs of a Cr–Ni–Mo steel made from arc furnace vacuum cast and ESR ingots.12,14

57

SEQ 0058 JOB IOM8627-003-004 PAGE-0058 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

58

Hatfield Memorial Lectures Vol. III

A further distinct advantage of the ESR ingot is represented by its low microsegregation. In Fig. 14 the microsegregation ratios of chromium and molybdenum are compared for both conventionally cast and ESR ingots of H13 hot work die steel. This low microsegregation combined with the freedom from ingot segregation is an important presupposition for the production of practically isotropic wrought special steels. This already can be realised on a technical scale by using special hot working techniques and heat treatments which have been developed in order to minimise further the microsegregation. These steels have practically the same properties in all directions. From Fig. 15 the marked improvement of reduction of area in the transverse direction with decreasing microsegregation ratio of molybdenum can be seen.

Fig. 14 Primary structure and microsegregation of H13 in conventionally cast and ESR ingots:14 (a) conventional ingot (centre), 300 mm diameter, as cast; (b) ESR ingot centre, 300 mm diameter, as cast.

Fig. 15 Influence of microsegregation on the reduction of area in the transverse direction in forged bars of H13.15

SEQ 0059 JOB IOM8627-003-004 PAGE-0059 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Electroslag Remelting – a Modern Tool in Metallurgy

59

The extreme reduction of microsegregation is of definite importance in the production of high strength steels, tool steels and above all hot work die steels. A statistical evaluation of test results of a two year production period of H13 bar material in the diameter range 150–470 mm is presented in Fig. 16. Optimum results, particularly in the transverse direction, are obtained from ESR material specially treated to give minimum microsegregation. Conversely such a special treatment has only a rather limited effect on conventional cast ingots.

Fig. 16

Statistical distribution of toughness properties in the transverse direction in forged bars 150–470 mm diameter steel H13.15

Excellent and isotropic toughness properties result in an improved behaviour of materials under multiaxial stresses and in a reduction in notch sensitivity, as is shown in Fig. 17. This diagram contains the curves for the notch bar tensile strength v. tensile strength of open arc furnace, VAR and ESR material. It can be seen that remelting in the vacuum arc furnace or by the electroslag process not only improves the notch bar tensile strength, but also shifts the beginning of the decrease of the notch strength to higher values of the

Fig. 17 Notch tensile test diagram of steel H13 melted under various conditions.8,15

SEQ 0060 JOB IOM8627-003-004 PAGE-0060 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

60

Hatfield Memorial Lectures Vol. III

ultimate tensile strength. A further improvement is obtained by a special treatment for the removal of microsegregations. The fatigue properties are also markedly improved by electroslag remelting. Figure 18 shows that the fatigue strength of electroslag remelted steel H13 is superior to that of conventionally produced steel, and that the fatigue strength can also be further increased by the special treatment which has been mentioned before. This is true for both specimens with and without a notch. Indeed, the experience has proved that very remarkable improvements can be achieved in practice. For certain applications the life of H13 hot work tools has been increased by about 100 or 200% and more than the life of dies made from conventional H13.16

Fig. 18 Fatigue strength of steel H13, heat treated to UTS 1900 N.mm2, rotating bar test.8,15

CONCLUSIONS The improvements and achievements obtained by the ESR process to date can be briefly summarised as follows: 1. Remelting of case hardening steels is mainly carried out to improve the cleanness to an extent which cannot be attained in conventional melting and casting processes. The improved toughness properties in the transverse direction are of importance only when high strength levels are required. 2. For heat treatable steels the toughness properties in the transverse direction are of importance besides the improved cleanness. A typical example is represented by alloy steels used for aircraft landing gear which will safely meet the stringent specifications when electroslag remelted. 3. Improved toughness properties can be attained in high duty forgings such as turbine rotors and discs as well as generator shafts. The better overall yield attainable with this kind of production may to a certain extent compensate for the increased expenses due to the additional remelting operation.

SEQ 0061 JOB IOM8627-003-004 PAGE-0061 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Electroslag Remelting – a Modern Tool in Metallurgy

61

4. The excellent cleanness and toughness properties together with a special hot working and thermal treatment leads to the production of practically isotropic hot work die steels and ultra high strength steels, thus justifying the application of the ESR process for these alloys. 5. By remelting ledeburitic tool steels, an extraordinary uniformity in carbide distribution is achieved throughout the cross-section, which in turn results in less distortion in heat treatment and in many cases, especially for cold working applications, in an improved service life. 6. Precipitation hardening Ni and Co base super alloys today are remelted by ESR, attaining at least the same standards as by the VAR process. For these alloys cleanness and structural homogeneity are of the same importance as the safe production to close analytical tolerances. The rapid adoption of the ESR process by practically all major steel producing countries is easily understood if one considers the metallurgical progress already attained and still attainable by this new process. As shown in Table 2, 37 production scale ESR units capable of producing ingots with diameters above 400 mm were in operation at the end of 1972 and 6 more units are under construction. The production of ESR material in the western countries will have reached approximately 80,000 metric tons in 1972. In considering the units under construction and additional projects to be carried out in the next years, the annual production in 1977 can be estimated to be 200,000 t. This would represent an average annual growth rate of 20%. Table 2 ESR units with ingot diameter ≥ 400 mm in the western world and their estimated production.

Country

ESR units operation, 1972

Estimated production, 1972

Units under construction, 1972

Estimated production, 1977

Western Europe Austria Belgium France German Fed. Rep. Great Britain Italy Sweden Spain USA Japan Others

2 .. 2 7 8 1 6 .. 10 1 ..

5000 .. 4000 16 500 14 000 2000 14 000 .. 21 500 3000 ..

.. 2 .. .. 1 1 1 .. 1 .. ..

      200 000     

Total

37

80 000

6

200 000

The units to be put into operation in the next few years will mainly cover the ingot diameter range 500–1000 mm. A number of large scale ESR units for the production of heavy forging ingots up to 3000 mm diameter will be out into operation; one 2500 mm unit has already been started up recently. Furthermore there is a demand for heavy slab

SEQ 0062 JOB IOM8627-003-004 PAGE-0062 TWENTY-THIRD LEC REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

62

Hatfield Memorial Lectures Vol. III

units for the production of ESR slab ingots in excess of 50 tons, the construction of which, according to the present state of development, does not seem to be impossible. For ultra high quality products and special requirements remelting under controlled atmospheres will certainly gain importance. Generally speaking one can say that the ESR process will have to be used whenever the requirements of the properties of steels and alloys cannot be fulfilled by conventional melting and casting methods. It will, however, also be applied when improved yield and reduced rejection rates in certain cases will represent a cost saving factor, thus justifying the additional processing. The worldwide increasing application of the ESR process together with the experience gained so far may justify the statement that this new process represents a valuable tool for the metallurgical industry, enabling it to raise its technical standard to a higher level.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

¨ E. PLOCKINGER and W. HOLZGRUBER: Rev. M´et., 1968, 65, 463. ¨ W. HOLZGRUBER and E. PLOCKINGER : Stahl Eisen, 1968, 88, 638. M. WAHLSTER and H. SPITZER: Stahl Eisen, 1972, 92, 961. ¨ E. PLOCKINGER : Berg. H¨utten Monatsh., 1969, 114, 344. P. MACHNER and W. HOLZGRUBER: Paper presented at 7th internat. congress on electroheat, Warsaw, 18–22 Sept. 1972. W. HOLZGRUBER et al.: Paper presented at vacuum metallurgy conference, Pittsburgh, 16– 19 June 1969. W. HOLZGRUBER et al.: Paper presented at internat. symposium on special electrometallurgy, Kiev, 6–9 June 1972. ¨ E. PLOCKINGER : Stahl Eisen, 1972, 92, 972. A. RANDAK et al.: Stahl Eisen 1972, 92, 981. H. SPITZER: Stahl Eisen, 1972, 92, 994. W. HOLZGRUBER et al.: Radex Rund., 1970, 163. M. KRONEIS et al.: Paper presented at 5th internat. forging conference, Terni, 6–9 May 1970. M. KRONEIS et al.: Paper presented at 6th internat. forgemasters meeting, Cherry Hill, N.J., 1–6 Oct. 1972. A. SCHNEIDHOFER et al.: Berg. H¨utten. Monatsch., 1970, 115, 366. M. KRONEIS et al.: Berg. H¨utten. Monatsch., 1968, 113, 416. H. STRAUBE et al.: Schmiede. Mitt., Industrie-Anzeiger, 1971, 93, 609.

SEQ 0063 JOB IOM8627-004-005 PAGE-0063 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE TWENTY-FOURTH HATFIELD MEMORIAL LECTURE

Eurosteelresearch R. S. Barnes At the time the lecture was presented Dr Robert S. Barnes was Director of Research and Development, British Steel Corporation. The lecture was given at the Royal Commonwealth Society, London, on 14 November 1973.

The further evolution of Europe as a steel producing region will call for larger production units, yet with sufficient flexibility to adjust to fluctuations, both in raw material supplies and in the market for steel products. The access to bulk supplies of iron ore, reductants and fuels, will dominate the location of the large steelmaking plants which will be augmented by smaller, more flexible units concentrating on the finishing processes and located nearer to the steel user. In the future new sources of indigenous reductants and fuels will increasingly affect the steelmaking pattern and in the long term nuclear energy will reduce the consumption of fossil fuels. The research and development needed for these changes to come about is considerable, but if the opportunity is taken jointly to solve some of the major problems there could be a marked effect upon the long term viability of Europe’s steel industry. This will call for a clear identification of the most effective way to deploy the many research and development skills distributed throughout Europe.

It may be thought, particularly by those who belonged to the European Coal and Steel Community from the outset, that it is hardly appropriate, or even proper, for a Briton to hold forth on the subject of cooperative research and development for the European steel industry. That Hatfield was a Briton, however, gives me good reason on this occasion of the 24th lecture in his memory, and the first to be given since the European Community was enlarged. Dr Hatfield was one of the main driving forces in establishing cooperative research in the UK. He helped establish the research committees which became the focal point for the technical scene in the British steel industry between the wars and provided the basis for the formation of the British Iron and Steel Research Association (BISRA). He was a pioneer who early realised that only by common effort could the ever growing technical problems of the steel industry be tackled. The conditions that Hatfield recognised have become intensified in the thirty years since his death and many of the major technological problems of the steel industry can no longer be adequately dealt with by the industry of a single nation. It is therefore fitting that this lecture should be concerned with the widening of the research and development activity to include the British steel industry into the

63

SEQ 0064 JOB IOM8627-004-005 PAGE-0064 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

64 Hatfield Memorial Lectures Vol. III framework of the wider European scene. This is a challenge – which Hatfield would certainly have welcomed. Also, this lecture comes at a time when The Iron and Steel Institute and The Institute of Metals are amalgamating to form The Metals Society – a body which will bring together, on a worldwide basis, members of the ferrous and non-ferrous metals industries to draw mutual benefit from discussing openly the results of each other’s technical successes and failures. I am sure Dr Hatfield would have applauded this initiative. I am greatly honoured at having been invited to deliver this lecture, and particularly on this subject. Having been outside the Community until recently helps me to look back over its achievements and, not having been associated with them, to look dispassionately and without prejudice towards the future role for research and development for the European steel industry. The only advantage I can have over so many others much more qualified is the experience gained, over the last four years in particular, in establishing a new identity and role for the previously independent research and development organisations within the British steel industry. This task has involved bringing together, from a research and development point of view, fourteen independent companies and BISRA, which was the largest research association and itself pursued cooperative research, but on a national scale. As each already had its own goals, traditions and structures, the task of bringing them all together has taken time. It has required patience and a measure of that pragmatism so necessary when dealing with complex situations, and often associated with the British. While these experiences can be only of marginal value in the examination of the research needs on a European scale, I hope they can help to point out some of the directions and stimulate the debate which is now beginning on the future for steel and thus for steel research in Europe. I shall give my own view, first on the place for steel in Europe, secondly on the areas appropriate for community research, and thirdly on some of the organisational requirements for such Eurosteelresearch.

THE PLACE FOR STEEL IN EUROPE It is not my purpose to dwell on the history of the steel industry, nor on its technical evolution. Although the part played by the early pioneers, by the men of action, by the early craftsmen, by the scientists and by the technologists each would make a good subject for a talk such as this, I shall merely list (Table 1) a few of the technological achievements of Europe’s steel industry. The cooperative research stimulated by the inception of the European Coal and Steel Community (ECSC) in 1951 (which was itself the forerunner of Euratom and the European Economic Community in 1957) started a trend and has evolved some concepts which will continue. Cooperative research projects have been stimulated by funds from the Commission, these funds being provided from a levy upon the coal and steel companies of the ECSC; normally the research is conducted with 60% financial support from the Commission. The way in which these expenditures have grown in the various

SEQ 0065 JOB IOM8627-004-005 PAGE-0065 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

65

Table 1 Historical landmarks in steelmaking in Europe. (The modern steel industry was born in Europe as a result of Bessemer’s researches, and European inventions have continued to play a significant role, even though the major application of them has often occurred in other parts of the world). Year 1740 1856 1868 1878 1879 1877

1951 1971

Event Huntsman, Sheffield, England. Crucible steelmaking. About 20 lb per crucible. Bessemer in London, England. Pneumatic steelmaking 10 t in 15 min by the 1870s. Siemens Bros. in England, and Martin in France. Open-hearth steel-making: 20–30 t in 10 h. Thomas and Gilchrist. Paper on Basic Bessemer Steelmaking: main application in the 1880s in Europe. Siemens experiments with electric-arc steelmaking in England. 1900 Exploitation by Paul Heroult ´ in France. 1905 First Heroult ´ furnace in USA. Induction furnace experiments in Italy by de Ferranti. 1899 First commercial application, in Sweden. 1907 Experimental 2 t furnace, in Sheffield, England. 1914 First commercial application in USA. First LD experimental converter at Linz, Austria. Bottom-blown processes with oxygen and propane. Maximillianshutte, ¨ Germany

areas of research is shown in Table 2.1 There is also a contribution for pollution research, provided through the Social Affairs Directorate. (For the five years ending in 1972 the average amount was 0.8m units of account, while the average asked for, for the following five years, is 2m units of account.) This area of ‘Social Fund research’ is growing more rapidly than the Research Fund, which itself has increased to 8m units of account for 1973. The total ECSC support to research accounted for approximately 5% of the total research expenditure of the European steel industry in 1972. Loans have also been provided to stimulate technical development and there are indications that these mechanisms are becoming more common. Table 2 ECSC grants to steel research. Research areas supported by the Research Fund of the Commission, which increased to 8m units of account for 1973 (a unit of account is about 1.2 US dollars). Category

1955–1961

1962–1966

1967–1971

Iron ore Blast furnaces experimental other Sintering Direct reduction Steelworks Rolling mills Analysis and methods of measurement Properties in use Physical metallurgy Steel utilisation Various

2.8

1.9

0.7

2.1 1.6 0.6 1.2 0.6 0.1 .. .. .. .. 0.8

4.7 1.0 .. 2.7 2.3 2.6 1.9 1.7 0.1 0.6 1.1

1.3 2.5 0.7 0.1 3.9 3.2 2.8 4.1 1.5 1.2 1.3

Totals

9.8

20.6

23.3

Grand totals 5.4

}

13.2 1.3 4.0 6.8 5.9 4.7 5.8 1.6 1.8 3.2 53.7

SEQ 0066 JOB IOM8627-004-005 PAGE-0066 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

66

Hatfield Memorial Lectures Vol. III

While many of the projects have consisted of a number of small items often within a broad area of research, some have been larger in concept and capable of influencing more directly the future of Europe’s steel industry. Notable major projects were, for instance, the gas injection on the experimental blast furnace, the development of the Purofer direct reduction process, and the work on the IRSID continuous steelmaking process. The various research projects have succeeded in improving the contacts and the communications between the technical experts from the member countries. The research has been conducted in the laboratories of the various national research institutes, universities and the steel companies themselves. While these arrangements have not greatly enhanced the mobility of research workers within the Community, they have avoided the setting up of special research laboratories. Such laboratories have often been set up in other contexts, and experience has shown that this leads, in the long term, to problems of finding new roles when their immediate tasks have been achieved. Besides these considerable efforts on research itself, there is also the process of harmonising patent law and removing other technical barriers to trade within the Community. These could be most important to the future technical health of the industries of the Community and they can be greatly helped by a background of technical cooperation. The Market for Steel in Europe The growth in output of what is now the steel industry is shown in Fig. 1 from the year 1880 when iron was its main product.2 While the growth has not been steady, since the Second World War it has been impressive and shows no sign of abating.

Fig. 1 Iron and steel production in the UK and the Six 1880–1970. The output of what is now the steel industry is shown from 1880 when iron was its main product. The growth has not been steady, but since the Second World War has been impressive and shows no sign of abating.

SEQ 0067 JOB IOM8627-004-005 PAGE-0067 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

67

As shown in Fig. 2,3 in the Six the growth in steel consumption averaged 5.5% per year between 1955 and 1970, compared with an average growth in the UK of 2%. In their Projection 85 the IISI suggest that demand in 1985 will be 36m t for the UK and 153m t for the Six.4 They do not explicitly forecast for the Nine, but on this basis consumption in the enlarged Community would be estimated at around 200m t/year by 1985.

Fig. 2 Steel consumption in ECSC and in UK (1950–1971). For the last 20 years steel consumption in the Six has been growing at a rate of 5.5% per annum compared with the much lower UK rate of 2%. Demand forecasts made by IISI are 153m t for the Six in 1985 and 36m t for the UK.

During this period of rapid growth, however, the industry has suffered serious fluctuations in demand. These fluctuations pose a serious problem as modern large high throughput plants are only profitable when their utilisation exceeds 85–90%. Sources of Raw Material In the Six, as in the UK, the growth in steel production since the Second World War has been accompanied by a switch in raw material sources from local to imported supplies. For iron ore this has been brought about by the discovery of higher grade ores overseas paralleled by the depletion of the better-grade home ores. France is the only large scale ore producer in the Community (56m t in 1971) and even there output is declining. Currently France imports about 20% of its ore compared with 55–60% (70% if only the

SEQ 0068 JOB IOM8627-004-005 PAGE-0068 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

68

Hatfield Memorial Lectures Vol. III

iron content is considered) for the Six as a whole.5 The fall in the cost of shipping ore as a consequence of the increase in the sizes of ships used has been a major influence in this switch. As Fig. 3 shows, since 1963 the average size of bulk carrier in the world fleet6 has increased from 20,000 to 30,000 dwt t; for an average voyage (3255 miles for the UK steel industry in 1973) this will decrease the cost of shipping the ore by about 22%.7 Of course, the largest vessels available offer much greater reductions in the cost of transport, hence the development of deep water ore terminals to make provision for these very large ships. For instance, the development of Europort at Rotterdam has been of particular value to the West German steel industry (whose home ore production has fallen from 19m t in the late 1950s to 7m in 1970)5 where the disadvantage of the long distance from the coast has been partially offset by the use of Rhine waterway transport.

Fig. 3

Size of dry bulk carriers. Since 1963 the size of bulk carrier in the world fleet has increased, decreasing the cost of shipping ore.

While coal retains its dominance as the reductant for iron oxide, and the ECSC is still a major coal producer, output has declined in recent years and imports, particularly of coking coal, for the steel industry, have increased. West Germany is still more than self-sufficient, but the steel industries of the Netherlands and Italy, in particular, depend upon imports of coking coal, mainly from the USA. France and Belgium also import some coking coal. Luxembourg, which has no coke ovens, is dependent upon blast furnace coke from West Germany. The shipping of coal has not yet developed on the same scale as that for iron ores and this is partly because oil has been available, and relatively cheap, and is more readily transportable. It has thus increasingly replaced coke in the blast furnace. Limestone is normally found locally. The other major raw material with significance for the location of the industry is scrap. In the main the European steel industry is selfsufficient so far as scrap is concerned, although Italy, for instance, imports about 10% of its consumption. Location of the Steel Industry With the trend to importing raw materials by sea comes increasing pressure for bulk steelmaking plant to be located on the coast with ready access to the ore terminals at deep

SEQ 0069 JOB IOM8627-004-005 PAGE-0069 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

69

water sites. Currently, only about 20% of the steel production of the Six is on the coast compared with 40% for the UK, and the latter is planned to increase substantially by the early 1980s.8 Clearly there will be a trend for a greater proportion of the bulk steel production in Continental Europe to be located close to deep water. The cost of despatching the finished steel to the user, however, is a significant determinant in the economics of location. As Fig. 4 shows, the steel industry of the Six is heavily concentrated in the Ruhr–Lorraine–Luxembourg–Eastern Belgium region,5 and although the markets are more dispersed, there is a strong market concentration in the same areas. There are advantages in being close to the user, particularly for finishing mills, but also for steelmaking where the process and capital scrap originating there can be used most economically.

Fig. 4 Location of major European steelmaking regions. The steel industry of the Six is heavily in the Ruhr–Lorraine–Luxembourg–Eastern Belgium region although the markets are more dispersed.

SEQ 0070 JOB IOM8627-004-005 PAGE-0070 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

70

Hatfield Memorial Lectures Vol. III

The Technology The economies of scale to be gained by increasing the size of each production unit in the iron and steelmaking processes are being exploited in Europe as elsewhere. Figure 5 indicates the way in which the output per unit involved in the various iron- and steelmaking processes has been increasing over the last 20 years.9

Fig. 5 World’s largest iron- and steelmaking vessels. The size of the largest blast furnace has increased markedly and vessels used in steelmaking have also increased though already the maximum size of the BOF vessel is flattening out.

The output of the largest blast furnace has increased by a factor of 4 over the last 12 years and more than doubled in 5 years. The size of the blast furnace cannot go on increasing at this rate. Extrapolation of the present trends would indicate a single blast furnace providing 15m t/year in less than 10 years from now. With European annual steel production 200m t in 1985, 13 such blast furnaces would supply the whole of the Community. The sizes of the vessels used in making steel also show rapid growth, though less spectacular than that of the blast furnace, and already the size of the basic oxygen steelmaking vessel is flattening out. The economic benefits gained by increasing the size of the units diminish and ultimately become too small to offset the disadvantages of the organisational problems and the inflexibility which goes with such size. In a fluctuating demand situation the economics of operating the units below their maximum output penalises their overall economics. A further disadvantage in building very large units is that, unless they can come on full stream rapidly, they can involve a considerable financial penalty. These delays can

SEQ 0071 JOB IOM8627-004-005 PAGE-0071 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

71

stem from construction and commissioning difficulties, even if they only occur in one critical part of the entire system. Also, unless a market for the output of the new unit can be built rapidly, either by closing obsolete units or by intense marketing, the financial penalties can accumulate. Figure 6 illustrates the penalties incurred by delays in commissioning a works. Thus the optimum size of the units is determined not only by the size but also the rate of growth of the market available.

Fig. 6 Net present value versus build-up period: cost penalty for slow commissioning. The diagram shows how the profitability of a major investment is eroded as the time to build up to full production increases. In this example, increasing the build-up time from one to two years is equivalent to increasing the capital cost by 8%.

The Emerging Pattern of the European Steel Industry The emerging pattern is thus of an industry with a steadily increasing output, where bulk steelmaking is increasingly concentrated in a small number of locations, each chosen for its ready access to cheap, high quality raw materials and with good access to its principal markets. The advantages of large scale operations for iron- and steelmaking conflict with the advantages of specialisation at the finishing end and with the advantage of close contact between the works and the customer. The result may well be a trend towards

SEQ 0072 JOB IOM8627-004-005 PAGE-0072 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

72

Hatfield Memorial Lectures Vol. III

polarisation of the industry with large coastal plants (increasingly financed by joint ventures between different European steelmakers) making bulk and semi-finished products which will be transported to finishing works near the markets (probably owned by the individual steel companies). Supplies of scrap (normally centred in the major market areas) will become increasingly important in opening up economic niches for mini-mills. The major suppliers of raw materials may ultimately wish to upgrade their raw material exports. In those cases where iron ore and reductant occur in close proximity, some prereduction will be carried out before shipment, and this will be particularly the case where the reductant (e.g. natural gas) cannot be exported so economically in any other way. Thus direct reduction plants will be built in such parts of the world, but for security of supply and flexibility of operation, there will be a number within Europe, particularly where a suitable reductant and a suitable iron ore can be obtained economically. While deep water for iron ore imports will become essential, the proximity of indigenous supplies of coal, and increasingly oil, could determine which ore terminals will be developed. The landing points for the offshore oil and gas fields could eventually decide the locations of these large steel plants and also those for the smaller direct reduction plants. These possibilities would need the development of the appropriate technologies, particularly those for using oil and natural gas.

RESEARCH AND DEVELOPMENT NEEDED Having discussed broadly the likely pattern of the European steel industry of the future, I turn now to the research and development necessary to help bring this about and to meet the future with sufficient flexibility. Many areas for research are not appropriate for cooperative effort because of the need for competition, or they are too specialised. I shall indicate those areas where cooperative research between the countries of Europe could be most fruitful, summarised under the headings: steel utilisation, environment and human requirements, and raw materials and processes. Steel Utilisation Figure 7 indicates in a general way that iron based alloys can be formulated to cover a very wide strength spectrum at a reasonable cost compared with other basic materials.10 Notwithstanding its versatility and cost, steel is under constant pressure from a variety of competitive materials, each with a relatively circumscribed market range within which its peculiar properties enable it to compete with steel. In these areas there is continuous pressure to improve the properties of the various qualities. However, this is a field of product oriented research to improve competitive performance, and there is likely to be little scope for Eurosteelresearch so I will not dwell on it, other than to say that as steel products encompass a wider range of composites, for instance with non-ferrous metals and plastic laminates, so there will be a growing tendency for the scrap, which is

SEQ 0073 JOB IOM8627-004-005 PAGE-0073 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

73

Fig. 7 Cost versus tensile strength for various materials. Iron based alloys cover a very wide strength spectrum at reasonable cost. Carbon fibre in appropriate admixtures with a plastic matrix also offers a wide range of strength, but at a very much higher cost.

ultimately offered for recycling, to pose more difficult problems. In this area there should be scope for cooperative research as these problems will bear upon all members of the industry. The extent to which the steel industry, as against steel users, should engage in research on improving steel utilisation is discussed continually; the trend is certainly for increasing use of ECSC funds. One thing is clear: ECSC funds in this field, as in others, should not be dissipated on small projects. Much of this work should be reflected eventually in standards and codes of practice, the rationalisation of which should be an important task of Eurosteel promotional efforts: without it many of the advantages of the bigger home market made possible by the formation of the European Community would not be realised. There is a need for the industry clearly to inform its customers on the precise characteristics and uses for steel products in general. Coordination of the increasing volume of creep and fatigue data and their use in practical design concepts could help further the effective use of steel in many cases. As another example, work on fully instrumented steel structures with stresses and strains continuously recorded could both improve the design and enable the full benefit of the characteristics of the steel to be achieved. To illustrate this point Fig. 8 shows a box girder, such as is used in bridge construction, under test by

SEQ 0074 JOB IOM8627-004-005 PAGE-0074 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

74

Hatfield Memorial Lectures Vol. III

Fig. 8 Box girder test rig for improving design criteria for bridge construction.

the Corporate Engineering Laboratory of the BSC.11 Such an effort on a complete structure is expensive and would justify collaborative effort, preferably with major participation by user interests. Environment and Human Requirements The importance of environmental aspects to the steel industry is in no doubt. Environmental considerations need not themselves limit the general industrial growth which is necessary for worldwide prosperity, but it is essential for the steel industry to establish what is technically feasible to reduce air and water pollution and to pursue a vigorous policy of improvement, and to be seen to be doing so. This would be the best means of ensuring that environmental control requirements are based on what is practically attainable rather than just that which is theoretically possible. A framework of proper balance, as UK experience has shown, encourages industry to pursue socially responsible policies. Environmental control is an area particularly suitable for collaborative research, calling as it does for clearly defined tasks and a coordinated effort to fulfil them. ECSC funds have already been used to support two five year research programmes on air pollution in and around steelworks. The third will include water pollution and noise, this programme being aligned in the broadest sense to the recently approved Community environmental policy. The research programme should be drawn up with the help of representatives of those who will eventually apply the outcome of the research and development and should be closely connected with the research on the processes used by the industry; it is by the modification, sometimes radically, of these processes that reduction in pollution can be most effective in the long run. The need to ensure a sufficiently low level of noise within our plants leads naturally to a consideration of the role of our labour force. As Europe moves into the post-industrial phase of social and economic development, the labour force will increasingly expect more satisfying tasks. Increased use of mechanical handling and increased automation of

SEQ 0075 JOB IOM8627-004-005 PAGE-0075 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

75

plant will enable many areas of the works to involve the human operator, backed up by automatic sensors and substantial computing power, who will fulfil an essential decision making role. The whole area of social and technical understanding of men, and women, in the working environment is thus far poorly understood and there is clearly great scope for social and ergonomic research which should surely be conducted on a cooperative basis for the European steel industry at large. Raw Materials and Processes The main materials used for making liquid steel are given in Fig. 9,12 which shows the balance of materials which go to make 1000 t of steel in a typical integrated iron- and steelworks using the BOS process. The main materials involved are iron ore, scrap, reductants, and fuel, and the major need, I believe, is to increase the flexibility in the use of these materials, so as to minimise costs and adjust to shortages, whether short or long term. Figure 10 gives data taken from that much criticised publication, The Limits to Growth.13 Though the two projections, that of constant demand and that of exponentially growing demand, are neither likely to prove correct, it is clear from this data (using only reserves known in 1970) that there is no imminent shortage of iron ore or of coal (although of course the requirement at the moment is for low sulphur coking coals, and these are more restricted).

Fig. 9

Major materials used for bulk steelmaking. Balance of materials used for a typical basic oxygen steelmaking plant within the British Steel Corporation.

Iron ore is, however, not only plentiful, it occurs widely over the earth’s surface, in both developed and undeveloped countries, and it has little use other than for the production of steel so that its price should have relative stability over the years. The increasing trend away from lower grade home ores to the import of higher grade foreign ores is likely to continue. While iron ore is the only true source of our steel, scrap is a very important component of the steel made. Although the steel industry is perhaps an example to many others in its ability to recycle, and thus reduce its consumption of raw materials, there is still need for research. The steel industry recycles some 55% of its output. In making 1 t of steel on

SEQ 0076 JOB IOM8627-004-005 PAGE-0076 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

76

Hatfield Memorial Lectures Vol. III

Fig. 10 Lifetimes of known reserves of some natural resources (courtesy of Meadows et al.13). The chart shows two alternatives for the lifetimes of reserves known in 1970 for certain strategic raw materials.

average the UK steel industry consumes something like 0.55 t each of newly refined iron and of recycled scrap; of this 0.3 t is recycled scrap which has not crossed the works’ boundary, 0.12 t processed scrap which has been returned from industrial customers and only 0.13 t capital scrap, i.e. recycled fully used material.14 Allowing for the fact that in a growing market with an average product life of perhaps 10 years, the recycled capital scrap could never reach more than a fraction of the material actually turned into products by our customers, there is still room for improvement. Also, our 30% internally recycled material could be reduced by increasing the efficiency of our own processing techniques. The energy scene The principal reductant, coal, while plentiful in the earth’s crust, if often in deep seams and difficult and expensive to win, particularly in the much worked traditional coal mining areas, and the relative prices of coal and alternative reductants15 have to be considered (see Fig. 11). This is complicated because, in addition to using these for their chemical ability to reduce iron ore, they are commonly used as fuel, domestically, industrially and for the generation of electricity. The European steel industry operates within an energy market of which the UK pattern of consumption15 shown in Fig. 12 is fairly representative. It can be seen that oil’s

SEQ 0077 JOB IOM8627-004-005 PAGE-0077 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

77

Fig. 11 Prices paid by UK industry for fuel. With the advent of North Sea supplies the price of gas has fallen sharply to equal that of coal and fuel oil. Even allowing for its relatively greater efficiency in use, electricity is still substantially more expensive. Relative prices can change rapidly.

Fig. 12 UK primary energy consumption. Coal’s share of the market has fallen continuously since 1956. Oil has accounted for almost all the growth although natural gas and nuclear power are becoming significant.

share of the market has grown appreciably over the last 15 years, largely at the expense of coal, while othery energy sources together only contribute the remaining 16% of the total. The situation is now changing, however, as oil prices rise and security of supply becomes less certain (and now, 1973, is again threatened by unrest in the Middle East). Throughout the world, some resurgence of coal may be expected, nuclear generation of electricity will be increased, and the percentage contribution of oil is likely to begin to fall even if absolute consumption does not. Europe, and particularly the UK, is fortunate in

SEQ 0078 JOB IOM8627-004-005 PAGE-0078 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

78

Hatfield Memorial Lectures Vol. III

having appreciable quantities of indigenous oil, and natural gas coming on stream from the recent finds in the North Sea,16 as indicated in Figure 13. During the next five years oil production from the North Sea will increase rapidly and it is now confidently predicted that the UK will be self-sufficient in oil by 1985, and that up to 20% of European oil requirements will be met from indigenous sources by that date. In the next decade this must have a marked influence upon the energy pattern, at least in the countries bordering the North Sea.

Fig. 13 Possible oil and gas production from UK and Norwegian offshore fields. During the next five years oil production from the North Sea will increase rapidly and by 1985 should make the UK self-sufficient in oil supplies.

Energy in the Steel Industry Figure 14 clearly shows the major dependence on coking coal,17 due to its dominance as the means of reducing iron ore in the blast furnace, where coke makes an essential contribution to the strength and permeability of the burden. Coking coals, won by deep mining in Western Europe are expensive, as is the subsequent coking process, particularly with the increasing costs of anti-pollution equipment. Thus a major aim should be to reduce our dependence on them. Also we should increase the flexibility to take maximum advantage of a changing energy market and of new opportunities such as the increasing availability of indigenous off-shore hydrocarbons. Let us consider some developments and the possibilities for cooperative research. The consumption of coke in the blast furnace has been reduced steadily (Fig. 15) by better operation, improved burden preparation, and latterly by hydrocarbon injection through the tuyeres.18 Further dramatic reduction in coke consumption will require the development of the bosh injection of reducing gases into the blast furnace.19 A proposal for a major trial is being completed jointly by steel companies of Belgium, Italy and the UK to

SEQ 0079 JOB IOM8627-004-005 PAGE-0079 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

79

Fig. 14 Gross energy consumption (all purposes) per ingot tonne (all qualities) in UK (left) and ECSC (right). Both in the UK and the Six coking coal made a large contribution to the energy requirements of the iron and steel industry. The increase in oil consumption is largely for tuyere injection in the blast furnace.

Fig. 15 Coke consumption for ironmaking 1950–1972. Since 1952 coke rates have fallen steadily due to better operation, improved burden preparation and latterly to oil injection through the tuyeres.

SEQ 0080 JOB IOM8627-004-005 PAGE-0080 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

80

Hatfield Memorial Lectures Vol. III

inject a hydrogen–carbon monoxide mixture which could be generated from natural gas or oil. If successful this could lead to a halving of the coke required per tonne of iron produced. Figure 16 depicts the bosh injection system using hydrocarbon oxidation products (BISHOP). It shows the main alternatives being considered for the bosh injection trial currently being planned by the three countries. This is an example of a large project, of potential advantage to the industry as a whole, but beyond the means of one country’s research effort, and which is ideally suited for development on a European scale.

Fig. 16 Bosh injection system using hydrocarbon oxidation products (BISHOP). The main alternatives being considered for the bosh injection trial currently being planned by Belgium, Italy and the UK.

Whatever minimum coke rate is eventually attainable by replacement with hydrocarbons, some coke will still be required in the blast furnace. A further measure to increase flexibility is to produce this coke from non-coking coals. The usage of lower rank coals can be increased by pre-heating them before carbonisation and suitable technology is becoming available. Again this is of European interest, and indeed application has been made for partial ECSC funding for the building of a pipeline charging unit for two coke ovens at the British Coke Research Association to test the suitability of various coal blends for preheating pipeline or gravity charging and carbonisation in a tall or a conventional coke oven. Figure 17 gives a general view of the test plant at the BCRA. This currently includes a pre-heater, charging car and a conventional oven, and plans are in hand to add a tall oven and a pipeline charging system. There are other developments by which our dependence on coking coals can be reduced while still retaining the blast furnace, and there are alternative processes for reducing iron ore. Direct reduction processes in which the ore is reduced at a lower temperature to produce a solid product still containing the gangue, may use coal, hydrogen or a hydrogen–carbon monoxide mixture generated from a hydrocarbon.

SEQ 0081 JOB IOM8627-004-005 PAGE-0081 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

81

Fig. 17 Coke oven test plant at British Coke Research Association. General view of the test plant at BCRA. This currently includes a pre-heater charging car and a conventional oven, and there are plans to add a tall oven and a pipeline charging system.

Direct reduction has been developed to provide a raw material to augment scrap in the electric arc furnace so further increasing flexibility. In the long term, however, the direct reduction electric arc furnace route could become the major route to steel, for it is in this way that nuclear energy is most likely to be harnessed to steelmaking. Nuclear energy and steelmaking While it is necessary in the future to ensure flexibility in the raw materials we can use, this is nowhere more important than in our fuel supplies. This is an area with considerable scope for greater economy, which will be most important if the real price of energy is to increase markedly in the future. However, for reasons of conservation alone, we should be more stringent in our use of hydrocarbons and in the long run confine their use to chemical feeds rather than as providers of chemical energy. In these circumstances nuclear energy must increasingly become both the desirable and also the economic source of heat. Already power in the form of electricity is being produced economically from nuclear stations in competition with coal and oil fired stations, despite the present high investment costs of the former. Further increases in fossil fuel prices will hasten the move to nuclear stations, and will perhaps begin to close the gap between the price for electricity and fossil fuels for industrial heating. While this would favour the electric arc

SEQ 0082 JOB IOM8627-004-005 PAGE-0082 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

82

Hatfield Memorial Lectures Vol. III

route, the likelihood that electricity will seriously compete with fossil fuels for bulk industrial heating is some way off and must await the dominance of nuclear generation. However, before this is the case it should be economical to use nuclear heat directly for various process applications for, if uranium is the most economical energy source for electricity generation, it should also be the most economical source of heat for process application, including steelmaking. While this could be the case now, the high temperature needed for the essential steelmaking route does not lend itself to existing or readily foreseen reactor technology. On the other hand, a low temperature direct reduction process followed by electric melting could be adapted to use direct nuclear heat (the electric arc furnace is already consuming electricity generated from nuclear power). It is important to assess the role nuclear heat could play in the future European steel industry so that the market potential for such systems can be determined. Roughly, 50m t/year of steel (a quarter of the projected European steel production for 1985) using nuclear heat to the maximum, would only require 10 nuclear reactors each producing 2000 MW of heat. Clearly, the organisational and safety problems involved in going straight to a completely self-contained nuclear steelmaking plant would be considerable, and because the nuclear reactor needs to be large to be economical, such a single step would demand too big a decision. A more modest, yet more feasible approach would be to separate the nuclear reactor from the steelmaking complex, so keeping the inherent problems of each from interacting strongly upon one another. Also, it would no longer be necessary to match a single steel plant to the large nuclear, gas and electricity producing station. Figure 18 illustrates schematically a system proposed in essence by Hawkes and Hosegood20 some time ago and indicates a possible arrangement. Here the high temperature nuclear reactor reforms natural gas and then heats the reducing gas produced. This then converts the iron ore to a prereduced iron which is fed to the arc furnace supplied by electricity generated from the reject reactor heat, contained in the helium heat transferring gas, after it has passed through high temperature heat exchangers. With sufficient foresight the development both for the nuclear and the steelmaking plant could be pursued as a natural sequence and the least cost development path taken, so avoiding the astronomical costs that were incurred in developing the nuclear electricity generating reactors. European cooperation is essential if there is not to be a proliferation of approaches. It was with such thoughts that we founded a European Nuclear Steelmakers Club (ENSEC) to determine the appropriate strategy and clearly to define the research that will be necessary to develop the preferred scheme.

ORGANISATIONAL ISSUES After the 22 years it has taken for the cooperative research in the Community to evolve, perhaps it is time to take stock and for a new member to give some views.

SEQ 0083 JOB IOM8627-004-005 PAGE-0083 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

83

Fig. 18 Suggested nuclear steelmaking complex. Schematic diagram showing part of a high temperature nuclear reactor and a direct reduction electric steelmaking plant coupled to it. The parallel lines indicate two ways that the reactor complex could be separated from the steelworks.

Communications The wide range of interests throughout Europe makes it particularly difficult, both to distil broad and acceptable views of the research needed and to determine the priorities to be given to individual research projects. With so many interests involved it is essential that the role of representative groups is clear. There is need for a forum for debate on the scientific and technological problems and developments which relate to the industry. It is in this area that the ‘Learned Societies’ have an important role to play. The pattern in which the learned societies have developed their activities and their relationships with the iron and steel industry differ in each country. This diversity of approach could be adapted so that within the European Community each of the national societies could have a distinctive role, the one best suited to it, without the need itself to cover the whole field. The Directors and Secretaries of the metallurgical societies have been meeting twice a year for many years to coordinate their activities and avoid duplication of conferences, etc. I hope these contacts develop, certainly between the European societies, to produce a coordinated pattern of activity. The need to exchange technical views, and to construct and to develop cooperatively concepts for joint European research activity is now more necessary than ever it was.

SEQ 0084 JOB IOM8627-004-005 PAGE-0084 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

84

Hatfield Memorial Lectures Vol. III

Technical Service and Research Centres Eurosteel collaboration would be advantageous both in the technical service and the research field. The concept of ‘centres’ equipped with the best staff and equipment for their particular service or research, often too expensive to be run on purely national lines, could be encouraged. An appropriate allocation of central funds to such centres could be important. There are numerous potential ‘centres’ throughout Europe, some independent, some already national research centres, some acting as cooperative steel research institutes and others integral parts of steel companies. These could both provide services and conduct research in special areas. Clear, distinctive, yet complementary roles for these laboratories will take many years to crystallise. However, an early recognition of the need could do much to move effortlessly in this direction, without the need for any ‘master plan’ other than a desire to improve the effectiveness and efficiency of our always scarce research staff. Research Strategy The Commission is at the centre of the biggest European steel research collaborative effort, paid for by the steel industry through the levy upon it. It is therefore essential for the Commission to be fully informed of the steel industry’s preferred strategy for this collaborative research, and, subject to constraints, such as any imposed by the ECSC treaty, this strategy should be used to help in selecting the research proposals. It might be thought that with the extensive use of experts, many from the steel industry, on its advisory committees, the Commission was able to do this. However, these experts are appointed as individuals. Clearly there is need for a steel industry organisation to produce, and regularly update, a strategy for collaborative research into which the Commission could inject its views. This organisation should represent those who have the need for the research and who will develop and apply its successful outcome. L’Association Europ´eene pour la Promotion de la Recherche Technique en Sid´erurgie (AERES), associated with the Club de Sid´erurgists could fill this role. Allocation of Funds Once such a strategy is available and known, new proposals for collaborative research would increasingly accord with it. There would be few irrelevant proposals and the bulk of the proposals could be sorted and coordinated before final submissions were made to the Commission. Indeed at present the sheer task of processing and assessing the waves of uncoordinated and untranslated proposals that sweep through the Commission and its advisory committees several times a year, is too great. The Commission’s executive committees (there must be about a hundred of these) only deal with narrow technical

SEQ 0085 JOB IOM8627-004-005 PAGE-0085 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

85

areas and are often more concerned with the technical details rather than with the broad needs for the individual research projects or with the eventual application of their outcome. The numerous and often small research investigations have made it very difficult to take broad views. A concentration upon more clearly defined and important large projects would very much ease the whole communication and administrative problem involved in Eurosteelresearch. This need had become even greater with the enlargement of the Community. This and assistance in coordinating the individual projects in the light of agreed strategy by the industry could leave the Commission staff freer to deal with the broader Community aspects of the research, including the coordination of ECSC funded with other Community funded researches (in this context it is significant that non-ECSC funds are now becoming available for research on matters of steel industry interest, such as pollution control, corrosion, and plant engineering). Also, we should prepare for the agreed eventual merging of the three Communities and the consequential revision, or possibly the end, of the special levy which only the European coal and steel industries bear.

SUMMARY AND CONCLUSION The issues involved in determining the future pattern of steelmaking in Europe can be outlined as follows: (i)

(ii) (iii) (iv) (v)

The larger markets will lead to larger production units situated near to deep water to benefit from low transport costs of iron ore, while the finishing processes will be near to the customer. To achieve economic operation during demand cycles, there is need for plant with lower capital cost to augment the large production unit. There is a need for flexibility in the use of the main raw materials and particularly fuels, where indigenous supplies will be important. The availability of scrap will diminish as far as internally arising scrap is concerned, while capital scrap will become more difficult to process. The electric arc route could increasingly provide flexibility, provided any lack of scrap could be made up by prereduced iron.

The objectives for the research, which could with advantage be done cooperatively, might be as follows. (i) To develop plant and processes acceptable to our environment. (ii) To improve working conditions, develop sensors and instruments to enable automation and human control to be improved, and adapt mechanised handling devices to the needs of the industry. (iii) To widen the types of indigenous coals which may be used for coke in the blast furnace, and produce alternative reductants to coke.

SEQ 0086 JOB IOM8627-004-005 PAGE-0086 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

86

Hatfield Memorial Lectures Vol. III

(iv) To increase the ability to use oil and natural gas for direct reduction of iron ore while encouraging the development of a high temperature nuclear reactor eventually to minimise their consumption as fuel for steelmaking applications. (v) To improve design concepts to take maximum advantages of steel’s unique range of properties. Organisationally this would require the following. (i)

There should be a cooperative research strategy, indicating priorities, produced by the industry. (ii) Individual research projects should be of major importance to the European steel industry, cooperative in character, and fewer in number to ease communication. (iii) There should be encouragement for individual research institutions and company laboratories to specialise in particular research areas or on certain projects so that they will complement each other in areas appropriate for cooperative research. Similarly there should be encouragement of existing laboratories to provide technical service centres across Europe. All this points to a time of new opportunities for Eurosteelresearch, but to take advantage of them, the European steel industry must know exactly what it needs and be willing to take the necessary steps.

ACKNOWLEDGEMENTS I am most grateful for the help and understanding I have had from my European colleagues and I should like to thank all those who have helped in collecting data and illustrations for this paper; in particular R. Chapman, D. M. Cowie, Dr S. Klemantaski, A. Post, M. Smith and also D. Smith for preparing the illustrations.

REFERENCES 1. EUROPEAN COMMUNITIES: Medium Term General Directive on Research and Development in Steel, (2025/III-B/72F) European Commission, Brussels, 1972. 2. ‘Brown books’ (Statistics of the iron and steel industry of the United Kingdom) for the years 1938, 1955 and 1972. ‘Development in the iron and steel industry, Special Reports for 1957 and 1964’; pub. HMSO for Iron and Steel Board. ‘International steel statistics world tables 1970’; British Steel Corporation, London, 1973. 3. Estimates based on delivery, import and export data taken from Ref. 2 above. 4. Projection 85, International Iron and Steel Institute, Brussels, 1972, 4. 5. ‘Steel in the Nine, 1973’, reprinted from British Steel, Jan/Mar. 1973. 6. World Bulk Trades, Fearnley and Egers Chartering Co. Ltd, London, 1973. 7. L. R. P. PUGH: JISI, 1973, 211, 461.

SEQ 0087 JOB IOM8627-004-005 PAGE-0087 TWENTY-FOURTH LE REVISED 19DEC04 AT 20:18 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Eurosteelresearch

87

8. BRITISH STEEL CORPORATION: ‘Ten year development strategy’, Cmnd.5226, HMSO, London, Feb. 1973. 9. W. F. CARTWRIGHT: JISI, 1969, 207, 729, and 33 Magazine, 1972, 10, July, 25–27; Aug., 32–33; September, 34–38 and 47–49. 10. R. L. CRAIK and I. L. DILLAMORE: private communication. 11. Steelresearch 72, British Steel Corporation, London, 1972, 31. 12. BRITISH STEEL CORPORATION: Annual Statistics for the Corporation 1972, British Steel Corporation, London, 1973. 13. D. H. MEADOWS et al.: The Limits to Growth, Earth Island, London, 1972. 14. Scrap is Our Business, British Steel Corporation, London, 1971. 15. DEPARTMENT OF TRADE AND INDUSTRY: United Kingdom Energy Statistics 1973, HMSO, London, 1973. 16. D. M. COWIE: private communication based on published information in Petroleum and Petrochemical International, Oct. 1973. 17. Based on energy consumption data taken from ‘United Kingdom Energy Statistics 1973’, op. cit., and ‘Energy statistics’ (pub. by the Statistical Office of the European countries) taken together with production data shown in Fig. 1. 18. W. F. CARTWRIGHT: JISI, 1971, 209, 89, and private communication from Iron and Steel Statistics Bureau. 19. Adapted from Steelresearch 71, British Steel Corporation, London, 1971, 10. 20. D. A. HAWKES and S. B. HOSEGOOD: Alternative Routes to Steel, Iron and Steel Institute, London, 1971, 130.

SEQ 0089 JOB IOM8627-005-006 PAGE-0089 TWENTY-EIGHTH LE REVISED 19DEC04 AT 20:19 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE TWENTY-EIGHTH HATFIELD MEMORIAL LECTURE

From Invention to Industrial Development L. Coche At the time the lecture was presented Lucien Coche was Director General of l’Institut de Recherches de la Siderurgie Fran¸caise (IRSID). The lecture was given in Sheffield on 24 November 1978.

The structure and efficacy of industrial research organisations involves a high degree of cooperation: with other similar bodies, with universities and with industry itself. The director-general of one such European research body, Lucien Coche, stressed this theme in presenting the 28th lecture.

Dr Hatfield was an extremely bright applied scientist. In a book he wrote a few months before his death, he summarised his experience of the difficulties of creative scientific work in relation to industry; and I would like to quote what he wrote: When a piece of new scientific knowledge has been unearthed by the investigator it is some little time before either he, or someone in the world at large, realises how and where such knowledge can become of practical service; there is then very likely an even longer time to elapse before the conception of the utility of the knowledge can be construed into practical service. In the case of most applications of knowledge, many practical details have to be worked out and manufacturing plant devised and equipped. Also the potential users or consumers amongst the public must have their interest awakened, and that takes time. It is, therefore, obvious that from the inception of the idea as to how knowledge can be employed, considerable time elapses before the thing comes to fruition in everyday life and usage. That was written in 1943, 35 years ago. Since that time, industry has become more and more complicated. So much progress has occurred in many fields that further technical improvements have become a matter of cooperative work, involving many men or departments inside big companies, and usually also men working in other companies or organisations. This evolution has had a big influence on the structure of industrial research organisations. In the steel industry, possibly more than in many other industries, new technical developments can be directed to final industrial application only by the intimate cooperation of a

89

SEQ 0090 JOB IOM8627-005-006 PAGE-0090 TWENTY-EIGHTH LE REVISED 19DEC04 AT 20:19 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

90 Hatfield Memorial Lectures Vol. III number of researchers and engineers with different backgrounds, at the different stages which lead from an idea to what Dr Hatfield called ‘fruition in everyday life’. The history of a few technical developments that have recently entered industrial practice can be used for an investigation of what is necessary, taking into account the present structure of the European steel industry, to obtain that successful evolution from idea to fruition. As a first example, let’s look at the application of controlled rolling to the production of weldable plates of high properties. Much work was done in various countries, in the 1920s and 1930s, on the theory of rolling and plastic deformation of metal. When the French companies created an Institute for collective steel research, a group in charge of rolling techniques endeavoured to improve our knowledge of the fundamentals of rolling. In 1953, Paul Blain and Claude Rossard joined IRSID and started research work on the high temperature plastic deformation of steel. Experiments were first made on a conventional tensile testing machine, which was equipped with a heating furnace in order to control the sample temperature, and with a cam system to control the elongation speed applied to the sample. Subsequently, they used a hot torsion testing machine to attain higher elongations without breaking the sample. Such machines had been used for many years to measure the forgeability of steel. The one used by Blain at IRSID was modified to give an accurate measurement of the torque during the torsion test. From the shape of the curves representing the torque against number of revolutions, conclusions were possible about the structural changes of the steel. The hot torsion machine was then equipped with a fast quenching system, so that it became possible to see the grain structure at various stages of plastic deformation. Recrystallisation phenomena, either static or dynamic, were thus observed and better understood. In the University of Sheffield, at the same time, Dr Tegart was working on the same subject, with similar equipment, and a friendly and fruitful cooperation was established between him and Blain. Some work had also been done at IRSID by Blain and Fazan on the fundamentals of the rolling operation. They discovered that the stress and strain distribution in the superficial layer of a torsion sample was the same as inside the hot ingot during the rolling operation, so that one could expect to be able to predict the structure of a hot rolled plate by simulation on a hot torsion machine, with corresponding deformations, same number of passes, same temperatures and same waiting times between passes (Fig. 1). Unfortunately, this similarity of stress and strain distribution was valid only for a thin layer at the surface of the torsion test sample. It was possible to use microscopic observation as a means to observe the metal structure which was to be expected for the industrial rolling operation. But it was not possible to use an appreciable volume of metal for mechanical testing. This was undoubtedly a disadvantage of the torsion testing simulation; in order to predict physical properties of plates, it was necessary to calculate them from the metallurgical aspect, grain size distribution and similar characteristics.

SEQ 0091 JOB IOM8627-005-006 PAGE-0091 TWENTY-EIGHTH LE REVISED 19DEC04 AT 20:19 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

From Invention to Industrial Development

91

Fig. 1 Time to the beginning of static recrystallisation, as a function of temperature, for values of rolling reduction between 20 and 80%. Diagram on left is for C–Mn–Al steel, on right for a steel containing niobium, which makes for slower recrystallisation. Research at IRSID led to new controlled rolling methods for these steels.

In the 1960s, automation was introduced in rolling mills. It required a closer control of operating data such as temperature, speed, torque and roll pressure. Since another group of IRSID researchers was working on automation of a plate mill they had an opportunity to take measurements and to collect accurate data on an industrial plate mill operation. They had close relations with the research group working on torsion testing, and jointly conducted experiments to test the industrial validity of those simulation methods. The results were unexpectedly good; proof was given that it was possible to simulate the hot rolling of plates, and have an excellent knowledge of the steel structure and physical properties, just by looking at the torsion sample, and deducing the physical properties from the optical observation. It was also possible, from torque measurements on the torsion testing machine, to predict the strain in the stand and rolls of the rolling mill. Measurements were made, and found to be in good agreement with the predictions. So it became possible to imagine new types of rolling operations, try them on the torsion machine, observe the resulting structure, deduce their physical properties, use the torque measurement to calculate the strain and check that an industrial experiment could be done without any risk of breaking a roll or the stand itself. Thus, the next step was to realise an industrial scale rolling test. Practical experience confirmed that those tests were always immediately successful. In the same period, much development was being done in the field of high resistance weldable steels to be used in shipbuilding and pipe lines. Niobium semi-killed steels were first used. Then in Europe and America controlled rolling was used, with the idea of getting fine grain structures and better mechanical properties (higher elastic limits and lower transition temperatures) through plastic deformation at temperatures lower than those used in normal rolling operations. In the Soviet Union, controlled rolling was applied for beams and bars.

SEQ 0092 JOB IOM8627-005-006 PAGE-0092 TWENTY-EIGHTH LE REVISED 19DEC04 AT 20:19 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

92

Hatfield Memorial Lectures Vol. III

Plates became thicker and mechanical properties had to be higher. All major producing companies were competing for big markets. A new torsion machine specially designed for these plate rolling simulations was built at IRSID with induction heating, in order to get good temperature control, and with a computer to facilitate the simulation of multi-pass rolling operations. This machine was used for helping two plate producers. Usinor and Dillinger Huette. A new method for rolling plates was developed, the commercial Multiphi process. In 1974, using that process, these plate producers gained big orders, with profits which paid back more than the past research expenses. This new torsion machine has also been used to simulate hot strip mill and wire rod rolling operations. It is an extremely useful piece of equipment, which we are still using for all kinds of controlled rolling investigations.

FROM IDEA TO COMMERCIAL SUCCESS Industrial profits were first obtained about 20 years after the beginning of our research work in that field. A long succession of laboratory experiments has been necessary, since three successive testing machines had to be built; and industrial experiments lasted more than five years. A lone researcher, even though he might be a genius, would not have been able to lead that work from the beginning to industrial application; success could be attained only by cooperation by several teams which have been working on torsion testing, on the fundamentals of rolling, on precipitation hardening and weldable steels, and on rolling mill automation. Among the researchers involved in that development work, some had to be metallurgists. Others had to master mathematics and the theory of plastic deformation. Others had to be able to understand automation problems and to know a lot about computers, software and hardware. Others had to be field measurement engineers. All had to be experienced people in their fields, and no one single man could have had knowledge of all the various fields involved. Such intimate cooperation would have been extremely difficult, if not impossible, if those teams had been located in several distinct companies or organisations. Though Nippon Steel, US Steel, British Steel and the Soviet Union now have such size that they can maintain important research organisations, at the time that work was done it would have been practically impossible in Europe without the existence of a collective research organisation. The torsion test was not a fully satisfactory one, yet it could be used for practical application. The temptation of doing further research work in order to improve the laboratory testing methods was great, and it was a good decision not to try to look for a better test, but to try to use it as it was. From the Usinor management viewpoint, the industrial experiments could take place because there was a reasonable hope that this research could bring a solution to an

SEQ 0093 JOB IOM8627-005-006 PAGE-0093 TWENTY-EIGHTH LE REVISED 19DEC04 AT 20:19 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

From Invention to Industrial Development

93

important commercial problem: success in hard worldwide competition for plates to be supplied for northern countries’ pipelines. That was a large and important market.

CONTINUOUS STEELMAKING Another story of long lasting research, which has not yet come to industrial application, is that of continuous steelmaking. In the 1950s French steel companies had many Thomas converter plants. Improving the Thomas operation was a big problem for them. Foaming and overflow losses were a difficulty; the blowing rate had to be reduced at certain periods of the blowing operation, and yet metal losses were high. IRSID had a good team of physical chemists under a well known Russian scientist in that field, Professor Paul Kosakevitch. After the Second World War, M. Malcor, who was at that time President of IRSID, engaged him, engaged assistants to work with him, and asked that research group to work, among other subjects, on the problem of Thomas slag foaming. So Professor Kosakevitch, Dr Olette and others of that same group conducted basic investigations on steelmaking slags and such physical properties as melting point viscosity and surface tension. In close cooperation with the IRSID group working on steelmaking problems, they observed what happened in Thomas converters, investigated the foam properties and discovered that the analysis of the small steel balls trapped in Thomas converter foam was excellent, with very low phosphorus and sulphur contents. Then the idea of using that foaming effect for continuous steelmaking came into their minds. No person would pretend to be the one who had that idea first. It came naturally to the minds of a group of men working together, starting either from semi-fundamental studies, or from industrial problems. It would be a long story to describe all the steps which led continuous steelmaking from that idea to the last step currently attained, which was a semi-industrial operation at Hagondange. There were first laboratory experiments, then a pilot plant built at Maizi`eres-l`es-Metz for about 3 tonnes per hour, then another pilot unit, also at Maizi`eres-l`es-Metz, for about 10t/h, then the semi-industrial plant at Hagondange, which produced 20–25t/h. Two stage operations with two slags were also tried. Operation of the experimental plant, erected in the steelworks of Hagondange, began in 1969; improvements had to be made and tested. In 1971, that plant was operated for three months; the first campaigns had to be stopped after a few hours, but week after week the technological difficulties were eliminated. And finally the engineers were able to get the results which were wanted: a full five days operation with 108 hours of actual blowing time and a smooth operation throughout that week. In 1973 and 1975, the Hagondange steel plant was used again, for two kinds of additional research. One was to use scrap, and see how massive scrap could be melted; it

SEQ 0094 JOB IOM8627-005-006 PAGE-0094 TWENTY-EIGHTH LE REVISED 19DEC04 AT 20:19 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

94

Hatfield Memorial Lectures Vol. III

was quite successful since, though it was only a small scale unit with 1 meter diameter at the cylindrical part of the reactor and 0.6 meter at the bottom, bloom scraps 200 mm square could be melted as easily as sugar lumps in tea. The other line of research was on the use of preheated granulated pig iron’, in addition to or instead of liquid hot metal. With phosphorus iron preheated at 850°C, it has been possible to use granulated iron alone, without any liquid hot metal. In bigger units, of course, the minimum preheating temperature would be lower. A total of 22,000 t of steel has been produced in that Hagondange pilot plant. Yields and all elements of the cost were measured; practical problems for starting, stopping and maintenance were encountered and solved. Almost all the problems that could be investigated at pilot-plant scale have been investigated and solved, at least for a single-stage operation. More work could have been done in order to operate the two stage process, but that would have needed more money than IRSID budgets could afford.

CONVINCING THE SCEPTICS Though no larger plant has yet been constructed, it is worth reviewing the various steps between the original idea and the final application. As for what we may consider to be the successful first part, some remarks expressed about controlled rolling are valid for continuous steelmaking as well: the length of time between the beginning of research work and industrial application will be more than 20 years, with a long succession of laboratory and pilot plant experiments. The cooperation of several research teams with different kinds of knowledge and experience has been absolutely necessary, and was possible only because those men were in a single big research organisation. Too, the men leading the project had a deep belief in its future: they were convinced of the advantages of continuous steelmaking. Now, let us consider the reasons why no bigger plant has been built. The investment for a 1.5 million t/y steelmaking unit is probably not less than £20 million. Second, the advantages of continuous steelmaking are not very obvious for people without strong technical experience. In the case of controlled rolling, the hope of winning arctic pipeline markets was a good argument that any non-technical director could easily understand. In the case of continuous steelmaking, if the technical director explains that, in addition to lower investment costs, the company may take advantage of easier and better quality control, as well as environmental control, those arguments are not easily perceived or appreciated by his colleagues with less personal technical experience. And they also overestimate the risks. I do not see any practical means for overcoming that kind of difficulty, because there is too much distance, and not enough communication, between the men who can decide a £20 million investment, and those with a deep enough technical knowledge of the process to be aware of its advantages.

SEQ 0095 JOB IOM8627-005-006 PAGE-0095 TWENTY-EIGHTH LE REVISED 19DEC04 AT 20:19 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

From Invention to Industrial Development

95

From these two examples, we can draw conclusions about the necessary conditions for the industrial success of research. First, chance is important. But that does not lead to any practical conclusion. Second, the result of research work must be considered as really useful not only by researchers, but by the people who decide investment in the industry. Research projects which do not correspond to immediate commercial needs have little chance to come to realisation, even though they may have value for the remote future. Even for minor improvements, a close cooperation of various specialists is necessary. Blast furnace operators can decrease coke rates by several points per cent, just by applying well established rules for shaft operation control. But in order to do that, they must know enough physico-chemistry to understand the basic laws of blast furnace shaft operation. They also need to cooperate with measurement specialists for top gas temperature and analysis measurements, and with computer experts for the necessary calculations. Because so many different men, with different technical backgrounds, are involved, technical progress becomes impossible if we do not either have big companies, or a cooperative system to organise joint research. In both cases, the main qualities required from research people are knowledge, tenacity and imagination, together with the aptitude to cooperate in team work. At the time Dr Hatfield was working and writing, all these qualities were necessary: but the last has become more important. Since our steel industry must survive in extremely tough worldwide competition, I sincerely hope that in all the European countries men responsible for conducting steel research will be conscious of that evolution, and be able to create the structures for increasing cooperation.

SEQ 0096 JOB IOM8627-005-006 PAGE-0096 TWENTY-EIGHTH LE REVISED 19DEC04 AT 20:19 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0097 JOB IOM8627-006-004 PAGE-0097 37TH LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE THIRTY-SEVENTH HATFIELD MEMORIAL LECTURE

Net Shape Solidification Processing of Steel, 1945–1995 M. C. Flemings At the time the lecture was presented Professor Flemings was at the Massachusetts Institute of Technology. The lecture was given in Sheffield in 1989.

Professor Flemings’ Hatfield Memorial Lecture, the 37th, with the above title, was presented in 1989. No lecture manuscript is available, but the following entry is adapted from a short review which appeared in the journal Cast Metals, and is reproduced by kind permission of the Castings Technology Institute.

One of the most important papers to be delivered in the 1989 season was the 37th Hatfield Memorial Lecture, delivered in Sheffield, England, by Professor M. C. Flemings, of the Massachusetts Institute of Technology. The title of the paper was ‘Net-shape solidification processing of steel, 1945–1995.’ The Lecture, which has become a major annual event in the European metallurgical calendar, was instituted in honour of the late Dr W. H. Hatfield, the noted Sheffield metallurgist who had, during his long association with the Brown Firth Research Laboratories, been responsible for many significant advances in ferrous metallurgy, including important aspects of the solidification of steel. Professor Flemings, author of two books and over 200 papers on aspects of casting, joined a long list of eminent figures in the metallurgical world who have been invited to give the lecture; his twin themes, expanded before a large audience, were the development of casting processes and the understanding of solidification over the half-century leading to 1995. Professor Flemings began by reviewing the transformation that had occurred since 1945 in a field until then dominated by the labour intensive process of sand casting and by the production of ingots for the wrought steel industry. In the shaped casting field, growing mechanisation, developments in sands and binders and the new concept of resin shell casting had been followed in the late 1950s by the appearance of the lost foam principle and later the V process, both also embodying no-binder moulding technology. The increasing emphasis on the three attributes of accuracy, thin section capability and surface quality was also featured in the parallel emergence of lost wax investment casting and eventually the ceramic shell system on to the industrial scene. The 1970s brought the

97

SEQ 0098 JOB IOM8627-006-004 PAGE-0098 37TH LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

98

Hatfield Memorial Lectures Vol. III

counter gravity casting processes, using vacuum and pressure assisted filling to achieve half-millimetre wall thickness. Problems attending the use of pressure diecasting for steel were then emphasised, particularly the die heating aspect, to which no moulding material has yet offered a complete solution. The alternative approach was to keep the mould temperature low and cast below the liquidus, using semi-solid slurries, which offered an opportunity of achieving this aim in what was effectively a hybrid casting/forging system. Continuous casting had developed, over the period being considered, from an abstract perception to a process which may soon approach universal application for billet and slab production, with large multi-strand plants now a reality. This gave new impetus to ideas for the direct casting of strip and even wire; some work in these fields was described, including melt spinning. Pilot thin slab casters with capacities for production in the thickness range 20–40 mm and 2 m width, at rates of 2–10m/min, were in existence, but the eventual aim would be to achieve the cold rolling stage directly from the liquid or semi-solid state, requiring 1–5 mm cast thicknesses to be produced at 30 m/min. This could involve either twin or single roll systems. Professor Flemings then proceeded to review advances in the knowledge of solidification in steel castings and ingots over recent decades, first in the field of feeding principles for sound products, for which the main criteria were mentioned, based essentially on the total solidification time as emphasised in the equations produced by Adams and Caine. The mid 1950s had brought advances in the use of insulation and exothermic compounds and also the concept of feeding distance as developed by Pellini and others to attack the problem of centre line shrinkage. The early 1950s had seen the advent of the transistor and the requirement for the growth of single crystals of germanium and silicon, which gave impetus to the quantitative understanding of plane front solidification, but the extension of the same theoretical principles to dendritic growth had in the 1960s produced mistakes which indicated that the idealised concepts as envisaged by the physicists could not handle real systems. At a later stage new ideas from Scheil and others were successfully combined with the older to overcome this limitation. Special tribute was paid to the publications of Ruddle during the early phase of solidification research. The 1960s and 1970s had brought major advances in solidification science, especially application to microsegregation, macrosegregation and hot tearing. The audience was reminded of the earlier contribution of Dr Hatfield himself to the study of macrosegregation, through his chairmanship of the ISI committee on the Heterogeneity of Steel Ingots. It was pointed out that the phenomena in ingots and castings could now be simulated mathematically. Before the ’sixties, segregation had been seen mainly in terms of elements being ‘pushed ahead’ of the solidification front, together with the settling of crystals and tear filling. Between 1967 and 1970, however, the interdendritic flow of liquid, previously associated with inverse segregation, was seen to be relevant to all forms of segregation, whether in ingot, shaped, or continuous casting conditions. Mathematical and particularly computer modelling was not,

SEQ 0099 JOB IOM8627-006-004 PAGE-0099 37TH LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Net Shape Solidification Processing of Steel, 1945–1995

99

however, only important to the research worker, but also to the practical man for process simulation. Professor Flemings believed that the picture of the next five years was clear and indicated three major trends. These would be, first the progressively closer approach to net shape, second, a continued quest for improvement in the quality attibutes of soundness, cleanness and properties, and third, the drive towards lower total costs of the final component and of the whole system, thus justifying higher selling prices for the products. In specific terms, the coming period might see more automation of investment casting, with possible hybrids of investment and sand casting practice, the application of semisolid diecasting and the continuous casting of thin slab and strip and perhaps later even of wire. Other changes foreseen by Professor Flemings in the final part of his review were increased emphasis on uphill metal transfer and on the application of computers for both simulation and control. The Lecturer had been introduced by the Vice-Chancellor of the University of Sheffield and a concluding vote of thanks was proposed by Professor R. W. K. Honeycombe. Dr. P. R. Beeley, The University of Leeds, England.

SEQ 0100 JOB IOM8627-006-004 PAGE-0100 37TH LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0101 JOB IOM8627-007-006 PAGE-0101 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE FORTY-THIRD HATFIELD MEMORIAL LECTURE

The New World of Steel J. Edington At the time the lecture was given, Dr Jeff Edington was Executive Director Technology, British Steel plc. The lecture was given in Sheffield in December 1995.

Delivering the 43rd Hatfield lecture, Dr Jeff Edington promised his Sheffield audience a bright outlook for steel, the engineering material of choice for the twenty-first century. This presentation was delivered in December 1995 and the data included represent the state of technology development at that time.

SUMMARY In the early part of the next century the steel industry will be deploying worldwide new and even cleaner process technologies, new business processes and IT systems, as well as new products and systems based on symbiotic relationships with customers. It is an exciting and challenging time to be a technologist in the industry because most of the new technologies, on which all of this depends, are being developed and piloted now. The broad influence of technology on both the profitability and cleanliness of the industry as well as service to society has probably never been greater. Steel will continue to provide the world with the uniquely versatile, low cost engineering material. It fills a fundamental need in modern society and will continue to do so in such major markets as infrastructure construction, transportation, packaging, domestic white goods and engineering machinery. A key component in the overall acceptance of steel as the engineering material of choice is the fact that it is already by far the most recycled material in the world.

THE MARKET Steel is by far the largest metals industry and the second largest man made materials industry in the world, with an annual production of 750 Mt, compared with 1100 Mt cement, 100 mt plastic and 25 Mt of aluminium. As shown in Fig. 1, during the 1950s and 1960s production showed a steady growth of ∼ 6% per annum.1 However, in common with most other materials, the first oil shock transformed the situation into one

101

SEQ 0102 JOB IOM8627-007-006 PAGE-0102 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

102

Hatfield Memorial Lectures Vol. III

Fig. 1 World steel production (1950–1994).

of very slow growth with large fluctuations arising from a fine balance between supply and demand. Since the end of the Cold War, growth in Asia Pacific, mainly in China, has compensated for shrinkage in Eastern Europe. Although the market overall has not grown much, the total amount of steel traded has because of the entry of steel from Eastern Europe into world markets for the first time. During the past 20 years the distribution of steel sales between the major product market segments has remained largely constant. Recent figures for the European Union, North America, and Japan, the three major and mature geographic markets, are shown in Fig. 2 (1993 statistics). From an estimated total consumption of about 250 Mt (product tonnes), the manufacture of automobiles accounted for 16%, metal goods and containers 10%, whereas building and construction used 38%. All geographic and product markets are mature except the emerging economies of Asia Pacific. Here, growth in steel consumption is driven largely by urbanisation and construction of an infrastructure typical of the early stages of economic growth and is likely to move on to consumer goods as the populations become more wealthy.

Fig. 2 Steel consumption by sector: USA, EU, Japan (1993).

SEQ 0103 JOB IOM8627-007-006 PAGE-0103 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

103

In all markets the industry is pursuing a four pronged strategy of: ● creative product and systems development based on enhanced symbiotic relationships with customers ● improvement in product quality and customer service ● continued cost reduction ● development of a series of even cleaner, lower capital cost manufacturing processes that, depending on the technology deployed, can produce steel economically at small or large scale and from a wide range of raw materials; these will provide more flexibility to tailor a viable scale of operation and reduce environmental impact to meet the widely different needs of local markets throughout the world.

PRIMARY PROCESS DEVELOPMENTS The industry has a long history of innovation both in technology and in management systems. For example, in the last century the industry has invented and implemented three different ways of making steel via the iron ore route, namely, the Bessemer (1865), open hearth (1868) and basic oxygen (1960s) processes. More recently, combined managerial and technical innovation have led to the implementation of minimills based on electric arc steelmaking from scrap combined with near net shape casting. These are low capital cost, very high productivity plants focused on regional markets and a narrow range of products. At present there are two established process routes to make steel, the blast furnace (BF) which produces liquid iron for refining to steel in the basic oxygen furnace (BOF) and the electric arc furnace (EAF) which remelts in plant arising and (collected) merchant scrap.

BLAST FURNACE–OXYGEN STEELMAKING ROUTE The BF–BOF is a tried and tested route capable of producing 4–6 Mt/year of high quality steel from two large blast furnaces and a three vessel BOF configuration on one vary large site. For example, British Steel’s Teesside plant produces 4 Mt/year based on one blast furnace and the largest steelplant in the world at POSCO’s Pohang site produces 9 Mt/ year from four blast furnaces. Iron ore is converted to sinter by drawing air through a mixture of iron ore, limestone, and coke at 1000°C, while, in parallel, coal is heated in airtight coking ovens to produce a coke feed. Both are fed into the blast furnace which produces molten iron by reducing the sinter and lump ore with carbon monoxide generated by combustion of the coke. In the next step the BOF charge is usually 300 tons and consists of 80% liquid iron and 20% scrap to cool the metal during the exothermic chemical reactions that occur as high purity

SEQ 0104 JOB IOM8627-007-006 PAGE-0104 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

104

Hatfield Memorial Lectures Vol. III

oxygen is blown through the BOF. The oxygen combines with carbon and other unwanted elements, eliminating these impurities from the molten charge. During the ‘blow’ (30 min), lime is added as a flux. Once the steel has been refined, it is tapped into a ladle when alloy additions may be made to adjust the composition and/or the steel may be degassed. The integrated BF–BOS route is being developed continuously to increase productivity, lower costs and reduce the environmental impact.2 For example, British Steel has poineered injection of coal into the blast furnace through the tuyeres at 200 kg per tonne of hot metal (thm) allowing coke rates to be reduced from ∼ 500 to ∼ 300 kg/thm. The limit of this injection technology is probably ∼ 250 kg coal/thm. In the future more productivity and cost gains will come from oxygen enrichment of the blast beyond 60% and reinjection of reheated top gas. Further cost savings and reduction in environmental impact will come from newly developing technology that allows ore fines and fluxes to be injected with the coal. This could lead to coke rates as low as 175 kg/thm which is the minimum to service the heat requirements of the hearth. In addition there are techniques to feed the BOF with 40% or more scrap, although productivity is reduced. All of these developments will reduce the environmental impact of steelmaking by lowering the production from potentially the most polluting steps in the manufacturing chain, the sinterplant and coke ovens. In addition, as illustrated schematically in Fig. 3, the BOF can be adapted to accept a wider range of feedstock. As an alternative to scrap, direct reduced iron (DRI) or hot briquetted iron produced by reduction of solid ore by natural gas in commercial processes such as Midrix3 is already in use. One further approach is to feed the BOF from a smelting reduction furnace which produces liquid iron, but on a smaller scale than the blast furnace. Several new smelting reduction processes are under development and one, Corex,4 has already been deployed. The first major plant, operating at 300 kt/year started up in 1988

Fig. 3 The adaptable BOF.

SEQ 0105 JOB IOM8627-007-006 PAGE-0105 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

105

at ISCOR in South Africa and a 600 kt/year unit is now operating at an actual output of 700 kt/year as POSCO in South Korea. Design of a 1 Mt unit is under way. Currently, approximately 2.4 Mt of Corex capacity is on order around the world. The Corex process telescopes the coke ovens, sinterplant and blast furnace into one process and has the advantage that non-coking coal can be used. The quality of the iron is identical to that obtained from the blast furnace and, because this process substitutes one fully enclosed process for one enclosed (blast furnace) and two open to the air (coke ovens and sinterplant), its emissions are considerably less.5 This technology also has the advantage that it produces large amounts of spare energy in the form of a gas similar to natural gas which can be used to make electricity, DRI, or oxygen depending on the local needs. For example, the gas from a 600 kt Corex produces ∼ 125 MW of electricity. As a result of the the developments summarised here, there is now a choice of cleaner iron ore based processes for making steel via a BOF that will enable plant to be tailored both for the size of the market anywhere in the world and to fit the local feedstock (iron ore, coal or natural gas). The price of all of these raw materials is determined in a world market with plentiful supply and, therefore, the fundamentals of the BOF route will always be low cost with little fluctuation in the real price of raw materials.

ELECTRIC ARC STEELMAKING ROUTE (EAF) The electric arc furnace (EAF) is illustrated schematically in Fig. 4 and is a well known and reliable technology that has been used for 85 years to make steel. Traditionally it has been an ac three electrode furnace fed entirely with scrap and the quality of the steel produced is directly dependent on the purity of the scrap. However, alternative feedstock can be used, such as DRI, hot briquetted iron, iron carbide,6 or liquid iron from either a

Fig. 4 The adaptable EAF.

SEQ 0106 JOB IOM8627-007-006 PAGE-0106 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

106

Hatfield Memorial Lectures Vol. III

Corex or a blast furnace. Because they come from an ore based route, they all provide a clean feedstock and are a direct substitute for top quality low residual scrap. Several recent developments of the EAF have improved its efficiency.2–7 The dc furnace offers advantages over the ac furnace in reduced electrode costs and reduced electricity network disturbance as well as simpler construction and operation. Although ac furnaces are still preferred at power inputs above 90 MW, the recent development of larger diameter electrodes and a twin electrode dc furnace will probably raise the upper power limit. Furthermore, the use of natural gas for in vessel and external scrap preheating, including the use of twin shell furnaces, have all increased production rates, reduced electricity consumption and lowered operating costs. For example, Tokyo Steel is installing a twin electrode dc arc furnace with a central preheating shaft from which scrap is fed continuously between the electrodes. The target output is 150 t h–1 utilising an electricity consumption of 230 kWh/t, which compares favourably with 450 kWh/t for a conventional ac unit. One further potential development is a high aspect ratio furnace that can accommodate a charge of either all liquid iron or all coarse lightly packed scrap, or any mix of the two, depending on the available resources. However, this is not available commercially at present. Depending on which of these technologies is chosen, the economically viable range of an EAF based melting shop is 0.5–2 Mt/year and the costs can be minimised for the particular location. However, unlike the BOF route input costs are not determined in an international market with plentiful supply. Electricity prices in some parts of the world are determined by governments through control of supply and taxation, so the economics depend on geography. Furthermore scrap prices vary considerably (× 2) through the business cycle depending on the balance between supply and demand, although the increasing availability of DRI will place an upper limit on the price of best quality scrap. As a result, although this is the low capital cost route, it is not always the best option for low overall cost.

IMPLEMENTING NEW PROCESSES IN THE FUTURE As mentioned above, the different process combinations available to make liquid steel can be tailored for local needs for capacity, product mix and quality, and feedstock. By the year 2010 it is estimated that about 45% of the world’s steel will be produced via the arc furnace and the remainder will be BF–BOF or Corex. About 20% of the overall supply for the EAF will be DRI, 55% will be obsolete scrap, with the remaining 25% scrap from manufacturing process losses. The choice of the various technologies will be dependent on the local infrastructure, such as the availability and cost of natural gas, electricity, coking coal, ore, and scrap. For example, in Western Australia the availability of natural gas and iron ore could enable DRI to be produced at the mine then exported as a higher value added product than iron

SEQ 0107 JOB IOM8627-007-006 PAGE-0107 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

107

ore. Again, in Indonesia local iron ore and non-coking coal are more compatible with Corex than BF–BOF technology. Consequently, the key issue in making the appropriate technology choice is to have an accurate cost model which can be used in different parts of the world to compare the capital and operating costs of particular steelmaking processes. The output of one such model based on the 1995 input prices for energy, raw materials, and manpower in Southeast Asia is shown in Fig. 5. In this case, for a capacity of 2.5 Mt/year, the Corex/BOF route is the most economical.

Fig. 5 Liquid steel costs in Southeast Asia for different steelmaking combinations.

ENVIRONMENT The industry has invested heavily to ensure that modern environmental standards are met and considerable progress has been achieved. For example, most major companies have reduced discharges to air and water by 90% since the 1960s. Indeed, to meet HMIP requirements in the UK, the emission of particles to air by sinterplants has reduced from 460 to 60 mg m–3. For EAFs emissions have reduced since 1966 from 115 to 15 mg m–3. This has been achieved by fitting filtration systems which capture most of the dust and fume which then becomes an economic source of reclaimable material. New technology developments range from the application of advanced computing technology to enhance the capacity and effectiveness of air pollution control equipment to development of instruments for monitoring continuously air and water quality. Considerable success has been achieved in the economic use of waste material from the blast furnace. Now up to 80% of this process waste can be recycled, processed into cement or sold to the road construction industry as a replacement for quarried materials. Several opportunities are being investigated aimed at using BOF slag to build roads or condition soil. In addition, extensive work is also being done to develop more value added products from waste and new technologies to recycle wastes previously sent to landfill. In the future consumers are likely to begin to make buying decisions on the (perceived) environmental impact of the manufacture and use of materials and indeed of the

SEQ 0108 JOB IOM8627-007-006 PAGE-0108 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

108

Hatfield Memorial Lectures Vol. III

environmental performance of individual companies. The most obvious recent example was the disposal of the Brent Spar oil platform which led to a boycott of the company’s products in Germany. Another is the need in some Scandinavian countries to meet environmental standards in manufacturing before being allowed to bid on some contracts. Another is the increased emphasis on life cycle analysis as a means of comparing the environmental impact of different materials. Much work needs to be done here to ensure that the industry fully understands these developing issues and responds effectively. We start from a strong base in which we have by far the most recycled material as described below and we already recycle and use much of our process waste. We have many opportunities to improve.

RECYCLING Steel is the easiest to recycle of the materials used in large volumes. Furthermore, steel is not downgraded when it is recycled and reappears in buildings, cars, engines, earthmoving equipment, tractors and white goods. Approximately 80% of all steel available in the UK is recovered and recycled so that almost half of the UK steel production is derived from scrap. The total amount of steel recycled annually in the world is 320 Mt, just under half of the world production at 750 Mt. The automobile sector is a good example of the excellent recycling track record of steel. In the UK, about 2 million scrapped cars are recycled each year, representing 97% of all redundant vehicles. The challenge for the future, should society require it, is to design and build consumer goods for long life followed by reuse rather than recycling. This is already done to some degree in many small industries based on reconditioning industrial machines, car components and household equipment in which steel components figure large.

ENERGY The steel industry is a large consumer of energy because carbon is the key element in the chemical reduction of iron ore in the BF/Corex and the purification step in the BOF. All of this absorbs 45% of the fuel consumed, and basic thermodynamics controls the chemistry so there is no alternative here. Nevertheless, there are planty of opportunities to work on the balance of 55% and most companies have considerably reduced their energy consumption during the past decade. For example, British Steel now uses 25% less energy in total than in 1980. The future lies in improving the operating efficiency of the processes in each individual plant by making many small improvements in energy consumption, the Japanese kaizen approach. These range from burners that will strike the optimum balance between energy efficiency and air pollution to fan speed control in extractions systems.

SEQ 0109 JOB IOM8627-007-006 PAGE-0109 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

109

DOWNSTREAM PROCESS DEVELOPMENT The past decade has seen considerable activity to improve the processes used to manufacture flat or long products. Like steelmaking this has been driven by the need for lower capital and operating costs and again like steelmaking this has led to telescoping process steps. This is best illustrated for hot rolled coil production. The conventional way is to start with a continuously cast strand typically 230–300 mm thick and up to 3 m wide. This is processed through a roughing mill, then a finishing mill, to produce coil. The production line is typically 500–800 m in length. The hot working schedule combined with the correct cooling rate enables the desired final shape and metallurgical properties to be achieved. However, near net shape casting techniques have been developed to simplify the process by minimising the amount of rolling required for both flat and long products.

THIN SLAB CASTING This process route is illustrated schematically in Fig. 6, although the details vary between the various manufacturers.8 The cast strand can be up to 2 m wide and 15–125 mm thick. This can be further reduced by in line rolling of the strand to 15–30 mm which is the normal entry thickness to the tandem mill in conventional rolling. Alternatively, casting may be followed immediately with a squeeze reduction before the centre of the strand has solidified. In effect the production line is reduced in length from 800 to 260 m and at most only a small roughing mill is needed. Combined with electric arc steelmaking, capital costs per tonne are less than one-half those of a conventional BF–BOF hot strip mill, but operating costs are only competitive where electricity is cheap and/or scrap is in the lower part of its price cycle. Surface finish is the limiting factor. At present 50–60% of the product range is attainable in quality terms, but it is expected that 80–85% will be reached soon with only the highest automobile and white goods grades out of reach of the process.

Fig. 6 Schematic diagram of thin slab casting route.

SEQ 0110 JOB IOM8627-007-006 PAGE-0110 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

110

Hatfield Memorial Lectures Vol. III STRIP CASTING

Direct casting of strip < 8 mm thick is the next step in cost reduction because it reduces the investment in rolling mills. On paper, when combined with electric arc steelmaking, capital costs per tonne could be less than one-quarter and operating costs less than half those of a conventional BF–BOF hot strip mill. It is under intensive development. As shown in Fig. 7, steel is teemed via a protected shroud directly into the strip casting mould using a feeding system designed for minimum turbulence and even distribution across the mould width. The strip is fed, via pinch rollers, into a hot coiler and casting speeds of up to 100 m min–1 are possible, depending on product thickness. Througputs of up to 500 kt/year are possible for a single twin roll caster.

Fig. 7 Schematic diagram of direct casting of strip.

Strip casting of stainless steel is easiest because is has a lower melting point compared with conventional carbon steels and much greater oxidation resistance. It is close to commercial realisation, with Nippon Steel/Mitsubishi able to cast 10 t coils at 1330 mm wide. Ugine has reported production of 25 t coils of stainless at 860 mm wide and POSCO/Davy have recently installed a 1300 mm wide plant in Korea.9 BHP has reported success with carbon steels, a much more difficult product. Several 18 t coils 1400 mm wide × 1.9 mm thick have been produced at its Port Kembla works. Commercial processes for stainless production will be available within 2–3 years and perhaps for carbon steels in the early part of the next century. Initially surface quality will limit applications, but this will be overcome with time although probably not for all grades.

BEAM BLANK CASTING Beam blank casting in the process route to sections excludes typically five rolling stages in the breakdown mill before final rolling in the universal mill. The process was developed

SEQ 0111 JOB IOM8627-007-006 PAGE-0111 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

111

at BISRA in 1968 and has been adopted for mills with a limited product range. Manufacturing a wide product range entails too many moulds and mould changes which translates into high capital and operating costs. However, design improvements have enabled more final shapes to be produced from a given cast shape and have enabled thinner web, wider flange sections to be produced. Thus, recent installations at Arbed have enabled a range of sections from 186 × 200 mm × 34/6 kg m–1 to 1008 × 302 mm × 349 kg m–1 to be made from only three cast beam blanks.10

BUSINESS PROCESS SIMPLIFICATION Information Technology (IT) The steel industry manages one of the most challenging manufacturing and product delivery systems in the world, consisting of a mixture of sequential batch and continuous manufacturing processes. This is exacerbated by orders that are sometimes much smaller than the minimum manufactured batch size and this must be harmonised with the customer’s requirement for just in time delivery and the steel company’s need for minimal stock. All of this leads to considerable difficulty in maximising the profit by optimising the loading of the assets while maximising the profit ability of the order mix. In this world effective IT systems are essential to run manufacturing and all companies have them. However, the new nexus between computing, communications, and information is providing the opportunity to move to totally new levels of efficiency and customer service that are already beginning to revolutionise the industry and will produce winners (those who implement the new technologies creatively and quickly) and losers (those who don’t) on a grand scale. All companies are progressing up the ladder shown schematically in Fig. 8. Here, the overall principle is to share information more and more broadly as the organisation progresses from the bottom to the top of the diagram. This will enable people to operate

Fig. 8 Business transformation and the IT revolution.

SEQ 0112 JOB IOM8627-007-006 PAGE-0112 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

112

Hatfield Memorial Lectures Vol. III

more effectively as individuals and to cooperate in teams. As a result the organisation will deliver much greater levels of service and efficiency to customers. At the top level is a new model for the creation of profit, the internetworked business, and at the bottom is a more effectively contributing person. For each level there is an enabling technology that is already available. Very little needs to be invented; the initiative lies with management and technologists to use it creatively. The opportunity is to develop a business enterprise which is both internally and externally integrated in a completely different way. Internally it will be integrated to greatly simplify and reduce the activities associated with order acceptance, manufacturing, and delivery. Externally it will be integrated with its suppliers, with the knowledge base in universities and consultants and other service providers, with up to date market databases and most importantly with engineering product development software shared with the customer and other vendors that cuts product development time massively. The way in which these opportunities have already been taken varies from industry to industry.11 For example, in retailing, Wal-Mart Stores Inc. is the world’s largest and most successful retailer with 1994 sales of US$80 billion from stores in the USA, Canada and Mexico and is an innovator in the business use of IT. Bar codes, read at the cash register, are communicated directly to the factory so that replacement goods are sent directly to the shop, so that 97% of them never see a warehouse. Sales drive the order process and the company can offer quality goods at low prices because stocks are minimal and there is no investment in warehouses. Levi’s Personal Pair Jeans mass customised jeans is another example in retailing. A customer can be measured by a clerk in a shop who is guided by touch screen software that does not require special computing skills. Using Lotus Notes, the measurements are transmitted over the Internet to the manufacturer who mails the jeans to the shop or the home within 3 weeks. The customer happily pays US$10 more for the jeans because the fit is guaranteed and profit is high because there is neither inventory nor distribution, which together account for 75% of the costs in the clothing industry. If the software is available at home from the Levi’s Jeans homepage on the Internet, there is no need for a shop; we have electronic shopping. In an industry where US$25 billion of manufactured clothing remains unsold or sells only after deep discounts, such preselling is revolutionary. Finally, the Boeing 777 is an example of a revolution in product design and development. This was a US$4 billion 5 year programme to design an aircraft digitally. Design packages Catia, Elfini, and Epic were shared on a network between Boeing, its 500 suppliers worldwide, and its four initial customers. The aircraft was designed and flown in digital space where its flying characteristics could be tested. Customers could contribute to the design helping to ensure that their requirements and innovative ideas could be incorporated and that the aeroplane worked. The network of suppliers ensured that the aircraft could be manufactured. The result was a reduction of design time by 30–40% and a more efficient and less expensive aeroplane that is both cheaper and easier to service. What will happen in the steel industry remains to be seen, but it will surely be exciting and challenging, and it will produce opportunities that are at present only shadows in the minds of our people. Unlike most technology development where the outcome is reas-

SEQ 0113 JOB IOM8627-007-006 PAGE-0113 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

113

onably predictable over the next two decades, this is not and is only limited by the ability of management, engineers, and technologists to see and act on the opportunities more creatively and faster than the competition.

PRODUCT DEVELOPMENTS Success in the marketplace is the key measure of the viability of the industry. Steel is a major material in the automotive, construction, packaging, industrial plant and mechanical engineering markets. The distribution was shown above in Fig. 2. Development of new steels for all of these markets has been extremely successful. For example, more than half of the steels used in cars built in Europe today were invented in the past 5 years. However, the future lies with developing a combination of better steels and systems solutions that meet consumer needs. We must help design our material into consumer products to improve performance, simplify manufacturing and improve reliability. It is this approach that will create demand and maintain the profitability of the industry. Six examples that point the way are discussed below.

THE AUTOMOBILE In 1994, world production of passenger cars was about 35.5 million, of which the European Union produced 13 million. Despite the trend towards weight reduction and the increased competition from alternative materials, about 0.5 t of sheet steel is used in the body and chassis of an average midsized saloon. This represents approximately 50% of the total vehicle weight and is about 25% of sheet steel production in the three major mature geographic markets, the European Union, North America and Japan. The factors governing consumer choice are complex and car manufacturers can choose areas to compete in such as engine performance, ride, or level of luxury. Increasingly a key area of competition for mass manufacturers is in minimising the environmental effect of the car. This has arisen because of worsening air quality in cities (increased nitrogen oxides, sulphur dioxide, ozone, and carbon monoxide) and the perceived connection with asthma and other respiratory diseases. In addition there is concern about the potential greenhouse effect arising from accumulation of carbon dioxide in the upper atmosphere. Cars are a major contributor to all of these and manufacturers must act. Although catalytic converters and lead free petrol have already helped to reduce air pollution, further improvement will require lower petrol consumption. There are many ways to achieve it including evolutionary improvement of the engine, the transmission, tyres and the running gear, as well as reducing the weight of the car. All are being studied and implemented in parallel. However, the steel industry is primarily concerned with reducing weight without increasing cost and here an intense competition is developing between steel, aluminium and plastics.

SEQ 0114 JOB IOM8627-007-006 PAGE-0114 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

114

Hatfield Memorial Lectures Vol. III

In the spring of 1994 a consortium of 32 steel companies from 15 countries, including British Steel, commissioned Porsche Engineering to design an ultralight steel auto body (ULSAB).12 The objective was to achieve a significant weight reduction of the steel body in white (the unpainted structural shell of the car) of a typical midsize saloon without any increase in cost. Phase 1, a computer based design and crash test study, was completed successfully in August 1995 using a new design and manufacturing approach. Figure 9 shows a computer simulation of the body in white. It will weigh 25% less than current designs and also cost US$150 less. Furthermore, the car will be safer. Its head on crash survivability is 35 mph compared with the current government mandated 30 mph, an improvement in energy absorption of 36%. In addition, the driveability of a car will be better because its predicted torsional rigidity is 50% higher. Phase 2, the manufacture and testing of bodies in white, is under way and will be demonstrated to the public in the first quarter of 1998.

Fig. 9 Computer simulation of the ULSAB body shape.

Two advanced manufacturing methods are crucial to the assembly of the ULSAB. The first substitutes hydroformed tubes for spot welded box sections and saves weight, reduces the number of parts, lowers the cost and increases design freedom. Hydroforms are intricately bent and shaped tubes that avoid the welded flanges of box sections which add unnecessary weight, limit effective space utilisation and only provide a weak discontinuous spot welded joint. Hydroforming involves three stages. Straight welded tubes (1– 4 m long) are bent on a conventional bending machine into a preform which is then expanded hydraulically into a die to form an accurate shape with a uniform wall thickness. Holes can be punched into the part using tool inserts. Finally, the part is removed from the tool and cut to shape. Hydroformed parts are already in mass production and Fig. 10 shows a Ford Mondeo engine cradle made this way. Compared with the same component made by stamping and spot welding, the major U shaped hydroformed component highlighted is one piece not six, is 30% lighter and 40% stiffer with 50% improvement in materials utilisation, and requires three manufacturing steps not 32. The second substitutes, for a single sheet of steel, preforms made up of different grades of steel that are laser welded together then pressed into a body panel. These laser welded blanks save weight and cost, not only because they reduce waste, the number of parts and the number of dies, but also because using different grades of steel produces the right

SEQ 0115 JOB IOM8627-007-006 PAGE-0115 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

Fig. 10

115

Hydroformed engine cradle for Ford Mondeo, highlighting the major component by a white outline (courtesy TI Vari-form).

properties everywhere in a finished part. With the present technology the whole part is overengineered because large panels are pressed from single sheet whose properties are determined by the most highly stressed or corrosion prone point in the part. An example of a laser welded side frame is shown in Fig. 11.

Fig. 11 The laser welded side frame of the Cadillac Seville.

In the future, these and other new design concepts combined with a series of new high strength steels will continue to contribute to reducing petrol consumption while improving passenger safety and car driveability, but without consumers having to pay more. The competition from plastics and aluminium will be rolled back because steel based systems provide the best solution from a consumer’s viewpoint.

PACKAGING The Two Piece Beverage Can Packaging is a £320 billion industry worldwide and the European Union accounts for around £100 billion. Of the materials used in consumer packaging, metals account for 27.5%; paper and cardboard, 28.3%; plastics, 30.3%; and glass, 14.1%.13 Steel competes in the rigid packaging sector with glass, aluminium, and plastics and for some products such as food cans steel has essentially all of the market. However, in others, notably beverage

SEQ 0116 JOB IOM8627-007-006 PAGE-0116 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

116

Hatfield Memorial Lectures Vol. III

containers, competition is extremely fierce and is based on cost and transport efficiency (the weight of the can) as well as its recyclability or reuseability. This is now moving towards differentiation visible on the shelf in supermarkets and so includes decoration, texture, end design and shape. In 1992 three major European tinplate manufacturers, British Steel, Rasselstein and Hoogovens, began to design jointly a new can together with CarnaudMetalBox (now part of Crown Cork and Seal) which is a global can manufacturer. The objective was to reduce both the cost and weight of the can as well as to introduce a new, visibly different can end. Consequently, like ULSAB, the design activity involved both product performance and manufacturability. Cost and weight reduction involve reducing the thickness of the feedstock and modifying the manufacturing process. The bodies of two piece beverage cans are manufactured by a process of drawing and wall ironing (DWI) and this involves several operations. First, cups are drawn from discs cut from a continuously fed coil of steel, then subsequently are redrawn and ironed in three steps to produce a shell that is the full height of the can. The process is very fast, each bodymaking machine operating at a rate of 250 + units min–1. The lightweight can body necessitated the development of a new superclean steel, new lubricants, and design of a new DWI tool. The first phase of the project has successfully enabled the weight of the steel body of a 33 cL can to be reduced from 28 to 22 g, reducing the thickness of the strip feedstock from 0.27 to 0.22 mm. This product is now in the market and further developments to reduce the weight to 18 g are underway. To differentiate the steel can, a new can end design in steel has been developed that looks different and is both easier to open and provides evidence of tampering. This ‘Ecotop’ was originally designed by BHP and American Can, but has been developed further by the consortium. It features two small buttons; pressing the smaller button releases the internal pressure and pressing the larger button enables the drink to be poured. Both buttons stay attached to the end because a small hinge of metal remains after pressing. For this reason the metal is almost completely punctured around the periphery of each button and only a small hinge remains. As a consequence, sharp edges around the apertures are completely eliminated. A plastisol sealant is used with any one of several different designs of aperture to ensure that tampering produces an unacceptable flat drink. This product development has contributed to increased market share for steel in the UK. In mid 1995 a number of leading brands announced a move to steel cans, reversing the earlier trend toward aluminium, with the results shown in Fig. 12. Can decoration is a further differentiator and will form a key element of future can development. One approach involves the use of shrink wrap and coloured ends. Examples of shrink wrap decoration are shown in Fig. 13. These have a strong colour density, but more importantly may exploit current events for marketing purposes. For example, pictures of players from winning teams in major sporting events such as the FA Cup Final, the World Cup, or the Superbowl may be on a can in a supermarket within 48 h of the event.

SEQ 0117 JOB IOM8627-007-006 PAGE-0117 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

117

Fig. 12 Market growth in cans in the UK.

Fig. 13 Ecotop cans decorated with shrinkwrap film.

CONSTRUCTION Construction is the largest market for steel. For example, in the European Union it is estimated at £420 billion (1991 prices), equivalent to 10% of the GDP.14 Depending on the market segment, steel competes against concrete, masonry, and timber. In 1994, in the European Union the production of constructional steelwork was 6.1 Mt, of which the industrial buildings sector used 2.4 Mt and non-residential building accounted for nearly 2.1 Mt. There are many opportunities for revenue and margin enhancing product development here also. Commercial Buildings The competing construction materials for commercial and single storey industrial buildings are steel and concrete. Over the past decade competition has been based on a combination of cost and fire resistance and, within the European Union, the steel industry has been most effective in gaining market share in the UK. Here, steel costs have been relentlessly driven down to be the lowest in Europe. Furthermore, together with the Steel Construction Institute, British Steel has successfully developed a range of cost

SEQ 0118 JOB IOM8627-007-006 PAGE-0118 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

118

Hatfield Memorial Lectures Vol. III

effective insulation systems that provide protection for specific times. These have been accepted by regulatory authorities and have become part of the building codes. Consequently, a system can be selected that meets the needs of an individual building dependent on its use. Together with engineering developments that have reduced the time spent in design and construction, this work raised steel’s market share in the UK from 30% in 1981 to 62% in 1993 where it has remained stable. The basis of competition is now moving towards producing effective and low cost mechanical and electrical services. Many large buildings are highly serviced with the mechanical and electrical installation accounting for 40% of the cost of an office block and 80% of a laboratory building. This compares with 10% for the building frame, so the basis of steel’s contribution must be to offer a low cost system with services that will be flexible in use and easy to access for maintenance or modification throughout the life of the building. British Steel, supported by the Steel Construction Institute, has developed a series of new building systems that addresses these issues and which has culminated in ‘Slimdek’ (launched in May 1997), which is illustrated in Fig. 14. The design is based on an asymmetric beam that supports a profiled steel deck which incorporates a poured reinforced conctrete floor. Services may be run through the tube which can be directly accessed from the ceiling of the room beneath. From the consumer’s point of view the system offers several advantages. For example, for the same floor area, the height of the building is 10% less, reducing the cost of both the structure and the mechanicals and electricals. Furthermore, because there are no service obstructions, room partitioning can be moved easily to provide maximum flexibility or use. In addition, easy access to the services ensures cheap maintenance and uprating if the use of the building changes. Again, fire tests have shown that protection of the steelwork is not normally required for up to 1 h fire resistance and longer periods can be achieved by protecting only the bottom flange of the asymmetric beam. Between 1992 and the end of 1995, approximately 50 buildings with ‘Slimflor’, the predecessor to ‘Slimdek’, have been completed in the past 4 years, one of which, the Ionica headquarters building at St John’s Innovation Park, Cambridge, is shown in Fig. 15. From a manufacturing point of view advances include

Fig. 14 Schematic of ‘Slimdek’.

SEQ 0119 JOB IOM8627-007-006 PAGE-0119 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

119

Fig. 15 The Ionica building, Cambridge, built with ‘Slimflor’ (courtesy Timothy Soar).

development of a new rolling technique to produce the asymmetric beam as well as innovative assembly techniques that speed construction and increase site safety. In the future, competition is likely to include ease of dismantling and resuse of the building components at the end of its life. Here, steel structures already have an advantage over concrete and will be further developed with a modular construction approach. Family Housing The basis of competition in this market sector has been cost, national preference, and personal taste. Clear differences emerge between different countries. For example, in North America and Japan the material of choice for the structure is timber, whereas in Britain it is brick. Steel framed houses are constructed like wood framed houses except that galvanised rolled formed sections 1.0–8.0 mm thick are substituted for wood 2 × 4 in studs. Each house can contain between 3 and 5 tons of steel, so the market is potentially very large. Compared with wood, steel framed houses have better survival in earthquakes and are fire resistant. They are not attacked by termites, do not rot, and last longer. Most importantly, they are more acceptable to some consumers than timber because they avoid deforestation. In most markets they cost less than the wood framed equivalent. To penetrate the market, building systems have been developed that include techniques for cutting, joining, insulating, etc. with the accent on simplicity and speed of construction. Frames can be either constructed on site as they are from wood or prefabricated offsite. The key to success has been the considerable effort that has been put into educating architects, builders, inspectors, local planning authorities and house insurers on satisfactory design codes, materials and construction systems.

SEQ 0120 JOB IOM8627-007-006 PAGE-0120 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

120

Hatfield Memorial Lectures Vol. III

In countries where timber framed houses are common, steel framed housing is the fastest growing new market for steel. It is particularly successful in North America, Australia, Japan, and Sweden. For example, in North America about 80 000 homes were estimated to have been built in 1995, representing about 0.5 Mt of new business.15 The target for 1997 is 250 000 houses. Success is regional. For example, most houses built in California since the Los Angeles earthquake in 1994 have been steel framed because of their proven earthquake survivability. However, in countries where the preferred building material is brick, steel framed housing is growing much more slowly. For example, in the UK it is predicted that in 1997–98 only 1250 houses (< 1% of the annual total) will be built. However, through initiatives such as the British Steel ‘Surebuild’ system it is anticipated that the use of steel frames combined with traditional brick cladding will accelerate. An example of a house under construction using ‘Surebuild’ framing is shown in Fig. 16. The system includes computer aided design of the house and precision factory welding of the frames.

Fig. 16 A steel framed house under construction (courtesy Bett Homes).

PRODUCT DEVELOPMENT FOR NICHE MARKETS The industry is also undertaking many developments of materials systems aimed at growing individual niche markets. Two examples that provide a flavour of the approach are summarised below. Ferrolite Ferrolite is steel strip (0.1–0.4 mm thick) coated with a coloured and decorated polymer film (15–200 mm thick) that was developed by CarnaudMetalBox and is produced under licence by British Steel for niche applications. The films are polypropylene, polyethylene terephthalate (PET), and nylon. When the sheet is formed, e.g. into a can, the films remain intact and indeed enhance formability. The films have full FDA approval so applications include pet food containers, microwaveable meal trays, baby food cans, and the easy open beverage can end described above. Niche applications

SEQ 0121 JOB IOM8627-007-006 PAGE-0121 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The New World of Steel

121

such as aerosol components and strip light reflectors have also been developed. There is also potential to use ferrolite in any application that enables manufacturers to eliminate a paintline and so avoid environmental problems. Examples might include cladding sheet for the construction market or sheet material for the body in white of an automobile. Sound Deadened Sheet Sound deaded sheet has been developed by several steel companies throughout the world. It consists of two thin sheets of steel separated by a thin layer of viscoelastic polymer which can be selected to provide maximum damping at specific temperatures. The composite sheet can be readily cut and the formability characteristics are similar to conventional steels. It is finding application in the automotive industry in the engine compartment, rocker covers, dashboard panels, and oil sumps and has the potential for larger components, such as door panels between the engine and passenger compartments of the Mazda 929 and Lexus saloons are made from sound deadened steel. There is clearly an opportunity to develop non-auto markets. Already sound deadened steels are used to quieten domestic appliances in Japan and there is potential growth in North America, where noise reduction of all forms of machinery provides a competitive advantage.

CONCLUSIONS (1) Today’s steel industry is the result of a century of innovation and investment in both process and product. (2) The rich seam of technology development will continue well into the next century as will the opportunity to invest creatively worldwide. (3) In the past, success in process development was increased scale and fully integrated steelworks with a wide range of products. In the future, it will be telescoping several sequential processes into one more environmentally friendly and much less capital intensive process with output focused on a narrower range of products in regional markets. (4) In the past, success in product development was better steels. In the future, it will be new steels and engineered systems to meet consumer needs. (5) Continuous improvement in environmental performance and effectiveness of recycling and reuse will be essential to maintain the production base and the position of steel in the marketplace. (6) The future is ours to create.

SEQ 0122 JOB IOM8627-007-006 PAGE-0122 43RD LECTURE REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

122

Hatfield Memorial Lectures Vol. III REFERENCES

1. Steel Statistical Year Book: 1994; IISI, Brussels, 1995. 2. J. W. EDINGTON, M. J. CORBETT and F. G. WILSON: IMM Bull., September 1995, (1026), 7–13. 3. M. L. THOMPSON: Direct from Midrex, 1992, 17, (3), 10–11. ¨ 4. C. BOHM, A. EBERLE, L. GOULD, J. KRIECHMAIR and R. WODLINGER : Steel Times, May 1995, 176–179. 5. H. FAURE: Steel Times, December 1993, 511–512. 6. B. A. POLLOCK: Iron Steelmaker, July 1993, 25–30. 7. J. AYLEN: Ironmaking Steelmaking, 1990, 17, (2), 110–117. ¨ 8. K. WUNNENBERG and K. SCHWERDTFEGER: Iron Steelmaker, April 1995, 25–31. 9. Y. K. SHIN, T. KANG, T. REYNOLDS and L. WRIGHT: Ironmaking Steelmaking, 1995, 22, (1), 35–44. 10. G. LESSEL and M. GROBER: Cont. Cast. Suppl., May–June 1994, 12–14. 11. D. TAPSCOTT: The Internetworked Buisiness from the Digital Economy, McGraw-Hill, New York, NY, 1996. 12. E. F. WALKER AND K. LOWE: Proc. Conf. on ‘Materials for lean weight vehicles’, University of Warwick, 27–28 November 1995, The Institute of Materials, 19–28. 13. J. BANKS: MBM, January 1995, 22–28. 14. Construction Europe. European Commission, Brussels, 1994. 15. J. D. EWING: Proc. Int. Sem. ‘Steel in housing’, 25–26 January 1995, IISI, 21–26.

SEQ 0123 JOB IOM8627-008-008 PAGE-0123 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE FORTY-NINTH HATFIELD MEMORIAL LECTURE

Technology – Driving Steel Forward M. J. Pettifor At the time the lecture was given Dr. Mike Pettifor was Director, Technical of Corus CC&I. He was also Chairman of the Steel Division of the Institute of Materials, Minerals and Mining. The lecture was presented in Sheffield in December 2001.

It is argued that continuous innovation of process, product, and applications will be key to the future success of the steel industry. Innovative applications of standard steels are likely to be at least as important as the development of new products; the construction industry in particular offers business opportunities. The industry must bring its solutions to market faster and add value by exploiting its knowledge of the product. It is a great honour to present the Annual Hatfield Lecture, following in the steps of so many illustrious predecessors. The lecture was established in 1944, as a memorial to the late Dr William Hatfield, the renowned Sheffield metallurgist. This lecture is the third in a series on the iron and steel industry and materials science more generally: the others1,2 having dealt with the past and present, I shall attempt to look forward, although there is no doubt, to paraphrase Niels Bohr, that prediction is a very imprecise science, especially where it involves the future. In what will be a very personal view, based on a lifetime’s experience in the industry, with United Steel Companies, British Steel and finally Corus, I shall suggest where the foundations of a successful future for steel may lie. The emphasis will be on steel’s future, rather than on other materials or competitive threats.

DRIVERS FOR CHANGE One question that perplexed me throughout my career concerned the drivers for change: which came first, market demand or technology push? I was certain that no market survey established the requirements or specification of the first wheel or Walkman, although certainly marketing has been responsible for the rapid growth of the latter. Of the factors affecting demand, important market drivers include: ● the supply–demand balance ● entry cost

123

SEQ 0124 JOB IOM8627-008-008 PAGE-0124 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

124 ● ● ● ●

Hatfield Memorial Lectures Vol. III

legislation (particularly environmental) market needs business globalisation profit.

A breakdown of world metal consumption illustrates the pre-eminent position of steel. Of the 900 Mt market, carbon steel makes up 810 Mt, stainless steel 31 Mt, primary aluminium 29 Mt, then down through copper, zinc, nickel, and magnesium to titanium at 0䡠15 Mt. Figure 1 shows demand for steel products to 2000 and a forecast to 2005; overall demand increases steadily but with large geographical differences – marked growth in China and developing Asia, reductions in Central and Eastern Europe, and relative stagnation in Western Europe, the USA, and Japan. Production information shows a very similar trend with strong growth occurring only in the Asia–Pacific regions.

Fig. 1 World steel demand 1976–2005

These data illustrate the ongoing issue of surplus capacity from which steel has suffered for some years now. Overall capacity in 2000 amounted to 130 Mt, only China being without a surplus (Fig. 2); Western Europe and the former USSR have significant reductions to make. The situation in the UK dates back to the 1970s when a steel requirement of some 35 Mt was being predicted; this was followed by the rapid decline in the 1980s and then growth to the present position of 15–17 Mt (Fig. 3). The general decline in UK consuming markets is well known, but this is not the case in all sectors. Figure 4 shows a significant decline in the production of commercial vehicles, but the reduction in volumes of cars from UK manufacturers is offset by recent growth as a result of the transplanted plants and, perhaps a little surprisingly, the continuous steady growth of construction – a topic returned to below.

SEQ 0125 JOB IOM8627-008-008 PAGE-0125 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 125

Fig. 2 World steel production and surplus capacity: 2000.

Fig. 3 UK steel production 1955–2000.

Fig. 4 Relative production volumes of metal using industries in the UK.

SEQ 0126 JOB IOM8627-008-008 PAGE-0126 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

126

Hatfield Memorial Lectures Vol. III

Fig. 5 Growth strategies for commodity businesses.

While steel is usually regarded as a commodity product, growth in commodity products – steel, soft drinks, book retailing, music (Fig. 5) – is certainly possible, given the correct marketing strategy and business model: ‘There are no commodity products, just tired salesmen.’ It is worth recalling that while the top five automotive companies account for 70% of the global market and the top five for iron ore 90%, the top five steel producers represent only 15%. A snapshot of crude steel production by company in 1999 ranged from USX at 11䡠3 Mt to Posco at 26䡠5 Mt and included such examples of consolidation within in the industry as Corus and Thyssen-Krupp, and more will certainly follow. There is a new player being formed, a combination of Usinor, Arbed, and Aceralia, which will have combined sales of c 30bn, production of 44 Mt, and employ 110 000 people. The preeminent position this company would occupy in the ranks of steel producers is evident: certainly, it would provide an opportunity to establish a dominant position commercially and pursue aggressively business rationalisation, a route that it is familiar to most of us. To conclude, despite large surplus capacities and stagnant demand (other than in China and South East Asia), growth is possible in mature markets if the correct strategies are adopted. Further concentration of the industry is inevitable.

DRIVING STEEL FORWARD Process Innovation Over the past 150 years, processes have been invented, continuously developed, and reached maturity until overtaken by new processes (Fig. 6), the electric arc furnace and basic oxygen processes reaching maturity perhaps at the same time and now coming under competition from new processes and combinations that have still to achieve maturity but are certain to bite further into the market. First, however, some thoughts on continuous improvement. In the classical blast furnace process, coke and sinter are fed into the top of the furnace, hot air is blown in through the tuyeres, and iron exits from the tap hole. Significant improvements have been made in blast furnace performance over the years, and specifically in injecting materials into the furnace. The improvement in productivity achieved

SEQ 0127 JOB IOM8627-008-008 PAGE-0127 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 127

Fig. 6 Process innovation in steelmaking.

Fig. 7 Fuel rates and blast furness productivity 1980–2000.

and the increases in coal injection rate are shown in Fig. 7. Coal injection has been carried out for many years. The benefits – reduced fuel costs, improved productivity and stability of operation – are well known, but the high cost of pulverised fuel systems for preparing the coal made installation prohibitively expensive in some cases, until an innovative engineer and metallurgist at Scunthorpe learnt that a firm in Doncaster had developed a process for injecting coarse coal into boilers. Soon contact was made and the trials began on injecting material into single tuyeres of the blast furnace, soon extending to full furnace operation. Various materials can now be injected including coal and plastics (Fig. 8). The future, though, is even more interesting as far as injection technology is concerned. Perhaps the most significant innovation not been proceeded with to date is the simultaneous injection of granular coal, fine iron ore, and oxygen through the blast furnace tuyeres, by which it is believed to be possible to replace at least 50%

SEQ 0128 JOB IOM8627-008-008 PAGE-0128 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

128

Hatfield Memorial Lectures Vol. III

Fig. 8 Stills from videos of plastic and coal injection into blast furnace.

Fig. 9 World steel production by process route projected to 2010.

of the coke requirement and a considerable portion of the ore input from the sinter plant. The importance of this lies in the high capital cost of integrated plants. If it were feasible to inject such large amounts of coal and ore and avoid some of the costs associated with sinterand cokemaking the process would have a major impact on steelmaking economics. Looking at steel production by process route over the past 150 years, the pronounced growth of both oxygen and electric arc steelmaking in recent decades is clear (Fig. 9). New processes are coming to the fore and are predicted to increase in proportion to integrated steelmakers. I believe that this is one area where we have to progress much faster than in the past. These new processes, perhaps not in Europe, but in Asia and other areas of growing demand, will play a major role in the steel industry of the future. When I joined the industry, steel was produced in open hearth furnaces and by the Ajax process, in which oxygen was injected into the open hearth furnace: we made a lot

SEQ 0247 JOB IOM8627-008-009 PAGE-0129 123-142 49 PETTI REVISED 20DEC04 AT 21:42 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 129

Fig. 10

Secondary steelmaking complexity: production route for steel for transport of wet acid gases from the North Sea.

Fig. 11 Prototype direct strip caster design.

of steel, and burnt the furnace down occasionally! Ingots were cast, but soon we moved to continuous casting and realised that to provide the caster with steel that was in specification, at temperature, at the right time, the introduction of ladle refining was required. Market demands for purer and purer steels led to more complex process routes being designed to manufacture them. Figure 10 shows an actual process route used to produce steel for transmission of wet acid gases from the North Sea: tapping, gas stirring, reladling to remove any slag, reheating, perhaps some dephosphorisation, powder injection to moderately reduce sulphur and modify the inclusions, followed by vacuum

SEQ 0248 JOB IOM8627-008-009 PAGE-0130 123-142 49 PETTI REVISED 20DEC04 AT 21:42 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

130

Hatfield Memorial Lectures Vol. III

Fig. 12 Production direct strip caster layout.

Fig. 13

Cost comparison of single integrated plant and two minimills, both producing 4 Mt/year of steel.

degassing and casting. It is absolutely wonderful for the metallurgist but unfortunately; it can take about three hours to work through the cycle and can effectively destroy the profit that might be made on such projects. Looking at the process routes to various products, the electric arc–ladle furnace–caster– rolling mill route is compact and simple, or has the potential to be. Relative to minimills, integrated plants always have to carry high capital baggage – coal and ore blending, sintermaking, cokemaking, ironmaking and steelmaking facilities – and it is not until the secondary steelmaking stage is reached that the processes become equivalent. Even then, in most plants occupying ground traditionally held by big steel, the continuous casting, slab reheating, and rolling processes are discrete and in different parts of the works. These high cost routes provided a major driver for innovation to reduce capital and operating costs and to resize to plants capable of working in small regional markets. This was achieved by collapsing the process route, for example by the continuous casting of shapes as close in dimension to the final product as possible, i.e. near net shape casting, to avoid primary rolling costs. A good example of this is beam blanks, a well known

SEQ 0131 JOB IOM8627-008-008 PAGE-0131 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 131 technology available to all. Another example, thin slab casting, can permit the linking of the casting machine and rolling mill. Long cast slabs are processed through equalising furnaces, making use of the energy contained within the solidified strand, through roughing and finishing mills and cooling beds, before coiling, often continuously as far as is practicable. The approach is equally applicable to hot strip, bar, and shape products. Figure 11 should look familiar. Shown by Frank Fitzgerald 1990 and Jeff Edington in 1995, it illustrates the strip casting process under development at the Grangetown Laboratories on Teeside. Figure 12 shows how the dream has become reality with the introduction of some inline reduction and cooling and a double coiling system. In fact such a plant was described at a recent SMEA lecture3 by Gerald Hochenblicher of VAI. EuroStrip, in which VAI is a partner, claims to be casting carbon steels and stainless successfully at speeds up to 150 m min–1. Thicknesses of 1䡠5–3䡠5 mm are scheduled for production in plants of capacity up to 0䡠5 Mt/year. These developments have led to significant cost reductions in the capital associated with steel products production and Fig. 13 shows a capital cost comparison between a single integrated works scheduled to produce 4 Mt/year and that associated with two minimill plants also producing 4 Mt but using 25% briquetted reduced iron (BRI). Avoidance of the cost of BRI leads to the inevitable conclusion that further significant reductions in capital are possible. However, the product must be fit for the market place, and while minimills certainly have had major successes and rapidly replaced traditional steelmaking in the production of many commercial products, they still have some way to go in terms of large structural sections, heavy plates, special wire rod, and strip grades. Some of these may always be the province of the integrated manufacturers. In summary, the primary process has survived so far by innovating and reacting to competitive threats. The complexity of the process routes in high productivity, multiproduct plants severely limits their potential to operate at minimum cost. I do not believe further new integrated plants will ever be built in the developed countries. The most likely economic solution is for a policy of ‘make do and mend’ as the industry is rationalised further, while new investment will be concentrated in many compact plants. These may come to play a part in steelmaking in Europe, but certainly will continue to do so with a vengeance in the Far East. Product Innovations Here I would like to describe some examples in heavy plate, heavy sections and wire rod. An important driver in the post-war effort to produce tough steels was the experience with the Liberty Ships built in the USA during the Second World War, which graphically illustrated the need to match the quality of the product, the process route, and design considerations. The Liberty Ships were designed for riveting. The steel was brittle yet the ships were welded together. The combination of very high operating stresses, workmanship that left much to be desired and unfamiliarity with the welding process resulted in ships breaking apart, often before they were launched.

SEQ 0132 JOB IOM8627-008-008 PAGE-0132 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

132

Hatfield Memorial Lectures Vol. III

Fig. 14

Microstructure control of rolled products by thermo-mechanical controlled processing.

Fig. 15 Mechanical property benefits of thermomechanical controlled rolling.

Much progress has since been made in the production of strong, tough, weldable steels through control of chemistry and microstructure and Fig. 14 illustrates the opportunities for controlling the structure by thermomechanical controlled processing (TMCP): reheating the steel slabs to various temperatures, rolling in the recrystallisation range, holding until the temperature reaches the range where no recrystallisation occurs, and rolling further to give a deformed austenite microstructure, which can then be allowed to cool normally in the ‘classic’ hot rolling (HR) process route, transforming to fine grained ferrite plus pearlite. Accelerating the cooling rate in water treatment plants (ACC) gives fine ferrite plus bainite, or in the case of direct quenching (DQ) ferrite plus martensite, which can be tempered to generate the required properties. These developments led to significant improvements in both strength and toughness. Figure 15 illustrates the

SEQ 0133 JOB IOM8627-008-008 PAGE-0133 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 133 implications as far as strength is concerned; note the marked reduction in carbon equivalent value that can be achieved in developing plate of the same strength using thermomechanical processing as opposed to heat treatment. There is little doubt that without development of these processes, exploration, extraction, and transmission of gas from the North Sea would have been considerably delayed.

Fig. 16 Some products from the large structural mill at Teeside.

Figure 16 shows some of the vast range of products from the large structural mill at Teeside, from the small column to the lower left beam which weighs >1 t m–1 to the beam on the right which is actually a metre in depth. In this mill the cast slab is vertically rolled, and then edge rolled with a combination of horizontal rolling to get the final shape. The prime target of section mills has always been to obtain correct product shape with the maximum yield. Gradually this requirement is being integrated with the metallurgical requirements, a process in which mathematical modelling is playing an increasingly important role. Linking these models with those related to the development of the microstructure of the steel proves extremely exciting. As the bar is deformed in the rolling process, it is possible to predict how the initial microstructure formed during reheating changes during dynamic recrystallisation in the roll gap, static recrystallisation, and grain growth. The industry is now beginning to bring together the two technologies and by putting a finite element mesh on the models of the slab rolling process discussed above (Fig. 17), it is possible to predict the distributions of temperature and deformation throughout the rolling process, produce plots of the equivalent plastic strain across the section, and link this to the microstructural developments that occur – allowing development of new products much faster than in the past.

SEQ 0134 JOB IOM8627-008-008 PAGE-0134 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

134

Hatfield Memorial Lectures Vol. III

Fig. 17 Prediction of temperature and deformation during rolling.

SEQ 0249 JOB IOM8627-008-009 PAGE-0135 123-142 49 PETTI REVISED 20DEC04 AT 21:42 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 135

Fig. 18 World’s first asymmetric beam rolled in routine production.

Fig. 19 Development of strength in steel wires for suspension bridge construction.

One example of this is the asymmetric beam: it is difficult to roll sections that are asymmetric for a number of reasons and I believe that the beam shown in Fig. 18 is the first asymmetric section rolled in routine production operations. However, as well as allowing rapid development of completely new products, model driven development also speeds reaction to the marketplace, making changes to the dimensions of existing products possible very quickly. Turning to wire and rod products, understanding the relationship between rod and wire properties and the influence of processing on these has led again to the development of sophisticated mathematical models to simulate a single slug of wire passing through guides and dies in the wire drawing process; accurate prediction of the high strains achieved at the throat of the die is particularly important. This has permitted development of total wire drawing models that make it possible to understand what goes on in the wire drawing process far more accurately, again facilitating development of new products.

SEQ 0136 JOB IOM8627-008-008 PAGE-0136 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

136 Hatfield Memorial Lectures Vol. III A good example of the impact of continuous development of wire strength is that of suspension bridges, which are an important market for high carbon steel wire ropes; 2 or 3 mm diameter hard drawn high carbon wires are formed into ropes and subsequently into cables to form part of the suspension rig itself. The development of wire strength has been fundamental to the increased span of suspension bridges (Fig. 19), from the Brooklyn Bridge in the 1980s with wires of 1100 MPa strength to the Golden Gate Bridge (1280 m span, wire strength 1520 MPa), and the Humber Bridge (1410 m, 1560 MPa). A significant hike then occurred to the Akashi Kaikyo Bridge (1991 m, 1770 MPa); this phenomenal increase in strength certainly made a suspension bridge more attractive to build by changing the design from a four rope to a two rope bridge with consequent reductions in costs, material weights, and foundations. The Strait of Messina Bridge (3300 m span), not yet a firm project, will join Sicily to mainland Italy and will contain 166,000 t of steel in the cables. In conclusion, the benefits of mathematical modelling of the production and metallurgical processes cannot be overestimated. Detailed knowledge of the market will allow these models to be applied to satisfy the market needs. Focused, systematic innovation will enable faster development of better products and enable far faster inroads into the marketplace. It is possible to predict with confidence that thermomechanical processing will grow in importance in all process sectors. Innovation in Application In 1981, a British Steel study into the costs of fire protection in buildings concluded that the cost of insulating fibre and insulating boards for protecting structural steelwork in multirise buildings amounted to 31% of the cost of the building, more than the cost of the steel; this was certainly the major reason why steel could make no inroad into the market for concrete. The reaction was to undertake a series of tests at the Fire Research Centre at Cardington in which fires were carefully set in, for example, a four storey building, fully instrumented to obtain the maximum information on the rate of heat transfer to the steel and the effect on stress and strain loading. This work led to very significant reductions in the quantity and cost of fire protection required and very significant changes in the designs codes for steel, bringing advantages for steel over concrete. Adjusting the 1981 cost of structural steelwork of £985/t to its equivalent 2000 value of £2400/t, Fig. 20 shows that the total cost has been reduced by 55%, fire protection costs by 64%, and the steel cost by 42% in real terms. The beneficial effects in the UK marketplace have been marked, with steel’s market share rising from 33 to 67% between 1980 and 2000, while concrete declined from 52 to 20%. However, there is much to be done to achieve similar levels of penetration in mainland Europe, where steel usage ranges from 30% in Spain and The Netherlands to as low as 10% in Italy.

SEQ 0137 JOB IOM8627-008-008 PAGE-0137 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 137

Fig. 20 Reduction in cost of multistorey buildings achieved following changes in steel design codes.

Fig. 21

Corus Bi-steel: applications of concrete filled sections include blast protection and to increase resistance to seismic shock.

Another application of interest is Corus Bi-Steel, which consists of two plates connected by friction welded bars (Fig. 21), filled with concrete to improve the properties. Bi-Steel has excellent impact strength and behaves in a semiductile manner. One major application for this product is in blast protection structures; another is in seismic applications, which has received much interest following the earthquakes in Chiba in Japan and North Ridge in Los Angeles.

SEQ 0298 JOB IOM8627-008-009 PAGE-0138 123-142 49 PETTI REVISED 21DEC04 AT 02:50 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

138

Hatfield Memorial Lectures Vol. III

Fig. 22

Location of connection cracking in joints of steel structures following earthquake.

These earthquakes resulted in a major, international work programme covering design, materials, fabrication and in-service stress patterns during seismic events. Figure 22 depicts schematically how failures were generally concentrated in the junction areas in the corners of the steel structures. Any fractures propagated down the weld metal and subsequently in the base material parallel to the weld, with some cracking through the material on occasions. However, the major concern was lamellar tearing, which always occurred in the base metal, always outside the visible heat affected zone, and generally parallel to the weld fusion boundaries. These issues led to a major development programme examining the effect of steel chemistry, cleanness, segregation, and rolling on toughness and susceptibility to lamellar tearing. The output of this work was a series of beams and columns with improved properties, and quite a large rig was constructed to test these materials, 19 m in width and 9 m high. The beam–column connections were welded with very high levels of restraint and hydraulically jacked to preset levels, examined, then rejacked to failure: no connection failures occurred, failure being by buckling of the section itself, a very satisfactory outcome. Steels having significantly better ability to withstand the rigours of seismic events have been developed, but, as with the Liberty Ships, the real requirement is to get the welding processes correct – the weld metals in the failed structures did not possess adequate toughness, design of connections left something to be desired, and workmanship was a key area responsible for the failures. Perhaps the most important conclusion as far as steel is concerned in respect to both North Ridge and Chiba is that in neither of those failures was there any loss of life in a steel framed building. Laser cutting and welding of thin gauge materials, for example in light goods and automotive applications, has been common practice for some time. More recently, ship builders and earth moving equipment manufacturers have become interested in laser processing for heavier gauge material. Corus has constructed a heavy product laser processing facility with a 25 kW CO2 laser bank, transmitted by optical fibres along a

SEQ 0139 JOB IOM8627-008-008 PAGE-0139 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 139

Fig. 23 Heat affected zones of submerged arc (top) and laser (below) welded joints.

gantry to a laser head, which is manipulated by a five axis robot. The objective of this work was to develop expertise in laser cutting and welding of heavy gauge products and specifically to ensure that products were available that would meet any new requirements/demands coming from the market place, to build a competitive edge and to develop better, more exciting products. Submerged arc and laser weld cross-sections are compared in Fig. 23. The submerged arc weld has a large fusion zone with the heat affected zone extending beyond the extremities of the photograph, whereas the laser weld, which uses no filler material, fuses the two parts together with almost no heat affected zone. This is reflected in the two main advantages of laser processing: minimal or no distortion and reduced cost owing to the lack of expenditure on the welding electrodes. The opportunities examined included butt welding of wide plates or bespoke sections, stake welding to form sections, T joints, perhaps of different materials, and cladding of high performance materials onto low cost base material. Although higher

SEQ 0140 JOB IOM8627-008-008 PAGE-0140 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

140

Hatfield Memorial Lectures Vol. III

capital costs are incurred compared with the conventional routes, laser welding has the capability to make products that are impossible by other processes. The final example concerns knowledge based construction. Egan4 characterised construction industry performance in the UK as one of fragmentation and low margins, inefficiency, waste, and cost and time overruns being common with little learning and innovation. This all leads, as might be expected, to pronounced client dissatisfaction. Egan pointed out the potential of information and communications technology as a key enabler in the necessary improvement of the construction industry, to bring together design of buildings and components: New technologies have proved very useful in the design of buildings and their components, and in the exchange of design information throughout the construction team. There are enormous benefits to be gained, in terms of eliminating waste and rework for example, from using modern CAD technology to prototype buildings and by rapidly exchanging information on design changes.4

This led Corus to launch the Knowledge Based Construction project, the vision of which was to enable an architect, client, and consulting engineers to design and modify proposed constructions very fast, test and cost concepts rapidly, construct buildings on screen, in two and three dimensions, and to support the designs with product libraries which describe the characteristics, availability, and costs of all the products available. To be able to redesign structures rapidly, for example to reposition lift shafts, where the technology promises the capability to allow automatic redesign of the structure to cope with the basic change in design via the use of ‘intelligent objects’. And finally to provide a platform where all the associated data and knowledge could be shared in real time between those involved. A typical construction sequence would begin with the boundaries of the site being defined with security fencing; piling being driven for retaining walls, containment, or for basement structures; bearing piles being introduced as necessary; perhaps for high rise buildings, the concrete pad; the structure secured to foundations with anchor bolts, structural steel work installed; floor decking being input to the asymmetric beam mentioned above; reinforcing materials, rebar, remesh, the concrete being laid. The objective is to get the secondary steelwork and the roofing in position as rapidly as possible to ensure the building is watertight as early as possible in the process, for rapid completion. All sorts of options are available for wall materials, decorative and roofing materials – different colours, different textures for the claddings. The system that the project envisaged would allow a building to be designed on a laptop computer between people talking, then converted into a two-dimensional drawing for people in the field to work on. It is absolutely vital that this work is continued to maintain and extend the dominant position of steel in construction.

SEQ 0141 JOB IOM8627-008-008 PAGE-0141 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Technology – Driving Steel Forward 141 CONCLUSIONS In drawing overall conclusions, it is important to emphasise again the difficulty of predicting the future: the one thing I soon learnt about forecasts was that they were never right. The future may be possible to predict if you are talking about tomorrow but accurate long term predictions are impossible. Nevertheless, I would suggest that the successful steelmaker of the future would have to give serious consideration to the following points: ● continuous innovation – of process, product, and application technology – will be fundamental to the survival of the steel industry ● construction continues to provide the major opportunities for both UK and world steel ● innovative applications for standard products will show significant benefits, compared with the past emphasis on development of new products ● solutions must be brought to market faster to maximise early benefits ● knowledge is the source of competitive edge: we have to move from simply selling bars of steel to encapsulating our knowledge of the product and its effective use and ensuring we achieve adequate value from the market place. However in striving to achieve success in new applications and products, we must remember to keep hold of existing processes. Remember that the past actually did exist: the future exists only as a place we are going to. The secret of success is to remember and learn from the past, live the present, and create the future.

ACKNOWLEDGEMENTS The assistance of friends and former colleagues at Corus in the preparation of the lecture is gratefully acknowledged, in particular Colin Honess of Swinden Technology Centre. This paper is based on the 49th Hatfield Memorial Lecture given in Sheffield on 4 December 2001.

REFERENCES 1. R. BOOM: 47th Hatfield Memorial Lecture: ‘Iron, the hidden element: the role of iron and steel in the twentieth century’, Steel World, 2000, 5, (1), 88–96. 2. C. J. HUMPHREYS: 48th Hatfield Memorial Lecture: ‘From Hatfield to high-technology: designing materials for the twenty-first century’, Sheffield, UK, December 2000.

SEQ 0142 JOB IOM8627-008-008 PAGE-0142 123-142 49 PETTI REVISED 19DEC04 AT 21:00 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

142

Hatfield Memorial Lectures Vol. III

3. G. HOCHENBICHLER: ‘EuroStrip – strip casting now a reality’, SMEA Lecture, Sheffield, UK, December 2001, Sheffield Metallurgical and Engineering Association. 4. J. EGAN: ‘Rethinking construction’, Department for Transport, Environment and the Regions, London, UK, 1998; available at www.dti.gov.uk/construction/ rethink/report/index.htm.

SEQ 0143 JOB IOM8627-009-007 PAGE-0143 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE FIFTIETH HATFIELD MEMORIAL LECTURE

Metallurgical Modelling of Thermomechanical Processing C. M. Sellars FREng, FNAE, FIMMM At the time the lecture was given Professor Mike Sellars was Senior Research Fellow in the Institute for Microstructural and Mechanical Process Engineering: The University of Sheffield. The lecture was delivered in the Octagon Centre at the University on 3 December 2002.

I am deeply honoured to have been invited to deliver this 50th Hatfield Memorial Lecture. As a Sheffield person, I am particularly conscious of the significance of the Hatfield Lectures in the metallurgical calendar. I attended my first lecture in 1959. This, the 12th lecture, was delivered by Academician Kurjumov on the subject ‘Phenomena Occurring in the Quenching and Tempering of Steels’. Since then, I have attended 32 of the intervening 37 lectures and tonight makes a very special 33rd occasion. The subject of tonight’s lecture is one which would have been unimaginable to William Herbert Hatfield, because at the time of his death in 1943, powerful digital computers were things of science fiction and ‘A Brave New World‘. However, from my reading of his achievements and attitudes, I feel certain that he would have been one of the first to recognise the potential contribution of computer modelling of the metallurgical effects of thermomechanical processing variables to the better understanding of their consequences on the properties of the products. He would then have made the maximum commercial benefit from this understanding.

INTRODUCTION The aims of this lecture are to consider: 1) why computer modelling of thermomechanical processing is carried out, 2) how the modelling process works, 3) where modelling is today, and, somewhat rashly, where it might be going in the future. However, first let me ensure that we all have the same understanding of the meaning of the term ‘thermomechanical processing’. To me, the term is a generic one for all forms of hot working of metals, but with the implication that these processes involve much more than simply heating the material and beating it into the required shape. The old definition

143

SEQ 0144 JOB IOM8627-009-007 PAGE-0144 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

144

Hatfield Memorial Lectures Vol. III

Fig. 1

Schematic representation of the field of variables in thermomechanical processing operations.

of ‘hot working’ as ‘forming metals above their recrystallisation temperature’ is totally inadequate to recognise the many variables that influence the properties of thermomechanically processed products. These may be ‘semi-finished’ or ‘finished’ in the sense that they do or do not have further processing, other than machining, to end up as one of the artefacts on which modern life depends. The term thermomechanical processing thus encompasses processing of all types of metallic materials to a wide range of shapes in various sizes or gauges, involving plastic deformation at different speeds and temperatures. It is therefore multi-dimensional in the numbers of variables involved. If, however, for simplicity we represent the full range of variables in two-dimensions as the grey area in Fig. 1, this can be considered as the generic field for thermomechanical processing. Within this field, each industrial process will occupy a much smaller area but will have variables in common with other industrial processes, with some ranges of variables unique to the particular process. For example, all the illustrated processes, except aluminium extrusion, involve deforming the material in a number of hits or passes separated by intervals of time, in which temperature changes and important changes in the internal microstructure of the material can take place. Sophisticated mechanical working equipment with complex control systems is required for these processes. The extreme example is probably for the hot rolling of steel strip, for which the mill has a series of roughing stands followed by a series of finishing stands, which are ‘in tandem’ so the strip is passing through all the stands at the same time and its speed becomes faster and faster as the strip becomes thinner in each pass. Finally, it undergoes rapid cooling before being formed into a coil. Each of these steps requires close control to give the strip the ‘attributes’ such as gauge, shape and surface finish and the ‘properties’ such as strength, ductility and formability, required by the customer. These must all be achieved at minimum cost.

SEQ 0145 JOB IOM8627-009-007 PAGE-0145 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 145

Fig. 2

Frequency distribution curves showing the effects of reducing the standard deviation, σ, by 1⁄2 and by 1⁄4 about a constant mean value.

AIMS OF MODELLING The main aim of computer modelling is to contribute to making better products cheaper and/or faster. There are many ways in which better may be defined, but for the purposes of this lecture, I consider it to mean improved consistency of the attributes and properties of the product. If one measures any one of these, there will be a certain scatter of values about the mean value. This is best represented by a frequency distribution curve, as illustrated in Fig. 2. Considering the lowest curve first, one can see that the values of the measured property of individual samples of the product most frequently fall close to the mean value, but with decreasing frequency some sample properties deviate significantly from the mean. The width of the distribution is measured by the standard deviation, σ, and improving consistency is defined as reducing σ. The narrower curves result when σ is halved and then halved again. Clearly, more results fall closer to the mean, and the extreme deviations are reduced. In fact, if one takes a range of ± 3σ about the mean as acceptable for the product, then only about one sample per thousand will fall outside the acceptable range. This might be considered a reasonable process window for supplier rejects, but in design of components, which may be used in their millions, it could lead to an unacceptably high failure rate. To obtain only one failure per million means multiplying the process window by a factor of 1.4 and for safety critical applications, where one can risk only one failure per billion, a factor of 2 is applied, i.e. the design limits are ± 6 σ about the mean. Thus from the application point of view, the benefit of improved consistency is seen directly in improved efficiency of design, so there will always be customer pressure for continued improvement. The benefit to the producer arises first from the need to stay ahead of the competition to stay in business! Also, for many products, it is not the mean but the minimum value of the property that is specified. The significance of this is illustrated in Fig. 3, in which it can be seen that reducing σ leads to a reduced target value for the mean, which should reduce alloying or processing costs. In

SEQ 0146 JOB IOM8627-009-007 PAGE-0146 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

146

Hatfield Memorial Lectures Vol. III

Fig. 3

Frequency distribution curves showing the effects of reducing the standard deviation when the minimum value of a property is specified.

Fig. 4

Schematic diagram of the cost benefits arising from reduced values of standard deviation, σ, achieved by different control strategies.

this figure the specified minimum would be the target if σ were zero. However, even if such identical properties could be achieved, it would never be known because no property can be measured with absolute precision. The cost benefit for the producer in improving consistency, i.e. in reducing σ, is shown schematically in Fig. 4. Traditionally, reduced σ has been achieved by the application of feed-back control systems in which sensors on the processing plant send signals to actuators, which adjust the plant settings when an error signal is received. Below some value of σ the cost benefit decreases as the frequency with which the target is achieved decreases. Advances in sensor and actuator design have decreased this limiting value of σ, but still an error must be detected before corrective action can be taken. The potential of computer modelling is that it can predict when action should be taken to prevent an error occurring, i.e. it enables feed-forward control to be applied. In principle, with improved

SEQ 0147 JOB IOM8627-009-007 PAGE-0147 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 147

Fig. 5

Effect of recrystallisation between passes on the flow stress of a microalloyed steel during deformation at 900°C and a strain rate of 5 s–1 (after (1)).

sophistication of models, consistency at a level of the precision of the test method for measuring the property may be achievable. This provides clear justification for the investment involved in developing models. The subject of this lecture is ‘metallurgical modelling’, so I now briefly consider why it is necessary to include metallurgy in the modelling. The simplest example relates to consistency of shape, size or gauge of the product, which is constrained by the fact that the working equipment, e.g. a mill stand and rolls, stretches elastically under the loads imposed in deforming the stock. These loads depend on the flow stress (strength) of the material under the specific conditions of temperature and strain rate (speed) of a particular pass. For a specific temperature and strain rate, a continuous deformation to a high strain causes the flow stress to increase from its initial value to a higher steady state value, as shown by the continuous curve in Fig. 5. In multi-pass processing, if no metallurgical changes take place in the time interval between passes, the work hardening introduced by the deformation in each pass is retained and the flow stress for the sequence of reductions (15%, 15%, 15%, 15%, 10% and 5% in this figure) follows the continuous curve. However, because of the high temperatures of processing, it is possible for the material to recrystallise fully between each pass. The recrystallisation process completely removes the work hardening, so each pass has a flow stress curve similar to the first pass. This results in much lower flow stresses in the later passes, so it is important to know whether or not recrystallisation takes place to model the process accurately. Furthermore, when recrystallisation does not take place, the original grain structure is simply flattened and elongated, whereas when recrystallisation occurs repeatedly, the grain structure is progressively refined, so, as illustrated schematically in the figure, the microstructure and hence the properties of the product differ significantly.

SEQ 0148 JOB IOM8627-009-007 PAGE-0148 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

148

Hatfield Memorial Lectures Vol. III

The most extreme example of metallurgical effects that I know occurred in finish plate rolling a niobium microalloyed steel compared with rolling a simple carbon-manganese steel under the same conditions (1). The difference in composition between these steels is essentially the presence of 0.03 wt%Nb in the microalloyed steel, yet the difference in mean flow stress calculated from the rolling loads in each pass is dramatic, as shown in Fig. 6. The crosses at the bottom show the results calculated for the C-Mn steel and the filled circles at the top show the observed values for the Nb microalloyed steel. The differences arise from several contributions: a smaller grain size generated by recrystallisation during roughing rolling of the niobium steel, and some hardening by the small amount of niobium in solution, caused by the large atomic size difference between the niobium atoms and the iron atoms of the steel matrix. The major differences arise from retained work hardening in the niobium steel because of retarded recrystallisation, as discussed above, and after pass 6 from strengthening by the formation of very fine niobium carbonitride, Nb(C,N), precipitates as the temperature falls between passes. The final

Fig. 6

Mean flow stress as a function of finishing pass number in industrial plate rolling of X60 microalloyed steel (2).

SEQ 0250 JOB IOM8627-009-008 PAGE-0149 143-164 50 SELLA REVISED 20DEC04 AT 21:46 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 149

Fig. 7 Schematic diagram showing the function of modelling.

difference of about a factor of 3 in flow stress has serious consequences and requires careful metallurgical modelling.

MODELLING METHODOLOGY Mathematical modelling is simply a method of turning information about a set of input variables, pi, into information about a set of output variables, ai, as illustrated in Fig. 7. The criteria to define the type of modelling required depend upon the action to be taken as a result of predicting the output variables. This has resulted in a range of modelling methodologies, which in the limits can be purely data-based or purely knowledge-based. For control purposes the models have been essentially data-based, using results from production or trials on the specific process, for which control actions are required. Traditionally, multiple linear regression analysis was used to obtain a series of empirical equations:

a1 = b0 + b1p1 + b2p2 + b3p3 + ........ (1)

a2 = c0 + c1p1 + c2p2 + c3p3 + .......

etc.

More recently, the application of artificial neural network analysis has removed the need for the relationships to be linear and so has improved the precision of the relationships between ai and pi. Artificial neural network modelling has been developed over the last twenty years (3). It is based on the biological neural networks in the brain, in which a complex set of some 1011 neurons and their interconnections each carry out some simple function on input data and transmit the modified data to other neurons. In this way the complex functions, including memory, are built up. Artificial neural networks are much simpler, but can be trained using certain learning rules to obtain close correlations between the inputs and outputs of a set of training data. The training is a computer intensive process, but once trained, the networks can operate extremely rapidly. However, the way in which they arrive at the correlations between input and output variables is even more mysterious than multiple regression analysis. Such models can therefore be considered to be ‘black box’. They require little knowledge input and they give no additional knowledge output.

SEQ 0251 JOB IOM8627-009-008 PAGE-0150 143-164 50 SELLA REVISED 20DEC04 AT 21:46 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

150

Hatfield Memorial Lectures Vol. III

Nevertheless, they provide rapid, accurate interpolation, which is all that is required for control of a specific process. For process development and optimisation, it is necessary to have understanding of the mechanisms involved. This leads to knowledge-based modelling, which gives a series of physically based equations

a1=f1(p1,p2,p3, ............) (2)

a2=f2(p1,p2,p3, ............) etc. where the functions fi may be complex. This type of modelling may therefore be computer intensive and so not suitable for control purposes. It may also require long development times for research to achieve the knowledge-base. Such models can be considered as ‘white box’ in the sense that each step in the model is transparent in terms of its physical basis. They add to understanding of the effects of process variables and are capable of extrapolation outside the range of tuning data. Used off-line, this type of modelling can predict optimum combinations of process variables and can guide investment in new or enhanced process equipment. As discussed later, the benefits of each type of modelling can be combined using a hybrid modelling methodology. Whatever system of modelling is adopted, the overall model is generally divided into a system of interconnected sub-models, e.g. Fig. 8. In this figure the sub-model that computes the temperature as a function of time during processing is placed at the top because of the major influence of temperature on the kinetics of metallurgical changes at hot working temperatures. This sub-model requires input of heat transfer equations for heat loss to the environment or the working tools, and inputs of the time intervals, t, involved in each process step and of the plastic work done, W, which generates heat in the workpiece. The latter inputs come from the mechanics sub-model, which considers the geometry of the process and has inputs of dimensions and speeds. This model may involve simple analytical equations, or complex finite element analysis, but in order to calculate the work done and the forces involved it also needs an input of the flow stress of the material, σ, under the appropriate conditions. This is provided by a deformation model, which computes the flow stress from constitutive equations involving the strain, . ε, the strain rate, ε, temperature, T, and (as seen earlier) a description of the microstructure at each stage of processing, S, which comes from the structure sub-model. This essential information flow between sub-models is indicated in the figure by the arrowed lines. The microstructure at the end of deformation and cooling provides one of the inputs into the behaviour sub-model, which predicts the product properties. In the case of most steels, the microstructure model must include the transformation from austenite on cooling, which itself requires a complex set of inter-relationships. Within the structure sub-model, the most important feature during multi-pass deformation processing is to quantify the extent to which recrystallisation takes place between

SEQ 0151 JOB IOM8627-009-007 PAGE-0151 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 151

Fig. 8 Schematic diagram of the sub-models involved in metallurgical modelling of thermomechanical processing operations, showing the required information flow (4).

Fig. 9 Polarised light optical micrographs showing the progress of recrystallisation in an aluminium – 5% magnesium alloy deformed in plane strain compression to a strain of 1.0 at 400°C and annealed at 380°C for (a) 25 s, (b) 50 s and (c)120 s.

SEQ 0152 JOB IOM8627-009-007 PAGE-0152 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

152

Hatfield Memorial Lectures Vol. III

each pass because, as discussed earlier, this has a large effect on the flow stress in subsequent passes. Recrystallisation also has a major effect on the microstructure as illustrated by the optical micrographs in Fig. 9. This figure shows that, with increasing time, recrystallisation progressively replaces the elongated, work hardened grains by new equiaxed strain-free grains. In this particular example, the annealing times are relatively long, but at higher temperatures and particularly in austenite, the recrystallisation process can be very rapid. The extent to which it occurs before any pass affects the deformed structure after the pass and hence the kinetics of recrystallisation in subsequent interpass time intervals. The fraction of recrystallised grains (X) within the structure can be measured and the results plotted as a function of time to obtain a curve, which is typically as illustrated in Fig. 10. Recrystallisation may ‘start’ (X = 5%) in a short time, t05, but may take a much longer time, t95, to ‘finish’ (X = 95%). This is disguised in Fig. 10 by the logarithmic time scale, but typically the ratio t95/t05 is 10 to 100. In today’s models, the whole curve is described by an equation based on the time for 50% recrystallisation, t50. This time is very sensitive to all the process variables and this sensitivity is captured by semi-empirical equations, to give the fraction recrystallised after each pass, and the size of the recrystallised grains, drex. In the case of microalloyed steels and some other alloys, the model must also include relationships for the kinetics of strain-induced precipitation of fine second phase particles. When precipitation starts at time, t0.05,ppt, in Fig. 11, there is a dramatic retardation of recrystallisation. The models developed by different groups, e.g. SLIMMER (5), Corus (6, 7), VAI (8, 9), differ in detail in the form of relationships used in each sub-model, but essentially they operate in the same way to compute the evolution of microstructure and its consequences pass by pass. This is illustrated for the rolling of microalloyed steel plate in Fig. 12. Figure 12(a) shows how the temperature at mid-length of the plate changes throughout the rolling sequence of 8 roughing and 9 finishing passes, which take place over a time period of nearly 400s from discharge of the slab from the reheat furnace. The mean temperature

Fig. 10 Schematic diagram of the evolution of the fraction recrystallised as a function of the logarithm of time of annealing after deformation.

SEQ 0153 JOB IOM8627-009-007 PAGE-0153 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 153

Fig. 11

Schematic diagram of the influence of strain induced precipitation in microalloyed steel on the static recrystallisation kinetics of austenite.

of the plate falls gradually with time, but the centre of the thickness heats up initially because of the work done, whereas the surface cools rapidly in air and then is severely chilled as it passes through the water of the descalers and through the rolls in each pass. Between passes the surface temperature recalesces by heat flow from inside the plate. The only temperatures that can be measured on the mill are surface temperatures, and these measurements are only reliable when recalescence has taken place. The points on the graph show such pyrometer readings and clearly validate the output of the temperature model. Figure 12(b) shows the evolution of the austenite grain size, which indicates by the rapid drop after each of the roughing passes and the first six finishing passes that complete recrystallisation has taken place. This is followed by grain coarsening for the remainder of the interpass periods. This process is very temperature sensitive so is most pronounced after the roughing passes. Because of the temperature gradients through the plate thickness, significant gradients in microstructure develop. These are, however, reduced as the plate becomes thinner and cooler in finish rolling. The predicted start of strain-induced precipitation of niobium carbo-nitride, Nb(C,N), after finishing pass six is indicated in the figure. Figure 12(c) shows the consequences of the changes in temperature on the hot strength of the steel. In this figure the filled circular points are the values calculated from the rolling loads measured at each stand of the mill. The triangular points are predicted by the model on the basis of the recrystallisation and grain coarsening behaviour. It can be seen that these agree reasonably well with the measured values until the 8th finishing pass, but the prediction for the final pass is 24% too low. The model was therefore modified to take account of the direct strengthening effect of the strain induced Nb(C,N) precipitation. The results of this modification are shown by the open circular points, which are now within 7% of the measured value for the final pass. Overall, the current form of modelling involving microstructures observable on an optical scale (≤x1000) leads to results in good agreement with the measurements that can be made on the mill, and it also fills in the gaps when measurements cannot be made.

SEQ 0154 JOB IOM8627-009-007 PAGE-0154 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

154

Hatfield Memorial Lectures Vol. III

Fig. 12 Predictions of a microstructure based model (SLIMMER) for industrial plate rolling of microalloyed steel from 230mm slab to 11mm plate, (a) mean, centre and surface temperatures, showing comparison with pyrometer readings for the surface, (b) austenite grain size and (c) hot strength compared with values derived from rolling load measurements. The effect of precipitation strengthening is shown (5).

CURRENT STATUS OF MODELLING As discussed earlier, the investment in modelling can only be justified industrially by an improvement in consistency of product attributes and properties that lead to cost effective improvements in competitiveness. To illustrate the current status, I have therefore chosen two examples using models developed by industrial companies. The first example, relates

SEQ 0155 JOB IOM8627-009-007 PAGE-0155 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 155

Fig. 13 Comparison of rolling loads predicted by the Corus microstructure based model (Metmodel) and loads measured during roughing and finishing rolling on the Teesside coil plate mill (6).

Fig. 14 Centreline gauge performance as a function of time before and after installation of the microstructure based model (Metmodel) in the control system of the Teesside coil plate mill in 1998 (7).

to the model developed by Corus (6,7) for the effects of microstructure on rolling loads. Figure 13 shows the correlation between predicted and measured loads during rolling in the coil plate mill at Teesside. Bearing in mind the uncertainties in the actual temperatures on the mill, this figure shows good agreement between prediction and measurement over a wide range of conditions. The effect of introducing the microstructure model into the mill control program is shown in Fig. 14. Within a few weeks, the percentage of plates within thickness and shape tolerance increased from about 97% to 99%, with obvious economic benefits. However, this resulted in tighter specifications, which initially reduced the acceptable yield again, but only for a matter of weeks. The second example relates to the VAI model (8, 9) for the effect of rolling variables and coiling conditions on the microstructure and properties of C-Mn and high strength low alloy (HSLA) steel strip. Figure 15 shows the relationship between measured ferrite

SEQ 0156 JOB IOM8627-009-007 PAGE-0156 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

156

Hatfield Memorial Lectures Vol. III

Fig. 15 Comparison of ferrite grain sizes measured on C-Mn steel and on microalloyed high strength low alloy (HSLA) steel strip with sizes predicted by the VAI microstructure based model (9).

Fig. 16 Comparison of measured values of (a) yield strength and (b) tensile strength of strip from normal production of microalloyed steels with the values predicted by the VAI microstructure based model (9).

grain sizes and the values calculated by their model. Bearing in mind the uncertainties in the experimental measurement of grain size, this is an impressively good correlation. The ferrite grain size is a critical input into their behaviour model and, therefore, the correlation between measured and calculated mechanical properties of strip produced over a wide range of processing conditions, Fig. 16, is very satisfactory. The introduction of this model into the control program for the coiling temperature enhanced the consistency of the product properties by nearly halving the standard deviations (9). This improved control package justifies the investment in the model development for a mill builder.

SEQ 0157 JOB IOM8627-009-007 PAGE-0157 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 157

Fig. 17 Results of finite element modelling of experimental rolling of an aluminium–1% magnesium slab of 20mm initial thickness by 50% reduction at a peripheral roll speed of 235mms–1 at an ingoing temperature of 400°C, showing (a) the distortion of elements as a function of position in the stock thickness, (b) distribution of total equivalent strain, (c) distribution of local equivalent strain rate, s–1, and (d) distribution of temperature, °C.

It may be implied from these examples that the current modelling methodology has achieved its goals. However, when the semi-empirical microstructural relationships are applied to the more complex geometries of rod and section rolling, and forging of shaped products, the limitations become apparent. The reasons for these limitations can be illustrated by considering the deformation produced by flat product rolling in more detail, so that the expected variations in structure through the thickness can be predicted. Figure 17 (a) shows the results of finite element modelling of the deformation in a single pass experimental rolling. It can be seen that near the centre of the strip thickness, an initially square element is compressed to a rectangular shape expected from a simple analysis of the overall reduction in the pass. As one moves towards the surface, the friction between the stock and rolls, which is essential to draw the stock into the roll gap, causes additional shear strain. Figure 17 (b) shows how the total strain (compression plus shear) builds up through the roll gap to give a higher final strain near the surface. Figure 17 (c) shows that the rate of this build-up (i.e. the strain rate) at the centreline gradually increases from zero at entry to a maximum near the mid-plane of the roll gap and then falls gradually to zero at exit. Near the surface the strain rate rises rapidly to a high value near entry then falls to zero about two thirds of the way through the roll gap, where there is no slip between the stock and the rolls, then rises slightly again as the shear strain reverses, before again falling to zero at exit. To add to these complications, temperature gradients are developed,

SEQ 0158 JOB IOM8627-009-007 PAGE-0158 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

158

Hatfield Memorial Lectures Vol. III

Fig. 18 Comparison of the evolution of strain path angle through the roll gap close to the surface and at the centreline, computed using finite element modelling of a hot rolled aluminium–1% magnesium alloy subject to 50% reduction (10).

Fig. 19 Predicted through-thickness gradients in fraction recrystallised after annealing following laboratory rolling of Type 316L stainless steel, using two values of friction coefficient, compared with experimental measurements (11).

Fig. 17 (d), as a result of severe chilling by the rolls. Thus, even for the relatively simple geometry of flat product rolling, every element in a vertical column before entry undergoes a different strain history. As well as the changes in strain rate and temperature, shown in Fig. 17, the strain path, i.e. the angle of straining with respect to the orientation of the element, changes. This can be represented by a single angle, which is shown in Fig. 18 as a function of position of the

SEQ 0159 JOB IOM8627-009-007 PAGE-0159 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 159 (a)

(b)

Fig. 20 Thin foil electron micrographs showing the dislocation structures developed in (a) austenite in a model microalloyed 30%Ni, 0.1%Nb, 0.09%C steel deformed at 950°C (12) and (b) a commercial purity aluminium–1% magnesium alloy deformed at 385°C (13).

element as it passes through the roll gap. At the centre line, the strain path angle is low, except in local regions near entry and exit. If it were zero, this would mean monotonic straining, which is aimed for in the laboratory tests used to develop the microstructural equations in today’s models. When these equations are applied in the model for predicting the fraction recrystallised at a certain time after exit from the rolls, the results are shown as a function of the position of the element in the thickness from centre to surface in Fig. 19. It can be seen that near the centreline there is good agreement between the predicted and measured results, but as one moves towards the surface there are major disagreements. These arise because, near the surface, the strain path angle changes rapidly in the roll pass, Fig. 18, and is far from zero. Today’s models, which are based on equivalent tensile stresses and strains, cannot take account of changes in strain path, so if one is interested in local microstructures and properties a new approach is required.

DEVELOPMENTS IN MODELLING METHODOLOGY In order to account for the effects of strain path on the microstructures that can be observed in the optical microscope, advanced models must describe the structures developed inside the deformed grains. These can be observed in an electron microscope using thin foils of the deformed metal, as illustrated in Fig. 20. This shows the dislocation structures in (a) a model microalloyed steel austenite, and (b) an aluminium alloy. It is clear that the dislocations are not uniformly distributed and that the structures in Fig. 20 (a) and (b) have a number of features in common. The current debate is how best to describe the features as internal state variables in models. Clearly, the total density of dislocations, ρ, is important, but to describe the distribution of dislocations also needs the subgrain size, δ, i.e. the spacing between the dislocation walls, which is 1 to 2 µm in Fig. 20, and the misorientation across these walls, θ, which reflects the spacing of dislocations in the walls themselves. In addition, it can be seen in Fig. 20 that the subgrains are not equiaxed and that the walls are preferentially aligned in a specific

SEQ 0160 JOB IOM8627-009-007 PAGE-0160 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

160 Hatfield Memorial Lectures Vol. III direction with respect to the direction of straining. Describing this alignment in the model appears to be the key to predicting the effects of changing strain path angle. However, at present there are few data, with which to calibrate any model description, and much further experimental research is needed before a new generation of knowledge-based internal state variable models is available. As discussed earlier, development of such knowledge-based models requires long times, whereas industry would like to be able to take advantage of improved models incrementally as each step of improvement is made. To satisfy both needs requires a new approach to modelling. A promising approach being developed in the Institute for Microstructural and Mechanical Process Engineering: The University of Sheffield (IMMPETUS) is called Hybrid Modelling (14, 15) and is shown as a block diagram in Fig. 21. The core of the model is the description of the internal states variables as they change during the deformation process. These are computed in the Dynamic states box. This is based on a description of the internal state variables under the simpler Steady states conditions. As indicated above, there is limited understanding and few data even for these simpler conditions, so, at present, computation of the internal states from the input Deformation conditions and composition is carried out by a ‘black box’ or, at least, a ‘dark grey box’ fuzzy-neural network model (16) based on the limited data that are available. From these internal state variables the Recrystallisation behaviour and the Stress components of the material are computed. There is rather more understanding of the effects of internal state variables on the properties, so this step of the model is ‘white box’ or, at least, ‘pale grey box’ using physically based equations. In the aluminium magnesium alloys, for which this hybrid model is being developed and validated, there is a contribution to the flow stress that is not sensitive to the microstructure. This Friction stress is also modelled using the fuzzy-neural method. Thus, the overall model from the input deformation conditions to the outputs of flow stress and recrystallisation behaviour is a hybrid of data-based and knowledge-based modelling. This approach maximises the use of all the information available. To take advantage of this, the model is attached to a Knowledge Base, which contains the sources of all the data and physics used in the model, and which can be updated as more information becomes available through continuing research (15). This then feeds back into the model to enlighten the greyness. One other important feature of the modelling recognises the fact that there are more empirical data available for the effects of the input deformation conditions on the flow stress and recrystallisation than there are on the internal state variables. These data can be used in the Genetic Algorithm Optimiser to optimise the constants in the ‘black box’ parts of the overall model to improve the accuracy of the final outputs. This type of optimisation is called a genetic algorithm because it is based on Darwin’s theory of evolution, incorporating the concept of ‘survival of the fittest’ (17). An initial set of tuning constants is given a set of ‘genes’ and the values are then perturbed by a first generation mutation. The modified constants that reduce discrepancies between observed and predicted outputs survive and are allowed to mutate for many generations

SEQ 0161 JOB IOM8627-009-007 PAGE-0161 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 161

Fig. 21 Block diagram of the hybrid model developed to predict the flow stress during hot deformation and the subsequent static recrystallisation kinetics and recrystallised grain size of aluminium–magnesium alloys from the evolution of internal state variables as a funtion of deformation conditions (15).

Fig. 22 Block diagram showing use of the hybrid model in an inverse sense to compute virtual microstructures for comparison with measured dislocation structures (15).

until a specified level of precision is achieved, or further mutation leads to no further significant improvement of results. The method is a robust one and enables industrially useful accuracy to be obtained even from the present level of understanding. The hybrid model is also useful to researchers involved in the time-consuming electron microscopy required to characterise the internal state variables. As illustrated in Fig. 22 for the recrystallisation behaviour, the ‘white box’ part of the model can be used in an inverse sense so that empirical data on the effects of the deformation conditions on recrystallisation predict the internal state variables required i.e. the model can generate Virtual internal states. These virtual microstructures can be used to optimise the range of

SEQ 0162 JOB IOM8627-009-007 PAGE-0162 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

162

Hatfield Memorial Lectures Vol. III

Fig. 23 Schematic drawing of the detailed processes of the hybrid model (14).

deformation conditions to be used in the experimental research, so that differences in the internal state variables are well outside the confidence limits of the measurement techniques, ensuring that the most significant trends are determined first. The actual architecture of such hybrid models becomes rather complex, Fig. 23. One might ask if such complexity is necessary, to which the best reply is to quote Einstein: ‘everything should be made as simple as possible, but no simpler’. We have already seen that the present metallurgical models are too simple for some purposes. The new internal state variable models should therefore be capable of including all the metallurgical phenomena, which may be important in some circumstances, to act as a benchmark. For application to the more limited range of conditions in any specific industrial process (Fig. 1), the model may be simplified by using simpler algorithms or by using the outputs from the benchmark model to train a neural network model to operate over the range of process variables of interest. Within any particular process, the variables may be more restricted for a particular plant or mill. Application of the genetic algorithm to individual plant data may therefore lead to somewhat different optimised process variables depending on the constraints imposed by the particular plant. The essential feature is that, when plants are making the same product, each different plant should be able to achieve competitive quality. In order for hybrid models to achieve these goals, there are a number of essential features about their architecture, Fig. 23. We do not need to be concerned with the detail of the model, except to appreciate that it is modular in nature, and 1. it should be possible to ‘unplug’ any module and replace it, either with a simpler algorithm, or with a better module as understanding develops with time;

SEQ 0163 JOB IOM8627-009-007 PAGE-0163 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Metallurgical Modelling of Thermomechanical Processing 163 2. each module should have its own knowledge management scheme attached to it, which is accessible simply by a ‘mouse click’. This might show the equations and/or data base in the module, the range of proven validity, and the sources of information, with links to these sources; i.e. it should contain everything that the next generation of experts needs to understand and, if necessary, to modify the module. Ideally the knowledge management system will be developed to interact actively with the model so that, if the model is run for conditions outside its range of validation, the results will appear with a ‘health warning’, which should estimate the increased level of uncertainty in the output and highlight the sub-module from which it arises. This would help prioritise further research needs, but is still a step into the future. CONCLUDING REMARKS The benefits of applying metallurgical modelling to the control of thermomechanical processing operations have been established, particularly for flat product rolling. The current methodology, however, fails if applied to compute local microstructures and properties when the deformation conditions and the strain path change in complex ways during a pass. To overcome this limitation, a new generation of models is required, which consider the microstructure on a finer scale. Hybrid modelling of the evolution of internal state variables, which describe the dislocation structures in the deformed material during and after each pass, and their effects on flow stress and recrystallisation behaviour, provides a flexible methodology for through-process modelling. Such models will enhance process control but can also: 1. aid understanding of the complex interactions between the external variables and the resulting structure and properties of the products of industrial thermomechanical processing; 2. enable sensitivity analysis to be carried out to optimise the return on experimental research to determine the influence of deformation conditions on the internal state variables; 3. prioritise the research needs to improve the physical basis and the precision of the predictions of the models; 4. with an attached knowledge base and knowledge management system, they can also encapsulate corporate wisdom in a leaner, meaner, more volatile industrial world. ACKNOWLEDGEMENTS I am grateful for valuable discussions and for the provision of illustrations for the lecture to Professors John Beynon and Mark Rainforth, Drs Brad Wynne, Jesus TalamantesSilva, Maysam Abbod, Qiang Zhu and Mr Lucian Tipi of IMMPETUS, to Drs Peter Morris, Andrew Rose, Roger Beaverstock and Sally Parker of Corus, Swinden Technology Centre, and to Mr Mick Steeper of VAI (UK) Limited.

SEQ 0164 JOB IOM8627-009-007 PAGE-0164 143-164 50 SELLA REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

164

Hatfield Memorial Lectures Vol. III REFERENCES

1. C. M. SELLARS and T. HOPE, Conf. On Hot and Cold Deformability of Metals and Alloys, Czechoslovak Society of Science and Technology, Ostrava, 1983, vol.II, 23–38. 2. B. DUTTA and C. M. SELLARS, Mater. Sci. and Technol. 1986, 2, 146–153. 3. M. T. HAGAN, H. B. DEMUTH and M. BEALE, Neural Network Design, PWS Publishing Company, Boston, 1996. 4. J. H. BEYNON and C. M. SELLARS, Proc. 13th Riso Internat. Symposium on Materials Science: Modelling of Plastic Deformation and Its Engineering Applications, Ed. S. I. Andersen et al, Riso National Laboratory, Roskilde, Denmark, 1992, pp 13–26. 5. J. H. BEYNON and C. M. SELLARS, ISIJ International, 1992, 32, 359–367. 6. R. C. BEAVERSTOCK, Corus Swinden Technology Centre, private communication. 7. A. J. TROWSDALE, J. P. TUNSTALL, K. RANDERSON, P. F. MORRIS and Z. HUSSAIN, Proc. Conf. on Modelling of Metal Rolling Processes 3, London, Dec.1999, Ed. J. H. Beynon et al, IOM Communications Ltd, London, 1999, pp 12–21. 8. P. STIASZNY, J. ANDORFER, G. HRIBERNIG, A. SAMOILOV, A. LUGER, G. HUBMER and D. AUZINGER, LA REVUE DE METALLURGIE – ATS – JSI, 2002, pp132–133. 9. J. ANDORFER, D. AUZINGER, G. HRIBERNIG, G. HUBMER, A. LUGER and P. SCHWAB, Internat. Conf. on Thermomechanical Processing: Mechanics, Microstructure and Control, Sheffield, June 2002, Ed. J. Palmiere, M. Mahfouf and C. Pinna, BBR Solutions Ltd, Chesterfield, 2003, pp 164–168. 10. J. H. BEYNON and B. P. WYNNE, Proc. Conf. Engineering Materials 2001, Melbourne, Sept. 2001, Ed. E. Pereloma and K. Raviprasad, The Institute of Materials Engineering, Australasia Ltd, 2001, pp 1–10. 11. A. J. MCLAREN, N. LEMAT, J. H. BEYNON and C. M. SELLARS, Ironmaking and Steelmaking, 1995, 22, 71–73. 12. W. M. RAINFORTH, M. P. BLACK, R. L. HIGGINSON, E. J. PALMIERE, C. M. SELLARS, I. PRABST, P. WARBICHLER and F. HOFER, Acta Mater. 2002, 50, 735–747. 13. C. M. SELLARS and Q. ZHU, Proc. 2nd Symposium on Hot Deformation of Aluminium Alloys, Rosemont, Il, Oct. 1998, Ed. T. R. Bieler, L. A. Lalli and S. R. MacEwan, The Minerals, Metals and Materials Soc., Warrendale, Pa, 1998, pp185–197. 14. Q. ZHU, M. F. ABBOD, J. TALAMANTES-SILVA, C. M. SELLARS, D. A. LINKENS and J. H. BEYNON, Acta Mater., 2003, 51, 5051–5062. 15. M. F. A BBOD , C. M. S ELLARS , D. A. L INKENS , Q. Z HU and M. M AFOUF , Mater. Sci. and Eng.A., 2004, accepted for publication. 16. J-S. R. J ANG , C-T. S UN and E. M IZUTANI , Neuro-fuzzy and Soft Computing: a Computational Approach to Learning and Machine Intelligence, Prentice Hall, Upper Saddle River, N. J., 1997. 17. J. R. K OZA , Genetic Programming: On the Programming of Computers by Means of Natural Selection, MIT Press, Cambridge, Ma., 1992.

SEQ 0254 JOB IOM8627-010-006 PAGE-0167 SEVENTH LECTURE REVISED 20DEC04 AT 21:55 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE SEVENTH HATFIELD MEMORIAL LECTURE

Development in the Iron and Steel Industry in Great Britain during the last Twenty-Five Years T. P. Colclough At the time the lecture was presented Dr Colclough was Technical Adviser to the British Iron and Steel Federation. The lecture was given at the Royal Institution, Albemarle Street, London, W.1, on Wednesday, 26th May, 1954.

The features that characterise the development of the iron and steel industry during the last 25 years are reviewed. The experience gained in the 1920s resulted in the formulation of certain principles for the reconstruction of the industry, and it is shown how these were applied during 1930–1939. The war years demonstrated the soundness and economies of the pre-war projects and proved the benefits to be derived from the joint discussion and co-ordination of effort to solve common problems. The rapid and extensive development of the post-war years are outlined, and it is shown how productivity has improved during the period under review. Reference is made to some aspects of future development and the growing importance and influence of research is stressed. The essence of the development of the last 25 years has been the creation of a new spirit and unity within the industry.

INTRODUCTION For those at present engaged in the steel industry, with all its activities and wide ramifications, it may sometimes be difficult to realise that the making of steel in bulk began less than 100 years ago and that this survey of development must cover one-quarter of the life of the industry. In magnitude, the growth of production, in spite of temporary variations, has been remarkably steady and progressive, and there can be no doubt that as new needs arise and new uses are devised the industry will continue to grow. The tempo of growth may fluctuate, but the growth must be unceasing if the standard of life of the nation is even to be maintained, much less raised. It is significant that no vital new principle for the making of iron or steel, or for the rolling and shaping of steel, has been discovered during this last quarter-century. The characteristic feature has been the application of scientific knowledge to the industry,

167

SEQ 0168 JOB IOM8627-010-005 PAGE-0168 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

168

Hatfield Memorial Lectures Vol. III

based on an intensive study of the principles underlying the processes employed, with a resulting far greater knowledge and better understanding of the operations involved. On the one hand this has led to a marked improvement in the efficiency of the manufacturing processes themselves, a raising of the general standard of quality, and the development of many new types of steel with specialised properties. On the other hand, the inventive genius of the engineering section of the industry has been devoted to the application of scientific knowledge to the development of larger and more efficient units of production; the mechanisation of plant with its reduction of manual labour; the greater use of electric power to replace steam as the source of energy and to replace raw fuel for melting and other purposes; and the development of instruments to give better control and greater precision to the manufacturing operations and, by reducing the dependence upon human judgment and reaction, to effect a remarkable increase in the speed and accuracy of operation. These developments have demanded and made possible a complete change in both the general lay-out and detail of the manufacturing operations, and this in turn has demanded considerable and important changes in the commercial and administrative organisation. It may be of interest to mention briefly a few salient examples of the changes which have taken place. In 1929 there were roughly 400 blast furnaces in existence in this country. About one-half, or 200 furnaces, had an average capacity of 400 tons per week, totalling together a capacity of, say, 80,000 tons per week. Today, eight operating furnaces have a capacity of over 50,000 tons per week, and when four other furnaces, at present under construction, come into operation the 12 furnaces are scheduled to produce 83,000 tons per week. In steelmaking, the production per open hearth furnace year has more than doubled, rising from 15,000 tons in 1929 to 39,000 tons in 1953. Electric furnace capacity has risen from about 100,000 tons to over one million tons per year. In the rolling mills, the growth in number and capacity of the continuous mills has been phenomenal and the capacity of the latest primary rolling units recently brought into operation is roughly four times that of their corresponding units in 1929. The target for production, three years hence, is 15 million tons of pig iron, practically double that of 1929, and roughly 20.5 million ingot tons of steel, an increase of more than 110% on 1929, which was at that time an all-time record. That is the general picture; the sequence of events will now be examined a little more closely.

CONDITIONS IN YEARS 1920–29 The years 1920–29 marked one of the most difficult periods in the history of the industry. The artificial demand of the war years 1914–18 had seen the installation of steelmaking and rolling plant far in excess of the normal growth of requirements. Much of the new plant had been built to meet the emergency conditions at relatively high

SEQ 0169 JOB IOM8627-010-005 PAGE-0169 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Development in the Industry in Great Britain during the last Twenty-Five Years 169 capital cost and was unsuited for normal commercial requirements. In addition, many old units of production, which normally would have been due for scrapping, were retained. As a result of the financial, economic, and industrial conditions after the war, the demand for steel fell away and many plants were either closed down or operated below capacity. The realistic capacity of the blast furnaces existing in 1920 may be assessed at about 10– 11 million tons. The production during the ten years 1920–29 averaged only 6 million tons, or even omitting the two years of prolonged strikes, 6.87 million tons, not more than 60% of capacity. In the steelmaking plants, the capacity in 1920 may be estimated at about 12 million tons, whilst the average production over the ten years was only 7.36 million tons, or again roughly 60% of capacity. The position was aggravated by severe competition from the Continent, particularly during the second half of this period. A large amount of new plant was installed in Western Europe, fostered by reparations and American investment, and the surplus output from these new plants was dumped in Britain. During this 10 year period the imports of steel averaged 2.08 million tons per year, or the equivalent of about 2.6 million tons of ingots. It may be estimated that, in spite of the general state of industrial depression in this country as a whole, the demand for steel averaged about 9.9 million tons per year, but of this amount only 7.36 million tons per year were actually made in this country, 26% of the steel required was imported. During the five years 1925–29, the total steel imported was roughly 3 million tons per year, and of this over 1.5 million tons were in the form of semi-finished steel, wire rod, or strip, mainly used for tube making. The salient facts were: (i) Much of the existing plant was old and obsolete. (ii) Not more than 75% of the steel requirements was being made in this country. (iii) About 50% of the steel imported was in the semi-finished form. The struggle for survival compelled a drastic revision of the financial structure of many companies; the scrapping of old, obsolete plants; and the closing of some works. Above all, it led to an intensive study of the methods by which plant and processes could be improved, operating costs lowered, and general efficiency raised so as to provide a sound financial basis for reconstruction. As is well known, the decision was made that it was imperative to place the steel industry on a sound economic basis. To this end it was determined that new plant must be installed; the industry must be re-organised to establish operating units of economic size; the imports of steel must be controlled, to prevent the unfair competition from abroad; and facilities should be created to provide at reasonable rates the financial resources necessary for the reconstruction projects. The steps taken to accomplish the provision of financial resources and the reorganisation of the industry as a whole are not within the scope of this review; attention can be given only to the progress made within the industry.

SEQ 0170 JOB IOM8627-010-005 PAGE-0170 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

170

Hatfield Memorial Lectures Vol. III RECONSTRUCTION

From the practical aspect, the last twenty-five years fall naturally into three divisions, the pre-war, war, and post-war years. The ten years 1930–39 can be regarded as a period of experiment and tentative progress, the war years were perforce relatively static, and the post-war years have marked a period of rapid and extensive development based on the knowledge and experience gained in the earlier stages. The Pre-War Years (1929–39) The experiences gained in this country and abroad during the years 1920–29 had demonstrated clearly the inherent advantages of integrated operation. The transfer of metal from one operation to the next while still in the liquid or hot condition made possible marked reductions in the cost of fuel and interdepartmental transport. The close association of coke- and iron-making gave a wider field of application of the valuable coke oven and blast furnace gases to the steelworks operations, thereby reducing the demand for raw coal and replacing it by a fuel which could be more readily controlled and more efficiently used. As a result, certain principles were formulated as a basis for the reconstruction programme. These were: (i)

The new plants should, wherever practicable, be integrated from raw coal and iron ore to the production of rails, heavy sections, plates and semi-finished steel. (ii) Coke and pig iron production should be centralised; the use of blended coal mixtures for coke making should be extended and the new coke ovens should be designed for under-firing with blast furnace gas. (iii) Blast furnace practice should be improved by the grading of the ores and sintering of the fines. (iv) Pig iron production should be assured a sound long term basis by consolidation of the available home ore reserves. (v) Steelmaking and rolling operations should be concentrated to the point at which the production would be adequate for large-scale, low cost operation. (vi) The operations of the primary steelmakers should be extended by the acquisition of certain re-rolling or finishing operations, so as to complete the integration of operations and provide an assured outlet for capacity production direct to the final customer. (vii) In determining the location of improvements or new installations, special regard should be paid to the freight costs involved in the assembly of the raw materials used and in the despatch to market of the finished products. These principles furnished the fundamental bases for the modernisation of the industry. What were the steps taken to implement these principles?

SEQ 0171 JOB IOM8627-010-005 PAGE-0171 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Development in the Industry in Great Britain during the last Twenty-Five Years 171 Whilst it is clearly impossible to give a detailed account of the many developments that occurred in the ten pre-war years, it is possible to select a few which illustrate the manner in which these principles were applied. Lancashire Steel Corporation The first of these deals with the position in South Lancashire, one of the centres of wire production. In this area there were blast furnace plants, steelmakers, re-rolling plants, wire drawers and wire fabricators, in total a substantial tonnage; but the different industries were almost totally unco-ordinated and the wire industry in a large measure relied on imports. The economic position of the iron and steelworks was definitely unsound, and reconstruction of the production units was imperative. To provide a sound basis for reconstruction, it was decided to co-ordinate these industries into a single corporate body, The Lancashire Steel Corporation. The manufacture of pig iron and ferro-alloys was concentrated in the blast furnaces at Irlam, which were furnished with a new dock, rapid unloading equipment, ore storage yard and ore crusher. New gas cleaning equipment was installed at the blast furnaces, and new coke ovens designed to operate on blast furnace gas were installed. The steel plant was rebuilt, embodying furnaces of new design with more rapid operation and equipped for the use of mixed gas and liquid fuel. A new continuous rod mill of the latest design was installed at Irlam for the re-rolling of their own billets; the wire rod was transferred to the nearby wire industry at Warrington. This development represented not only the integration of the iron and steel operations, but a complete integration from the raw materials, coal and iron ore, to the finished end product of wire and wire product. This plant is now in process of being expanded in capacity by about 50%. Guest, Keen Nettlefolds and Guest Keen Baldwins Similarly, a number of isolated iron and steel works in South Wales and Scunthorpe were integrated with users of semi-finished steel in South Wales and Birmingham to form the two companies, Guest, Keen, and Nettlefolds Ltd. and Guest Keen Baldwins. Their iron and steelmaking operations in South Wales were concentrated in the works at Port Talbot and in a new plant at Cardiff. The latter plant comprised a completely integrated manufacturing unit of ore unloading docks, coke ovens, blast furnaces, steel furnaces, continuous billet mill and continuous bar, strip, and wire rod mills, together with a mill for the re-rolling of sections suitable for colliery arches, etc. This plant, designed for a nominal production of 7000 tons per week, has since the war made and rolled over 15,000 tons of ingots per week and has made the two Companies largely self-contained in regard to their steel requirements in these forms. Stewarts and Lloyds A third example illustrates another form of plant and company integration. In 1929 the making of steel tubes was carried out by a large number of relatively small, widely scattered companies who relied to a marked extent on tube strip imported from the

SEQ 0172 JOB IOM8627-010-005 PAGE-0172 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

172

Hatfield Memorial Lectures Vol. III

Continent. After a thorough examination of the possibilities, Stewarts and Lloyds decided to concentrate certain sections of their manufacturing operations in a new plant to be erected at Corby, completely integrated from ore and coal to finished tube. This project involved a combination of new principles, which may be briefly summarised as follows. It was necessary to establish the first steelmaking operation on the Midland ore body, this ore previously having been regarded as unfavourable, when used alone, for the making of basic iron. Ore mining was organised on a scale hitherto unknown in this area and was equipped with, at that time, the largest ore stripper in the world. The small blast furnaces were replaced by four new furnaces, up to 20 ft. hearth diameter, larger than it was generally considered possible to operate on Northamptonshire burdens. The whole of the ore was thoroughly graded and the fines were sintered before charging. Coke ovens were installed to provide the necessary coke, and the coals were blended to give the type of coke required. The whole of the pig iron made was desulphurised before being used for steelmaking. For steelmaking the basic Bessemer process, which had gone out of use in 1925, was revived. The rolling mills comprised a semi-continuous slab and billet mill, a semicontinuous and a fully continuous mill for the rolling of tube strip. The tube mills included a new 40 ft. seamless pipe mill and a new innovation, the rolling of butt welded pipe by a fully continuous process. Simultaneously with the erection of the new plant, the Company pursued a policy of commercial amalgamation with a number of other tubemaking concerns, so as to give a fully economic load to the new plant and to provide an organisation capable of handling and marketing the whole of the output of the integrated operation. Richard Thomas and Co. and John Summers and Sons The fourth example deals with another facet of this many sided problem, a completely new manufacturing process. In the USA the 1920s were a period of marked development in the installation of continuous mills for the hot rolling of strip, and this was followed by the discovery that, given sufficient power and speed, steel strip in the form of coils could be reduced cold in continuous mills. This led to the installation of continuous mills for the hot rolling of slabs into coils of strip and the cold reduction of the strip to sheet and tinplate gauges. It became apparent that if Britain was to maintain her position as a manufacturer of these products it was essential to adopt this new process, and it was decided by Richard Thomas and Co., Ltd, to erect a completely new integrated plant on the site of the old works at Ebbw Vale for the manufacture of sheet and tinplate, and in particular to meet the requirements of the motor-car industry for car body sheets. Shortly afterwards, John Summers and Sons installed a semi-continuous mill for the hot-rolling of coiled strip and a fully continuous mill for cold-reduction to sheet gauges, in conjunction with their cold-metal melting shops at Shotton, United Steel Companies The last type which can be dealt with in this brief survey may be illustrated by the development of the United Steel Companies. This group, with its associates, was formed

SEQ 0173 JOB IOM8627-010-005 PAGE-0173 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Development in the Industry in Great Britain during the last Twenty-Five Years 173 to consolidate a wide range of manufacturing interests and products. The Company was a pioneer in many directions, and their development programme dates back to the first war period. The overriding feature may be regarded as the concentration of the different types of product into cognate groups suitable for production in one plant. Rail production is concentrated at Workington; plates and heavy sections at Appleby–Frodingham; billets, strip and bars at Templeborough; heavy forgings and the related products, wheels and tyres, at Rotherham; and alloy and special steel production at Stocksbridge. Each works is confined to the production of a definite range or type of product, and in each case the design of the works is on the scale which will give a productive capacity adequate to justify the capital expenditure that is required to provide the modern equipment installed. This record of initiative and enterprise for the years 1929–39 is probably without parallel in the history of the industry, and the success achieved by these courageous but well-based projects gave a sound foundation and pattern for the forward policy of the post-war years. The halt imposed on new building by the war conditions was probably of immense value, for it gave an opportunity to prove by experience the relative merits of the different types of development. The War Years Development during the war years 1939–44 showed quite a different pattern from that in the years 1914–18. In the first war, steel productive capacity was expanded as rapidly as possible and, as is well known, led to considerable difficulties in post-war years. In the second war it was, for several reasons, deemed expedient to concentrate efforts in this country on the production of vital steels and to provide all possible expansion overseas. The first obvious result was that, whereas the overall steel production in the UK in 1918 was 1.87 million tons (or 24%) higher than in 1913, the maximum steel production during the second war was never substantially above that of the peak pre-war year, 1937. On the other hand, there was a marked change in the pattern of the quality of steel made. This is demonstrated by two salient factors, the development of electric furnace production and the increased production of alloy steel. In the years 1937–38 the production of electric steel averaged about 220,000 tons, or 1.9% of the total steel made. This rose to a peak tonnage in 1943 of 990,000 tons, or 7.5% of the total production, a fourfold increase. In addition, a marked contribution to the war effort was made by the change-over of several of the open hearth steelmakers from ordinary commercial qualities to the manufacture of alloy and special steels. The production of alloy steels rose in 1943 to a peak figure of 1.6 million tons, or over 12% of the total steel production. Mere figures, however, are totally inadequate to present the true picture of the real development which British technical men and operators made in regard to quality during this period, and still fully maintain. This work placed Britain in a pre-eminent position in the manufacture of the highest qualities of steel, and it is to be hoped that the full story will be placed on record by a competent authority.

SEQ 0174 JOB IOM8627-010-005 PAGE-0174 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

174

Hatfield Memorial Lectures Vol. III

Probably the two most important factors emerging from the war years were: (a) the experience of these years demonstrated the soundness and economies of the pre-war projects and provided a solid foundation for further developments, and (b) the cooperative efforts imposed by war conditions proved the necessity for, and benefits to be derived from, joint discussion and co-ordination of effort to solve the common problems. This second factor applied particularly to the need for ensuring adequate supplies and correct distribution of the essential raw materials, which led to the establishment of a central body for the purchase and distribution of imported ore and the strengthening of the organisation built up before the war for the control of the scrap and steel required to fill the gaps in the home market. It was also advisable further to improve the arrangements made in the pre-war period for the joint examination of the development programmes of individual companies in the light of probable market demands, to ensure that estimated requirements would be met in the most economic manner, and, at the same time, to avoid unnecessary capital expenditure and wasteful competition. The Post-War Years Before the end of the war, discussions were initiated for the preparation of an agreed development programme. The principles underlying this forward policy and its broad outlines were agreed by December, 1945, and were published in the Command Paper of May, 1946. This document is too familiar to need recapitulation, but it may be of interest to review briefly how far and in what manner the objects set forth in that programme have been attained. The primary objects of the Five Year Plan, of which the principal projects were to be initiated by 1951 and completed by 1953, were: (i)

To increase steel production to about 15 million tons per year, an increase of 3 million tons (or 25%) over the pre-war maximum of 1937. (ii) To install about 5 million tons of iron and steelmaking plant of modern, efficient design. (iii) To install the most efficient design of rolling mills. (iv) To consolidate production in units of economic capacity and embodying the most recent methods of manufacture.

POST-WAR PRODUCTION The production target of 15 million ingot tons was achieved by 1949 and has been surpassed in each succeeding year. It was soon recognised that the target figure, optimistic as it appeared at the time, was not adequate to meet the growing demands of the country. After discussion with the Government Departments concerned, it was agreed that the target should be raised to 18 million ingot tons, to be achieved by the mid 1950s. In every

SEQ 0175 JOB IOM8627-010-005 PAGE-0175 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Development in the Industry in Great Britain during the last Twenty-Five Years 175 year except 1951, when there was a sharp fall in receipts of imported scrap, there has been a substantial increase in production. The production for 1953 was 17.6 million tons, and the present target is for a production of about 20.5 million tons in the years 1957–8. Pig Iron Production The first essential for steel production is an adequate supply of pig iron, and special attention has been given to this section of the industry. The production from the blast furnaces in 1953 totalled 11.2 million tons, an increase of 47% on the 1929 production. This has been achieved by a ruthless elimination of obsolete plant, a concentration of production in fewer works, a gradual but steady increase of size of furnace, and the wider application of burden preparation to improve the operating conditions within the furnace. The progress made is summarised in Table 1. Table 1

Blast furnace development, 1929–1954. Figures in parantheses denote percentages based on 1929.

Quantity Number of blast-furnace plants

1929

1937–38

1945

1953

88

57

53

44

Furnaces in existence

394

200

168

137

Average number in blast

158.25

111.5 (70.4%)

99.2 (62.7%)

104.7 (66.1%)

7589

7627 (105%)

7107 (94%)

11 175 (147%)

47 960

68 550 (143%)

71 640 (149%)

106 700 (222%)

Blast-furnace production, thousand tons Annual production per operating furnace, tons

The most significant factor is the increase in production per furnace operating, from 48,000 tons in 1929 to 107,000 tons in 1953; this increase is even greater if basic iron furnaces alone are considered. The outstanding feature is the growth in size of the furnace. The first ten years saw the experimental development from the maximum size of 15–17 ft. hearth diameter furnace to the 22 ft. hearth diameter of the two new furnaces of 1938 at Appleby– Frodingham. Since the war, these two furnaces have been expanded to 25 ft. diameter. Today, six furnaces of 25–28 ft. hearth diameter are operating and six furnaces of 27– 29 ft. hearth diameter are in course of construction. It is estimated that the target production of 15 million tons of iron in 1958 will be achieved by the operation of not more than 107 blast furnaces, an average of about 140,000 tons per furnace year, but with 12 furnaces of recent construction averaging 350,000 tons per furnace. It is not to be expected that there will be a further marked increase in blast furnace size in the near future. It is well established that the percentage effective hearth area diminishes as the actual size increases, and it is the general opinion that the optimum size has now been reached for present operating conditions.

SEQ 0176 JOB IOM8627-010-005 PAGE-0176 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

176

Hatfield Memorial Lectures Vol. III

The second dominating factor has been the growth of the ore preparation plants. Special attention may also be directed to the value of the ‘bedding’ plants installed at two works to secure a uniform composition of ore for charging into the furnace. The crushing and grading of the ore and the sintering of the ore fines are now accepted as essential for the economic operation of the blast furnace. The growth of this practice from practically nil in 1929 to about 9 million tons of sinter capacity for the current year is shown in Table 2. Table 2

Sinter plants in UK. Estimated production 1955: 81⁄2–9 million tons. Nominal Capacity, tons × 103

Production in 1953, tons × 103

Before 1929 1929–39 1940–53

120 3441 2247

77.1 2850.1 1806.9

Total to 1953 Projected 1954

5808 4000

4734.1

Total to end of 1954

9808

Installed

The beneficial effect of this treatment of the ore has been fully discussed at the recent Symposium on Sinter and needs no further elaboration. The extent to which the principle of sintering the ore before charging into the furnace can be applied to different types of ore and blast furnace practice with increasing economy is still subject to examination. It may well be that further investigations of this problem and the application of oxygen and/or higher furnace pressure operations may lead to modification of the present opinion on the optimum size of furnace. Steelmaking The principal features of the steelmaking development have been the increasing dominance of the basic open hearth process and a marked replacement of acid open hearth production by electric furnaces, together with the re-introduction of the basic Bessemer process in two special cases. The distribution of production by various processes is given in Table 3, which shows that the production of basic open hearth steel has doubled in the period under review and that 97% of the total increase in tonnage was made by this process. As in blast furnace development, the increased production of open hearth steel has been achieved by a reduction of the number of furnaces installed, the replacement of small units by furnaces of greater capacity, and a marked increase in the production per furnace employed. Table 4 shows that, while the number of furnaces installed has decreased from 595 to 397, the average size of furnace has increased from 55 tons to 95 tons per charge and the

SEQ 0177 JOB IOM8627-010-005 PAGE-0177 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Development in the Industry in Great Britain during the last Twenty-Five Years 177 Production of steel by various processes.

Table 3

1929 tons × 103

Process Bessemer: Acid Basic

1939

% of total

tons × 103

1953

% of total

tons × 103

% of total

559 Nil

5.8 ...

233 701

1.8 5.3

263 798

1.5 4.5

2451 6488

25.4 67.4

2157 9705

16.3 73.4

1133 14,298

6.4 81.2

Electric

87

0.9

292

2.2

929

5.3

Other

51

0.5

133

1.0

188

1.1

9636

100.0

13,221

100.0

17,609

100.0

Open-hearth: Acid Basic

Total

Table 4

Pattern of open-hearth furnace capacity 1929

Capacity, tons <60 60–79 80–99 >100 Total

1953

Average Capacity, tons

No. of Furnaces

%

Average Capacity, tons

No. of Furnaces

%

39.6 64.1 81.3 166.0

374 170 19 32

62.8 28.6 3.2 5.4

42.8 65.0 85.3 171.0

112 91 74 120

28.2 22.9 18.7 30.2

595

100.0

397

100.0

1929 Average capacity per charge, tons Production during year, tons × 103 Production per furnace year, tons Average number of heats per furnace year

54.7 8939 15,020 274

1953 94.5 15,431 38,870 411

production per furnace year from 15,000 tons to 38,870 tons. The development of integrated plants has increased the use of the hot metal process, and in 1953 practically 80% of the basic pig iron made was used for steelmaking in the liquid state. Tilting furnaces have been increased in number and in size from 150 tons to the latest nominal 400 tons per charge. Hot metal fixed furnaces are now of up to 200 tons capacity and cold metal furnaces range up to 150 tons. Apart from mere size and increased proportion of hot metal, marked improvements have been made in design, charging facilities, and particularly in the application of oil fuel and the instrumentation to facilitate increased speed of operation and automatic control. The electric furnace process, which was in its infancy in 1929, now gives a production of over 1 million tons per year, and individual furnaces of up to 60 tons capacity are under construction. The use of coke fired crucible furnaces has been almost entirely superseded by the new design of electric induction furnaces.

SEQ 0178 JOB IOM8627-010-005 PAGE-0178 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

178

Hatfield Memorial Lectures Vol. III

As stated earlier, new installations have revived the basic Bessemer process and production has approached 900,000 tons. It is probable that with the new arrangements for blowing with oxygen, with or without steam or CO2, the production and quality of this type of steel may show marked developments. Rolling Mills Unfortunately, the lack of statistical data will not permit the same type of assessment of rolling mill development. Nevertheless, this development is real. The striking feature is the increased application of electric drive and of continuous mills, the value of which had been so fully demonstrated before 1929 in the collection of mills at Templeborough. Primary mills with a capacity of up to 45,000 tons of ingots per week have been installed, and continuous billet mills with rolling capacity of up to 20,000 tons per week have replaced many of the old reversing type mills with a maximum of about 7000 tons per week. In the heavy structural and rail mills there has been a marked concentration of production, and the proposed continuous beam mill is under construction. Four-high plate mills have been installed for the more accurate rolling of plates, particularly of medium and light thicknesses. In regard to re-rolling, the new continuous mills for the rolling of wire rods are fully adequate to meet the requirements of the wire industry and to provide the material, in coils, required for the new processes developed for the manufacture of nuts, bolts, and rivets. A complete revolution has been effected in these industries. A striking innovation has been made by the introduction of the Sendzimir cold reduction mills for strip of special quality or heavy reductions to thin gauges. The most recent introduction of the Sendzimir hot rolling or planetary mill is still in the development stage. New continuous mills for the hot and cold rolling of strip, and increased supplies of steel in more highly finished form, have stimulated the growth of new cold forming industries, and here, also, marked expansion and developments may be anticipated. The outstanding feature in the rolling mill section is, of course, the development of continuous mills for the hot and cold rolling of sheets and tinplates. The pioneer installation at Ebbw Vale, with a designed capacity of 500,000 tons per year, represented a complete revolution in the manufacturing process, integration from ore and coal to finished product. The success of this project is apparent and the plant is in process of expansion to 1 million ingot tons per year. The second unit, at Shotton, was first installed in conjunction with the cold melting shop, and since the war the plant has been converted into a fully integrated unit and the capacity is also being increased to over 1 million ingot tons. The third unit, the largest, with a completely integrated operation up to the hot rolling stage of plates and coils at Port Talbot, and cold reduction units at Llanelly, Port Talbot and Newport, is already producing at the rate of 1.5 million ingot tons and is in process of

SEQ 0179 JOB IOM8627-010-005 PAGE-0179 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Development in the Industry in Great Britain during the last Twenty-Five Years 179 expansion to a capacity of 2.25 million tons. With the completion of these expansion schemes it is estimated that the requirements of the sheet and tinplate market will be fully met by the modern units and that there will be surplus production available as plates or coils for further processing.

RESEARCH No survey of the developments of this period, however inadequate, could fail to refer to the growing importance and influence of research. Development in this field owes much to the efforts and example of the late Dr Hatfield, to whose memory tribute is paid today. The activating force in his valuable work was not solely an endeavour to discover new facts or new steels, but was primarily the application of scientific facts and the scientific method to the study of works’ problems, and to give a sound, fundamental basis to the different manufacturing operations and processes. Today, this is recognised to be a necessary function of every operating company. One of the most important developments of the period has been the increasing stress laid on the provision of adequate facilities within a works organisation for the carrying out of research, technical supervision, and control of operation, and the technical training of personnel, both staff and operators, required for the successful operation of the powerful and sometimes complicated equipment now available. There is, however, an even broader aspect. Early in his career, Dr Hatfield recognised that no one company could deal adequately with the many problems demanding solution, and in 1924 he undertook the chairmanship of the committee set up by The Iron and Steel Institute for the study of ingot structure. This was the real beginning, within the industry, of co-operative research. Its value was recognised from the outset, and the experience gained during the war years by the research committees under Dr Hatfield, Dr Swinden and others, demonstrated that this work must be established as a fundamental feature of the permanent structure of the industry. The establishment of the central Research Association, with its band of highly skilled investigators, examining both fundamental principles and works’ problems, cannot fail to accelerate progress; and the opportunity which it affords within its structure for regular discussion, by the technical personnel of the works and the central staff, of mutual problems and new discoveries has already proved of inestimable value.

THE FUTURE Although the subject of this review is development over the last twenty-five years, it will be appreciated that development is not a static process. To be effective, it must be activated by a continuous and definite policy, with due regard to both immediate requirements and long term trends.

SEQ 0180 JOB IOM8627-010-005 PAGE-0180 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

180

Hatfield Memorial Lectures Vol. III

More than once some doubts have been expressed as to the wisdom of the expansion at present in hand. Not yet, however, has steel production attained the level at which the steel requirements of the country can be fully met. On the other hand, there can be no doubt that if this country is to maintain its economic stability, and if the demand for a higher standard of living both in this country and overseas is to be met, further expansion is imperative. Competent authorities have estimated that, on the average, provision must be made for an increase of 21⁄2% per annum in British steel production. In the course of ten years this would amount to over 25%, and, taking the forward view, it is not unreasonable to suggest that the steel production of this country should be raised to about 25 million ingot tons by the middle 1960s. It is not to be anticipated that the increase will be uniform over all types of product. So far as can be visualised at the moment, there will be a more marked increase in the requirement of flat rolled products and bars and other forms required by the engineering industry. This prospective increase raises the question of adequate supplies of raw materials. There can be no doubt that new resources of iron ore are available if development is put in hand. This additional ore, in general, is at somewhat greater distances from this country, and additional and improved forms of transport must be provided. For economy, it is essential that this ore shall be carried in the largest practicable type of ore carrier. It is urgent that steps should be taken immediately by the dock authorities for the further improvement of unloading facilities and the provision of berths for ships of deeper draught at at least four of our principal ore reception ports. It is evident that the supply of suitable coal will present some difficulties. Here again, research should be concentrated on methods of attaining greater fuel economy, and a study should be made of the possible extension of the use of fuel oil for heating purposes in steelworks. It will be agreed that one contributing factor to the overall problem is the relation between the steel and gas industries. If the total supply of coal suitable for carbonization is limited, it may well be desirable to consider a further extension of the transfer of coke oven gas from the steel industry’s ovens to the gas industry. The establishment of wider spread gas grids could do much to lower the demand on the available coking coals and, subject to reasonable transfer prices, add to the general overall economy. The location of this additional steelmaking capacity should be determined after careful examination of the vital factors of availability of labour and fuel and the freight costs of assembly of raw materials and despatch of product to market.

All that has been described so far represents merely the mechanism and form of the development. Statistics can never reveal the true man or his motive. The real essence of the development of the last twenty-five years is the creation of a new spirit and of a sense of unity within the industry which never existed previously. The discussion in common (at all levels) of individual difficulties and general problems, the sharing of individual

SEQ 0181 JOB IOM8627-010-005 PAGE-0181 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Development in the Industry in Great Britain during the last Twenty-Five Years 181 discoveries, the joint formulation of requirements and the means to provide them, the readiness to help those in special difficulty, and the willingness to modify individual proposals in the light of the common need, are the factors which alone have made possible the achievement of the common objective and are the real essence of the progress made. So long as this spirit pervades the activities of the steel industry, there can be made no question that the needs of the country will be met and the well being of the industry ensured.

SEQ 0182 JOB IOM8627-010-005 PAGE-0182 SEVENTH LECTURE REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0304 JOB IOM8627-011-006 PAGE-0183 TWENTY-SECOND L REVISED 22DEC04 AT 01:07 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE TWENTY-SECOND HATFIELD MEMORIAL LECTURE

The Place of Mini-Steelworks in the World W. F. Cartwright At the time the lecture was presented Mr Cartwright was Deputy Chairman, British Steel Corporation. The lecture was given in the Firth Hall, University of Sheffield, on 9 December 1971.

About 30% of UK steel production is in the form of light-rolled products and of special steels, i.e. products which can be economic to make on a miniworks scale. Scrap price is a key factor unless pre-reduced iron becomes available at an economic price. The advent of the BOS process has depressed scrap prices in the USA, favouring the development of miniworks. In the UK scrap prices are likely to rise. Scrap quality is also important. Electricity costs might be reduced by a miniworks generating its own power from gas turbines fired with cheap natural gas or reformed gas. Miniworks can be built and commissioned quickly and do not normally carry the high overhead costs borne by large corporations. The conclusion is that miniworks have a useful place and the circumstances where they are preferred can be calculated.

In about 1934 I had the good fortune to listen to Dr Hatfield lecturing in Cardiff. I little thought at that time that anyone would do me the honour of asking me to deliver a Hatfield Memorial Lecture. One thing I have learnt and that is if one has to give a lecture then the study of the subject on which one is to lecture results in one being very much better informed on it. I have enjoyed intensely the preparation of this lecture. From reading the newspapers there is no doubt that most people have been led to believe that the only way to make steel today is in a very large steelworks. A large steelworks has come to mean at least two 10 000 ton/day blast furnaces, three 300 ton BOS vessels and a continuous casting plant feeding, usually, two or three giant rolling mills typified by the new Oita works in Japan which has a strip mill of about 21⁄2 m and a 5 m plate mill. A feature of these new very large steelworks being built now in France, Italy and Japan is that they are flat-product based in their initial stages. This is because the percentage of the total world production of steel required as flat products has been steadily growing, and also because flat product units have grown so big. A modern hot strip mill can produce 6 million tons finished in a year and a modern wideplate mill can produce 3 million tons finished in a year, so that the two together may well require, at full output, 10 million tons of liquid steel a year.

183

SEQ 0184 JOB IOM8627-011-004 PAGE-0184 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

184

Hatfield Memorial Lectures Vol. III

At the other extreme, in the field of bars, rods and light sections, the individual units have not yet exceeded a size of between a half and three-quarters of a million tons per mill. Of course the idea of scale, many studies of which have been made,1,2 is now generally accepted. For instance, a modern power station with 1000 MW sets, which is a very far cry from the 30 MW sets of just before the War, has made such an advance in low capital cost per kilowatt installed, and in technical efficiency, that the actual cost of electricity in real terms has continued to fall though inflation has made an apparent rise. Figure 1 shows how the price of coal has galloped away from the price of electricity as no comparable scale effect has been available in coal mines. Many are the examples of scale effect in heavy industry which can be cited. Figure 2 shows the effect of scale in industry generally. This shows how the profitability of an industrial operation rises sharply at first, as small plants are replaced by larger units, but this rise becomes less steep later. There comes a point, which is ill defined, where the increase in profitability from making a very large plant still larger is quite small, and may be offset by unquantifiable disadvantages such as loss of flexibility. Figure 3 shows the way in which this curve has to be fitted to the huge units of a modern steelworks as they come into production. When a large integrated works is being built up there must be a sharp, if temporary, fall in profitability each time a new major plant unit such as a blast furnace, BOS vessel, or major mill complex is added, until the market can absorb the additional productive capacity. The steelmaker seldom has the good fortune to find a market waiting for him equal to the size of one of these big units. In the case of the Abbey steelworks there was of course the obsolete tinplate industry waiting to be replaced. Recently no such situation has developed except perhaps the death of open hearth plants when new BOS plants come on stream. By contrast the mini-steelworks of the type I am about to consider

Fig. 1 Index of production cost of coal and electricity for the period 1931–1969 for the UK.

SEQ 0185 JOB IOM8627-011-004 PAGE-0185 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Place of Mini-Steelworks in the World

185

Fig. 2 Full capacity profitability related to works annual capacity.

usually consists of two electric arc furnaces using local scrap, sometimes a continuous casting machine and sometimes small up-run ingots, followed by a relatively simple bar or rod mill. Although the addition of another arc furnace, or another small mill, will also affect profitability adversely for a short time (see the curve for miniplants also shown in Fig. 3), the effect is less marked because the market will be able to absorb the extra production very soon.

Fig. 3 Profitability/annual output with production growth.

SEQ 0186 JOB IOM8627-011-004 PAGE-0186 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

186

Hatfield Memorial Lectures Vol. III

The point I am making is best illustrated in Table 1 which gives the minimum capacity of most modern new items of plant units operating today. You will see that for most items you cannot operate on a scale corresponding to less than several million tons of steel a year, but for arc furnaces and mills making light rolled products, works making much less than half a million tons can be competitive by modern standards. Table 1

Minimum capacity of various modern items of new plant units presently operating.

Plant unit Ore terminal 2 × 150 000 ton ships/w Sinter strand 14 000 tons/day Blast furnace 2 × 12 000 tons/day BOS vessel 2 × 300 tons/40 min Electric arc 2 × 100 tons/3 h Wide strip mill Heavy or medium section mill Heavy or medium plate mill Rod mill Bar or light section mill

Capacity, tons/year

Corresponding crude liquid-steel capacity, tons/year

4 500 000

3.5–4 000 000

7 000 000 6 000 000 500 000 5 000 000 1 000 000 2 500 000 400 000 450 000

6 000 000 500 000 6–6 500 000 1 300 000 3 500 000 500 000 580 000

I first encountered this idea of a small steelworks about 20 years ago in Venezuela. Scrap was an unsaleable commodity in the capital city, Caracas, so an entrepreneur installed two electric arc furnaces and, using small up-run ingots, rolled the output into reinforcing bars with a very ready local sale. Today anyone driving around Tokyo and Tokyo Bay will see similar small plants each with two 50 ton arc furnaces, casting into up-run ingots which are then rolled out on a fairly simple bar mill. Such works make a habit of melting by night, when power is cheap, and rolling by day, and I believe in certain cases, collect their own scrap. While the big new steelworks in Europe and Japan have been making the headlines they have been actually outnumbered, although not surpassed in tonnage, by the mini-steelworks which have grown up in the USA and Japan. Figure 4 shows the distribution of ministeelworks by size in the USA.3 There have been more than 40 such works built in the USA in the last ten years. Figure 5 shows the comparatively neat and tidy appearance of these modern mini-mills, occupying as they do, a completely roofed in area of quite small size. To determine the likely percentage of steel to be made in mini-works of the future I have examined the trend in percentage of steel by product (Fig. 6). For the foreseeable future it is likely that almost all flat products, heavy sections and rails will be manufactured exclusively by the blast furnace–BOS route but alloy steels, bars, rods and light sections may be manufactured to an increasing extent in relatively small arc plants. Figure 7 shows that it would be possible for over 30% of bars, rods, light sections and alloys to be made in a small works relying on arc furnaces, but of course a considerable percentage will in fact continue to be made by the blast furnace–BOS route in which part of the production will be billets for transportation to re-rolling plants, as is at present so common.

SEQ 0256 JOB IOM8627-011-005 PAGE-0187 TWENTY-SECOND L REVISED 20DEC04 AT 21:58 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Place of Mini-Steelworks in the World

187

Fig. 4 Miniplants/capacity for USA, January 1971: maximum 4 000 000 tons/year.

Fig. 5 A modern mini-mill in the USA.

So far the key to the construction of miniplants has been a guaranteed local market for the product coupled with a highly reliable source of local scrap, and these market based, scrap based steelworks have now become the subject of worldwide discussion. Very little has been published showing the comparative costs of a product suitable for a mini-plant, delivered to the customer by the three possible routes:

SEQ 0188 JOB IOM8627-011-004 PAGE-0188 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

188

Hatfield Memorial Lectures Vol. III

Fig. 6 Percentage of UK steel in various categories of products.

Fig. 7 Percentage of UK steel production made in integrated works and suitable for miniworks, 1953–70.

(i)

the traditional route, making billets in a big coastal based steelworks and sending them to a number of re-rolling plants based on the market (ii) processing steel through to the finished bar or rod on the coast and shipping the product to the market (iii) manufacturing bars, rods and light sections in small steelworks using only local scrap or, where coast based, some pre-reduced pellets, the total finished product being delivered to customers within a very small radius.

SEQ 0189 JOB IOM8627-011-004 PAGE-0189 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Place of Mini-Steelworks in the World

189

The alternative of building a small coke oven and blast furnace plant based on a small market is no longer viable, although such existing plants continue to survive in many areas. It is generally considered that the difference in conversion cost between liquid steel made by the blast furnace–BOS route and made by the electric arc route is at least £3.5/ ton in favour of the blast furnace–BOS route, due to the high cost of electricity and electrodes compared with the very simple process of blowing oxygen on to liquid pig iron. But this difference in conversion cost is only one of many factors which must be examined and evaluated. The first and probably most important question is that of total transport cost. Figure 8 gives rough figures of transport costs for the main raw materials and finished products, although these costs must not be taken as values for any particular situation. If a market for, say, reinforced concrete bars exists in an area where scrap is readily available then a mini-plant built in that area will be at an advantage in terms of total transport costs over either the route of billets sent to a re-roller in a market area, or the more modern route of finished material being sent from the big steelworks distant from the market. For example, a mini-works based in London would incur a cost on scrap delivery of about 50p and on product delivery of about 50p, a total of, say, £1. On the other hand a steelworks 160 miles away would incur a cost on that same scrap of about £2, and on the product of about £2.2, but taking into account yield, call it a total of £4, or a difference between the two transport costs of some £3. The difference in these total transport costs varies widely between mini-works. This diagram also shows the high cost of transport on small lots of ore, showing why it is unlikely that there will be many inland arc plants based on prereduced pellets. The next most important question which affects the decision to install a mini-works is the price and availability of raw materials. Britain is so accustomed to a controlled scrap

Fig. 8

Notional transport costs/distance for finished products and main raw materials: not to be taken as actual.

SEQ 0190 JOB IOM8627-011-004 PAGE-0190 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

190

Hatfield Memorial Lectures Vol. III

market that we do not know what the effect of a free market will be on scrap price when we enter the Common Market. Figure 9 shows how the prices of scrap and pig iron have moved in the USA over a period of years, together with the parallel movement of steel production. Until the introduction of the BOS process, scrap was in great demand by the open hearth plants which used up to 50%, and the price of scrap tended to follow the demand for steel. With the introduction of the BOS process in about 1960, there was a sharp decline in the demand for scrap by the big integrated works, going as they did from around 50% of the steelworks burden to 25–30%, and it will be seen that the price of scrap is relatively stable even in spite of the sharp increase in steel production.

Fig. 9 US scrap and pig iron prices/steel production, 1950–1969.

In a country like Great Britain where one Corporation dominates the BOS plants, it is within the power of that Corporation to change the percentage of scrap used in its BOS plants. Figure 10 shows the tonnage of scrap likely to be available at various rates of steel production, i.e. the various percentages of works scrap, processed scrap and capital scrap which will be available; the balance has to be made up from iron. In a typical flat product BOS plant the scrap arising is about 25%; if such a BOS plant was on 25% scrap then under those circumstances the works would be neither receiving nor delivering scrap. A BOS plant manufacturing heavy sections which produce less arising scrap might in fact, at 25% scrap, have to bring in some bought scrap. BOS plants manufacturing flat products have to be so careful about nickel and copper content that they are very unwilling to use any scrap which is not of pedigree variety. Such works will not willingly use any material such as destructor scrap, motorcar scrap, or indeed any scrap of which the works is unaware of the origin. Some electric-arc plants manufacturing, say, cold headed steel, will be equally fussy about the content of nickel and copper

SEQ 0191 JOB IOM8627-011-004 PAGE-0191 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Place of Mini-Steelworks in the World

191

Fig. 10 Iron and scrap requirements at various steelmaking capacities.

but if they were adjacent to a motorcar pressworks, they would be able to obtain all their requirements of low residual scrap from this source. Motorcar pressworks continue to make between 37 and 43% of scrap/ton of sheets processed, so that such works are a major possible source of pedigree alloy free scrap for an arc plant. At the other extreme, a miniplant based, say, in London, and making reinforcing bars, can afford to use scrap containing small amounts of lead, tin, copper and nickel which is commonly found in scrap coming from motorcar disintegrating plants and may even be able to use some destructor scrap. At this stage it is difficult to guess the relative price of high grade pedigree presswork bundles and low grade disintegrator scrap when we enter the Common Market, but a big effect on the price will be the location of the nearest arc plant capable of using local arising scrap. Figure 11 shows how the relative proportions of BOS capacity and arc furnace capacity for a given amount of scrap would be affected by changing the percentage of scrap used in the BOS vessels. If we assume that the maximum amount of scrap used in all BOS plants averaged 35%, which is probably rather too high a figure, then with a steel production of, say, 30 million tons/year, and the amounts of scrap that Fig. 10 shows as likely to be available, some 25 million tons could be made by the BOS process, leaving 5 million tons for arc furnaces. Conversely, if the scrap percentage in BOS plants were reduced to an average of 15%, then out of 30 million tons of steel only 20 million tons would be BOS steel, 10 million tons would be electric arc. This can be expressed in another way, as in Fig. 12. In other words, if we change the percentage of scrap used in the BOS plant from 30 to 20% the arc furnace capacity in the UK could be increased from about 20% to around 30% of the total steel. In Great Britain the normal average percentage of scrap used in BOS plants is about 26% but it is known

SEQ 0192 JOB IOM8627-011-004 PAGE-0192 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

192

Hatfield Memorial Lectures Vol. III

Fig. 11

BOS–arc furnace capacities, using extreme BOS scrap ranges, related to UK capacity. Percentages of scrap refer to those used in BOS process.

that at Sakai, a large Japanese steelworks, they have recently been running at an average of 15%. In the UK it has been found that when making certain steels in the BOS plant a low percentage of scrap is desirable, but it does entail the construction of plant to ensure a steady feed of high grade iron ore pellets, otherwise foaming takes place.

Fig. 12

Percentage of possible arc furnace capacity related to percentage of scrap used in BOS plant mixture.

SEQ 0193 JOB IOM8627-011-004 PAGE-0193 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Place of Mini-Steelworks in the World

193

A further possibility, much discussed, for controlling the price and availability of scrap, and particularly for alloy free scrap is the use of pre-reduced pellets made from low silica high iron ores: such as those specially processed in the Labrador area, or the pellets made from magnetically separated ore from Sweden. Where natural gas is freely available at very low prices pre-reduced pellets can be manufactured by the HyL, Midrex, Purofer and other processes at a price low enough, when used as an arc furnace feed, to be considered competitive with the blast furnace– BOS route. Such plants are under construction in Venezuela, Canada, Brazil, Germany and the USA, and in Canada using natural gas; also in Brazil and Germany using gas made from naphtha. These systems seem to be reliable, controllable and produce a pre-reduced pellet of about 93% iron when using the highest grade of iron ore pellet. Naturally, with proper selection of ores the resultant steel, when made from these pellets plus pedgree scrap, can be of excellent quality for cold heading or other purposes requiring freedom from copper and other deleterious alloys. Thus a mini-works which is in an area where scrap is readily available, because of the expense of transport of the scrap to the nearest integrated works, will probably use scrap only. But a works such as that now being constructed on the St. Lawrence, where scrap is relatively scarce, but very high grade ore and cheap natural gas is readily available, may use pre-reduced pellets to quite a high percentage. Figure 13 shows that in the USA the lack of availability of very cheap natural gas or exceptionally high grade ore will probably restrict the development of the pre-reduced pellet route.4 This confirms the view I expressed in Tokyo last year5 that in most large steelmaking countries the blast furnace– BOS route will remain dominant for a long time to come. Since the cost of electric power and electrodes are the main causes of the conversion costs of the arc furnace route being high when compared with the blast furnace–BOS

Fig. 13 US forecasts of percentage change of steelmaking mix, 1970–85.

SEQ 0194 JOB IOM8627-011-004 PAGE-0194 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

194

Hatfield Memorial Lectures Vol. III

route, many studies have naturally been made to see whether these two costs cannot be reduced. The present consumption of electrodes is about 12 lb/ton. It is now believed that this consumption can be reduced to 8–9 lb/ton by use of the electrode coating process which, while it originated in Bulgaria, is being currently developed by the British Steel Corporation. There is, however, less hope of being able to reduce electrode consumption much below this figure, In endeavours to reduce the amount of electrical energy used, preheating of scrap, continuous charging of pre-reduced pellets and the use of very large furnaces have all been explored, but the total reduction in kWh/ton of liquid steel has not been very great. Even common steel requires a consumption of about 450 kWh/ton, and with typical electricity costs in Great Britain, this may well be equal to £2/ton and with the foreseeable trends of the price of electric power generated by the CEGB in Great Britain, this cost of £2 is unlikely to be reduced. Various other ideas have been explored. For instance, siting a steelworks adjacent to such a huge scheme as the Churchill Falls with its exceptionally low cost of power, 0.1p/ kW h, would reduce the cost of power per ton of steel, but the transport costs on both raw materials and finished product would more than wipe out the saving. It is still cheaper to convey electricity over a considerable distance rather than transport raw materials to the hydroelectric station and the finished products away. But one idea has received considerable thought: in many areas of the world today natural gas is readily available, and in some areas, e.g. Canada, it would be possible to install a simple gas turbine, directly coupled to an arc furnace, the waste gas from which being used to generate steam in waste gas boilers and then in steam turbines generating power for use in the mills. By avoiding any electrical connexion between the gas turbine generator and the waste heat boiler generator the wild fluctuations of voltage caused by the arc furnace would be avoided on the mill system. But to date neither American nor British gas turbine manufacturers feel that they can supply a gas turbine generator and transformer which could stand the wildly fluctuating loads of an arc furnace when melting down. In areas where natural gas is very cheap such an idea is extremely attractive, eliminating as it would the problem of flicker on the power supply and producing, if natural gas was at a very low figure, a low cost of power to the arc plant. To make this idea succeed what appears to be required is some form of electrical or mechanical flywheel which would take the wild fluctuations of load off the gas turbine. If such an arc plant was integrated with a pre-reduced plant of the Midrex, Purofer or other gas reduction type, then it can be imagined what an attractive idea it would be for natural gas to be used to pre-reduce the ore, drive the gas turbine and power the mills. Perhaps the problems involved will be solved before long. While using the term ‘natural gas’ it must be pointed out that already there are various other alternatives. Very large scale plants for the manufacture of natural gas equivalent from naphtha and other oil derivatives are now being designed and one of the prereduced pellet plants is to use a smaller gasifying plant integrated with the works.

SEQ 0195 JOB IOM8627-011-004 PAGE-0195 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Place of Mini-Steelworks in the World

195

The use of liquified natural gas conveyed by tankers at the moment produces a fuel cost which is too high. The next item which I will deal with when comparing the economics of the miniplant with the big plants is the employment cost. A major steelworks requires a large number of ancillary workers: such a plant is usually equipped with central engineering shops, transport divisions, mobile task forces, personnel departments, research and development, canteen and other non-productive staffs. In a small plant, such as will be seen in Japan, it is almost certain that there will be found many Jacks-of-all-trade turning their hands to anything. It is almost certain that there will be virtually no personnel department, no training department, no research labs, so that on paper it can be expected that a ministeelworks will have lower non-productive labour costs compared with a large works, due in effect to many specialist functions being carried out in the larger steelworks from which the mini-steelworks will ultimately profit. The running cost of making steel is made up of raw materials, energy, employment costs and yield. It is difficult to prove that there will be any material difference in the yield from liquid steel through to finished product against original order, between a well run blast furnace–BOS plant and an equally well run small arc plant, but it is felt by some that it is better to have one control over quality from start to finish than a split control which occurs under the traditional system of manufacturing billets in a blast furnace–BOS plant distant from the ultimate re-roller. I would also say that, due to the close supervision which is possible in well run small arc plants, it is easier to obtain the highest possible yield of prime product against order than in large blast furnace–BOS plants making the same product, with its wider span of command. At the bottom of any cost sheet is an item for overheads. Already I have mentioned the question of overheads on the personnel side but central research and development, advertising, overseas agents, charitable contributions and items of this nature will not figure in many of the mini-works devoted perhaps to making reinforced concrete bars only for a very limited market. Such mini-works normally export nothing and contribute but very little to research and development of the steel industry as a whole, though in one or two notable cases this is not true. Certainly big contributions have been made by miniworks in the study of pre-reduced pellets and in the operation of relatively small but efficient continuous casting machines. The difference between the ‘Buccaneer’ works typified by those small works in Tokyo and the large works manufacturing identical products but part of a huge complex, in overhead charges, is extremely difficult to estimate, but it may well be that it is something in excess of £2/ton. I think this whole attitude of mind was summed up by the overseas manager of a large Continental steel company who told me that when organising a mini-works overseas he had to be very careful that this works was completely divorced from the ‘overheads’ structure of his central company management in Europe, not only in terms of personnel, research and development, advertising and the like, but also in design. His mini-works,

SEQ 0196 JOB IOM8627-011-004 PAGE-0196 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

196

Hatfield Memorial Lectures Vol. III

he said, would have different engineering standards and would pay less for their equipment than if the European based central engineering department designed everything for them. We now come to a major factor in the comparison of cost which does not figure on an ordinary works cost sheet, viz. the capital cost of construction, commissioning, and the time from completion of construction to reaching final designed output, as well as the cost of working capital. It is generally considered that to build a large blast furnace–BOS steelworks processing down to such finished products as cold reduced sheet, bars and rods, the capital cost is between £60 and £100/ton, though new American figures indicate a higher figure in the USA. In recent years most mini-plants have been built for about £30/ton, from scrap to finished bar. Big blast furnace–BOS plants can take from three to four years to complete stage one from start of expenditure. Most miniplants are built in one and a half to two years. There is a wide variation in the time from completion of construction of the first stage of a blast furnace–BOS plant to bringing it to full output. Where it is a greenfield site it is seldom that such plants reach their full first stage output in under one year from commissioning. A typical miniplant would expect to reach full output in well under one year from commissioning. One reason for the saving in time of the miniplant is that the builder of such a plant will not be reaching out into the unknown. He will almost certainly order two arc furnaces of a type already in regular operation: his continuous casting plant will have very little degree of experimentation built into it, as will his roughing mill or bar mill. Thus the time to place the order to build the miniplant will be at a minimum and since the whole plant is so small, as shown by Fig. 5, the time to agree location, purchase of land, ground loadings and all the points which have to be agreed before the construction of a large integrated works can start, also will be months less, maybe years, for the miniplant. These are both important factors during a time of inflation. When building a large modern steelworks intended to last for many years, inevitably the builders will be asked ‘are you certain you are not building yesterday’s plant? Be sure you are building the finest in the world when it is complete and certainly better than anything presently in existence’, this is a very laudable ambition but it is difficult to achieve and almost certain to build-in new and unknown problems which may take time to solve. There is also the question of capital employed in the business. If a billet is produced at a blast furnace–BOS plant based works and shipped a considerable distance to a re-roller, or even from a new steelworks in Australia, say, to a re-rolling plant in Cardiff, then the working capital involved will be greater compared with that of a mini-steelworks processing direct from scrap to finished bar because of the need to hold a large variety of stocks at both despatching and receiving works. Similarly in the field of bars, rods and light sections, it must be evident that the miniworks based on the local market, with no export, is going to have a lower working capital than its big brother shipping bars long distances.

SEQ 0197 JOB IOM8627-011-004 PAGE-0197 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Place of Mini-Steelworks in the World

197

Figure 14 shows the calculated comparative returns on investment of the miniplant compared with the blast furnace–BOS plant on a straight return basis for varying scrap prices. This indicates that if scrap were to remain around its present price in relation to pig iron the mini-works based on scrap would be a more profitable proposition than the large integrated works up to the limits imposed by the availability of scrap. However, we are going to a ‘free market’ price for scrap in future and, obviously, in such circumstances the price of scrap will move upwards to an equilibrium price which will produce a situation more akin to that in the USA where large integrated works are the right answer in some circumstances, and the mini-works based on arc furnaces the right answer in others. Figure 14 can be only an indication, for of course one cannot exactly compare the local scrap local delivery mini-works using low-grade scrap into a low priced product such as reinforcing bars, with the integrated works selling a wide variety of products at different prices, but it is almost inevitable that the integrated works will have to use high grade material on occasions for low grade orders.

Fig. 14

Comparative investment returns mini/integrated works producing re-rolling billets.

Figure 15 brings in the ‘time of construction’ effect by making the comparison on a DCF basis; again it must be borne in mind that the miniplant may pay a lower price for scrap than the large integrated plant. Under favourable circumstances it is clear that in the field of bars, rods and light sections the miniplant is a serious challenge to the blast furnace–BOS plant right through to the finished bar or rod. I would not claim that Figs. 14 and 15 give precise guidance on which plant to use. On certain types of specialist marine chart they write ‘not to be used for navigation’ and this, I think, applies. Selection of the plant and the process to use is like sailing in waters where the sandbanks are constantly shifting and new charts are needed before every voyage. The figures I have used today are merely an indication of

SEQ 0198 JOB IOM8627-011-004 PAGE-0198 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

198

Hatfield Memorial Lectures Vol. III

Fig. 15

Per cent DCF returns on integrated works and mini-works under various conditions.

the way in which I think things are moving. I think we can expect throughout the world an increase in the number of miniplants manufacturing bars, rods and light sections, sometimes in relatively common grade steel for reinforcing bars and fencing wire, and sometimes in bars of the very highest quality for use in the engineering industry. It will depend on the availability of scrap, the quality of the scrap and the size and type of the local market. A coastal mini-works adjacent to a supply of natural gas may be suitable for supplementing the supply of scrap with pre-reduced pellets but I do not think, as I mentioned in connexion with total transport costs, that you can expect to see prereduced pellet plants used in miniplants far from the sea.

CONCLUSION To sum up, there is a clear place in the world for miniplants and the circumstances under which such plants can compete effectively with large integrated works are known and calculable. One remaining uncertainty is the future price and availability of pre-reduced iron to reinforce scrap supplies. The present rapid growth of miniplants in USA and elsewhere is not an inexplicable development but can be explained logically according to the principles I have talked about, and Table 2 summarises the conditions under which it might be expected that the miniplant has the advantage over the integrated plant, or vice versa. The idea of electric furnaces being used to make cheap reinforcing bars or common light rolled products would have been strange to Dr Hatfield, during whose lifetime arc furnaces were limited to making relatively expensive special steels. However, had someone come along with such a proposition, well thought out and with supporting figures to

SEQ 0199 JOB IOM8627-011-004 PAGE-0199 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

The Place of Mini-Steelworks in the World Table 2

199

Summary of advantages and disadvantages of miniplant v. integrated plant.

Large integrated works Advantages Higher eventual profitability Wider choice of materials Disadvantages Heavy capital investment Slow return on capital Higher transport costs

Miniworks Uses simple standard equipment Early return on capital Low transport costs Flexibility in meeting orders Depends on scrap as raw material Sensitive to power costs Limited range of products

back it, he would have shown a lively interest, and it is on this note that I want to end, viz. that the place of the miniworks in the world rests on its economics in any individual set of circumstances. There is no general or simple answer.

ACKNOWLEDGEMENTS I would like to thank my colleagues in the British Steel Corporation who provided much of the data used, in particular R. Scholey for making available an internal paper on which Figs. 10 and 11 are based, and Dr A. H. Leckie for general assistance with the preparation of the lecture. I would also like to thank Dr W. Rizk, Managing Director, GEC-English Electric Gas Turbines Ltd, for helpful discussions on the possibility of using gas turbines to generate cheaper electricity for arc furnaces, R. J. Dain, engineering consultant, also for ideas on the use of gas turbines in an electric-arc works, and the US Steel Corporation for Fig. 5.

REFERENCES 1. 2. 3. 4. 5.

A. H. LECKIE and A. J. MORRIS: JISI, 1968, 206, 442. A. R. PARISH et al.: JISI, 1971, 209, 420. ‘‘33’’ Magazine, March 1971, 57. ‘‘33 Magazine, June 1971, 38. W. F. CARTWRIGHT: JISI, 1971, 209, 89.

SEQ 0200 JOB IOM8627-011-004 PAGE-0200 TWENTY-SECOND L REVISED 19DEC04 AT 21:01 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0301 JOB IOM8627-012-005 PAGE-0201 TWENTY-FIFTH LEC REVISED 22DEC04 AT 00:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE TWENTY-FIFTH HATFIELD MEMORIAL LECTURE

Materials and Malthus Sir Alan Cottrell At the time the lecture was presented Sir Alan Cottrell was Master of Jesus College, Cambridge.

Sir Alan Cottrell’s Hatfield Memorial Lecture, the 25th, with the above title, was presented at the University of Sheffield in December 1974. No lecture manuscript is now available, but the following brief thematic review of the lecture was published in the March 1975 issue of Metals and Materials under the modified title: Of steel and self-sufficiency

In giving the 25th Hatfield Memorial Lecture at the University of Sheffield, Sir Alan Cottrell, Master of Jesus College, Cambridge, said that he was doubly honoured, firstly to be delivering his lecture in Sheffield, the cradle of modern metallurgy, and secondly in being asked to do so in the name of Dr Hatfield, whose ideas reach out to us still, reminding us that he was a true leader and pioneer of applied science.

PROBLEMS OF THE AGE The 1920s and ’30s, in which Hatfield worked, were like today, a time of great anxiety for the future of our country. Hatfield applied himself to developing new and improved materials, as urgently needed then as today. But today’s problems appear even more intractable. In Sir Alan’s view they stem from two central causes: an impending world shortage of materials and a global struggle between two opposing socio-political ideologies. This struggle, which has already erupted into civil war in some parts of the world, is the biggest single contributor to Britain’s economic disappointments since World War Two. Our problems seem worse now than in Dr Hatfield’s day, when the issues were at least clear cut, making it easier for governments to decide upon appropriate action. Nowadays the problems are more diffuse and general in character, more intrinsic and enveloping. The problem of dwindling material resources is at least open to scientific attack, but how can public willpower be harnessed when the world is obsessed with rival political philosophies? Recent UN conferences on World Food and World Population have been ruined by this rivalry.

201

SEQ 0202 JOB IOM8627-012-003 PAGE-0202 TWENTY-FIFTH LEC REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

202

Hatfield Memorial Lectures Vol. III

Doctrinaires of Left and Right are squabbling while the world starves. Although the fate imagined for us by Thomas Malthus, an 18th Century Fellow of Sir Alan’s own Cambridge college, seems inescapable, the world has the means to produce enough food for the next 30 years, albeit this would entail a three- or four-fold increase in production.This pre-supposes intensive agriculture all over the world, on the same scale as in the developed areas. It would mean financial aid of the order of £3000 million, plus greatly increased supplies of energy, water and machinery. To raise the use of nitrogen fertiliser alone to the required level would absorb one fifth of the world’s present energy production. The agricultural revolution needed would necessitate large scale migrations to sparcely populated tropical regions, which seem unlikely to occur under the present political climate. For Britain’s survival, a massive increase in food production is the first priority, not only for our own sakes, but also to ease the demand on world resources.

METALS DEMAND DOUBLES IN 20 YEARS The world demand for food, Sir Alan reminded us, doubles every 25 years, the demand for metals every 20 years, for energy every 15 years. Consumption of aluminium has been growing at the rate of 10% a year; steel consumption increased by 44% per person between 1957 and 1967, representing 12% in the case of the United States, but 270% for Japan. The rarer non-ferrous metals, such as mercury and tin, will not be with us for many more years. Even the more common ones are in a doubtful position, because there are no low grade deposits to turn to when the present ore bodies are exhausted. Copper and nickel are in a much healthier state, with larger reserves opening up as the minimum working grade is reduced. Thus there appear to be several decades of supplies ahead for these materials as extraction techniques improve. Bauxite supplies are still good, and will last until the end of the century, after which we will have to turn to shale and igneous rock for our sources of aluminium. In both cases, though, winning the metal is at the expense of vastly higher energy consumption. For iron the outlook is much happier. Here the trend is towards increasing the grade of ore used, a trend made possible by the discovery of vast deposits which geologists calculate should last for centuries at present rates of use. It is time, therefore, to start thinking of steel as a substitute for less plentiful materials.

ENERGY – THE VITAL RESOURCE Energy is the one fundamental resource; with it, all material problems are soluble, without it, none are. And it is truly non-renewable. Given the energy and the ‘know how’, we can regroup our atoms and molecules to make foodstuffs, separate pure water from sea water or effluent, convert oil into plastics, even extract minerals from the sea.

SEQ 0725 JOB IOM8627-012-006 PAGE-0203 TWENTY-FIFTH LEC REVISED 06FEB05 AT 18:52 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Materials and Malthus

203

But with energy in short supply, our industrial and social activities will be in competition for it, and energy is becoming scarcer and more expensive while the demand for it grows. Modern agriculture is increasingly energy intensive; a third of the energy in America’s corn crop, for example, is supplied from fuel. The desalination of water, too, is highly energy intensive, and the trend will inevitably increase in mineral processing industries. During the 1950s mechanical power in the United States metal mines increased four-fold for no increase in output (Fig. 1). The metallurgical industries consume energy far more intensively than industry in general, energy accounting for 20% of the cost of steel compared with a 4% average for manufactured goods. The UK’s iron and steel industry used 11% of the country’s energy output in 1972, and some 85% of the total energy consumed in our entire metals industry. Other metals are even more energy-intensive: the high affinity of aluminium

Fig. 1

Winning metalliferous ores demands ever more costly machinery and expenditure of energy.

SEQ 0204 JOB IOM8627-012-003 PAGE-0204 TWENTY-FIFTH LEC REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

204

Hatfield Memorial Lectures Vol. III

for oxygen, for example, and the low thermal efficiency in the production of copper (due to the low metal content of the ore), results in a given amount of energy producing much more steel than any other major metal. The same comparison holds good for steel and plastics, with the added factor that plastics production involves the consumption of oil as a main constituent. Furthermore, continuous progress in steelmaking technology is steadily increasing the output per unit of energy. Looking into the future, Sir Alan foresees a more efficient process than the blast furnace, in which direct reduction will change the oxide into solid iron by means of a stream of reducing gas in a simple heating chamber at a temperature of 800°C. Some thought has already been given to direct reduction by means of nuclear power, though nuclear iron making will not contribute significantly yet awhile. On the other hand, worldwide plans for future reliance on nuclear power should make electricity relatively abundant, and the implications of this should be studied. Today, such recent developments as the replacement of open hearth furnaces by the basic oxygen LD process, and the introduction of continuous casting, are already helping to save energy and hydrocarbon fuels.

THE CHALLENGE FOR BRITAIN The world resources picture, then, ranges from bleak for food to quite bright for iron and steel. However, the iron and steel industry is a heavy consumer of the fundamental resource, energy, of which we shall be very lucky to find anything like the amount we shall need. Huge new oil fields will have to be found, as will new uranium finds and/or strikingly successful development of the fast breeder reactor. Above all, the investment of capital on a vast scale is vital for the setting up of new power industries. What of Britain’s materials transactions with the rest of the world? International business is largely the exchange of materials, goods and services between countries. Our traditional strategy of exporting far less than we import, and balancing this by exporting manufactured goods of high added value, is no longer viable in a world of dwindling natural resources. We must make ourselves more self-sufficient in food and materials. Sir Alan sees the imbalance of cash flow as merely an expression of the imbalance of material flow in and out of the country. Tropical produce accounts for a third of our food imports, and it should be possible to produce the rest ourselves, assuming some changes in our eating habits. We already provide almost all of our own barley, oats, pork, poultry and dairy produce and 80% of our beef, and our agricultural output equals the combined outputs of Australia and New Zealand. The spectacular growth of Britain’s agriculture is due to higher technical and economic efficiency at a time when the work force has dropped and 50,000 acres of good farmland is being lost annually to road building, housing and the like. The demand for animal protein and imported cattle food can be reduced by the manufacture of factory made protein, possibly from micro-organisms in mineral oil waste,

SEQ 0726 JOB IOM8627-012-006 PAGE-0205 TWENTY-FIFTH LEC REVISED 06FEB05 AT 18:52 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Materials and Malthus

205

and also by reducing our excessive consumption of animal protein and substituting vegetable protein, perhaps in the form of artificial meat. Several years will be needed to reduce our imports of oil, our next biggest material import after food. With North Sea oil, good coal deposits and an advanced position in nuclear technology, the outlook can be bright for some decades ahead, but this hinges upon, respectively, a more advanced coal technology and development of the fast breeder reactor. In the field of non-ferrous metals, despite Britain’s one-time pre-eminence in this sector, we must remain highly dependent on imports. Globally the picture looks grim indeed. The actions being taken to ensure that the people of the world do not starve appear quite inadequate. On the contrary, we seem bent upon proving Malthus right before the end of the century. Shortage of food, backed up by shortage of energy, are almost guaranteed to bring this about. Greatly increased prices will make it impossible for the poorer nations, and difficult even for the richer ones, to feed themselves. Britain’s problems are different. We can clearly find ways to escape the Malthusian fate; but can we muster the national awareness, social will and political cohesion to follow them?

SEQ 0206 JOB IOM8627-012-003 PAGE-0206 TWENTY-FIFTH LEC REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0258 JOB IOM8627-013-012 PAGE-0207 207-226 35 SCHOL REVISED 20DEC04 AT 22:10 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE THIRTY-FIFTH HATFIELD MEMORIAL LECTURE

European Steel: What Future? Sir Robert Scholey At the time the lecture was presented Sir Robert Scholey was Chairman of the British Steel Corporation. The lecture was given at Sheffield University on 8th December 1987.

I am very conscious of the privilege of delivering the Hatfield Memorial Lecture for 1987, and feel this the more keenly as I look through the list of previous lecturers. So far as the approach adopted in the past is concerned, I note that many lecturers have treated subjects of a fairly specialized character in a detailed manner, whereas some others have handled a broad and strategic canvass. It is the second approach that I intend to adopt. I am encouraged in this by the thought that William Hatfield, who made such a distinguished contribution both directly to the steel industry in Sheffield and through professional bodies and in scientific papers, began his experience in industry in the works of Sir Henry Bessemer and Co.

The steel scene changed very greatly between the time of Bessemer’s pioneering work and William Hatfield’s period in the industry during the first four decades of this century. It has changed again very much between the time of his death 44 years ago and now; and nothing is more sure than that it will continue to change into the future. We have been very preoccupied in the steel industry for the past decade and more with meeting the immediate acute challenges that have come upon us, seeking to survive, almost on a day to day basis. Here in Sheffield we are very conscious of that experience, for the steel industry landscape here has changed in a very radical fashion over those years. This evening, however, I would like us to lift our eyes to the more distant horizon and ask ourselves a fundamental and very concerning question: namely, in the light of changes in the world steel industry, including particularly the growth of production in developing countries, does the steel industry in Europe realistically have a future? I shall first briefly set the overall steel scene as I see it, and then go on to suggest certain advantages and disadvantages in steel production experienced by developing and industrialized countries respectively, before reaching a conclusion and seeking to draw a few lessons. Looking back from the perspective of today, it is evident that the history of the steel industry in Europe has generally been a rather troubled one. The halcyon days of the decade and a half following the Second World War, when many of us were growing up in the industry, can now be seen to have been the exception rather than the norm.

207

SEQ 0208 JOB IOM8627-013-011 PAGE-0208 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

208

Hatfield Memorial Lectures Vol. III

Fig. 1 Average growth rates of world steel production.

Figure 1 illustrates this from average growth rates in world steel production in each of the decades of this century. The post-war period was certainly one of major change both in the international sphere and in the steel industry. At the world level, the European countries, not just Britain, but France, Belgium, and The Netherlands too, progressively relinquished their imperial links with the potential implications that had for the thrust of their economic and commercial activities. However, at the same time the European Communities were being created and developed: first the pilot European Coal and Steel Community established by the Treaty of Paris of 1951, and then the wider European Economic Community established by the Rome Treaty of 1957. The dynamism in trade which these Communities encouraged in those years substituted for the diminishing significance of the former imperial links. In the steel industry too, major changes took place which are sometimes overlooked nowadays as a result of our preoccupation with the traumas of the past dozen years. Perhaps most dramatic was the rationalization on the strip products side where, in the UK alone, a multitude of hand cold mills and tinplate lines were reduced at that time to five cold rolling plants, and three tinplate works, in South Wales. The main point about this is that all the changes that took place during those years were eased by a background of strong economic expansion. During the 1960s the position in that respect began to waver, in cycles, and just after the UK joined the European Communities the first oil crisis broke. Since then the experience of manufacturing business, and of steel even more acutely, has ranged between uncertainty at best and severe recession at worst. Figure 2 shows the development of apparent steel consumption in ‘Western’ industrialized countries, in developing countries, and in the world as a whole from 1960 to 1986. Certainly this autumn’s turmoil in the financial markets and, lying behind that, the huge budgetary and trade deficits of the United States, the associated fears of recession in

SEQ 0209 JOB IOM8627-013-011 PAGE-0209 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future?

209

Fig. 2 Apparent steel consumption, 1960–1986.

the American economy, and the continuing fall in the international value of the dollar, with all that these financial and economic factors could mean for Europe, are discouraging to confidence and stability. But the biggest long term challenge, and the area I intend to concentrate on, is the expansion of steelmaking in developing countries, and the effect of this on traditional producing areas, especially on Europe. Figure 3 is a map of the world showing: communist countries, developing countries, and traditional steel producing countries. This expansion of developing countries’ steel production was long anticipated: certainly for at least a quarter of a century. In the event, for reasons of political instability, difficulty of access in many cases to enough finance, and problems with the practical application of technology, the challenge was later in making a broad impact than once seemed likely. When it did arise on a really significant scale, it coincided with the general recession in world steel business following the first oil crisis. This meant that the competition from new or expanding overseas producers was felt more acutely in traditional steelmaking areas than if it had happened in the earlier generally expansive period. Moreover, the pressures created by recession and Third World competition are ones to which the traditional producers, including very much the Europeans, are finding it extremely hard to adjust. In summary, while world total steel production and demand have remained static over the years 1974 to the present, the share of that constant production claimed by the large traditional producers, by which I mean the United States, the European Community, and Japan, has declined from almost 60% to 40%. On the other hand, the share of newer producers has grown from a very low 5% to 13%. Table 1 sets out the position.

SEQ 0259 JOB IOM8627-013-012 PAGE-0210 207-226 35 SCHOL REVISED 20DEC04 AT 22:10 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

210

Hatfield Memorial Lectures Vol. III

Fig. 3 Steel in the world, 1986 Table 1

% of total produced by: Newer producers (mainly developing countries) Smaller traditional Western producers Communist countries Big 3 Western groups Japan ECSC USA

Pattern of world steel production

1974 (704 Mt)

1984 (710 Mt)

1986 (714 Mt)

15

11

13

7

8

8

30 58 17 22 19

37 44 15 17 12

39 40 14 16 10

At the political level the industrialized countries face a dilemma in this field. Development and growth of prosperity in the Third World are reasonably seen as antidotes to the risk of political instability there, with all that such instability can mean generally, and in the East-West context specifically. Moreover, practically speaking, trade between industrialized and developing countries cannot all be one way: we in the developed world must be willing to buy from emergent countries if we want them to take our finished goods. On the other hand, challenges on the trade front mean loss of production and jobs at least in some sectors of the aid providing industrialized countries’ own industries, and that is a price they are very naturally less anxious to pay. Yet it is also fair to ask whether developing countries always act wisely in investing in an industry such as steel which is capital intensive, no longer a major employer, and seldom flourishes in isolation from developed industrial structures and markets. Businesses that operate on a technological base of smaller scale could prove better suited to many of

SEQ 0211 JOB IOM8627-013-011 PAGE-0211 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future?

211

these countries’ situations and needs, and certainly create more employment than does modern day steel. The starting point in examining the future challenge to European steel must sensibly be the prospect for world steel requirements overall into the next century. Here I judge it feasible to assume that there will be an ongoing need for steel worldwide, although growth will almost certainly be slower than in the past. This is consistent with most projections. Looking at the prospect a little more closely, it seems likely that steel demand in the industrialized countries will stagnate or even decline slightly in the 1990s. This will result from a slow growth in the output of steel using industries in those countries being offset by continuing losses of steel intensity. This reduction in steel intensity will be associated partly with increasing sophistication of goods containing steel and improvement in the quality and strength of steel products, and also with market progress likely to be made by competing materials. This last factor impacts especially strongly on industrialized regions, where design techniques are most advanced. In the Third World, economic activity is more concerned with basic infrastructure development which tends overall to offer less scope to competing materials. Figure 4 is of interest in indicating the change in the share of construction in GDP in Western industrialized countries and in developing countries, taking the two decades of the 1960s and 1970s by way of illustration. Figure 5 shows the movement in steel intensity in industrialized Western countries and in developing countries over the period 1965– 1986. The general prospect for growth of steel demand in developing countries is rather better than in the industrialized ones. Infrastructure development is likely to be stronger there into the next century, and investment goods and consumer durables will all increasingly be required. Provided that sufficient finance is available – and that is, of course, a big proviso – steel demand could grow by 2–3%/year in the developing world. However, it must be anticipated that developing countries will themselves largely install capacity to meet their domestic demand so that the traditional producers will face a declining market. The pattern of world trade in steel illustrates the changes that have taken place, and may be expected to continue in coming years. In the early 1950s, the European Community countries and the USA accounted for over 80% of world exports, which, at that time represented in total only about 10% of world production. By the mid 1980s, in contrast, almost 30% of world steel production was sold outside the country in which the tonnage was produced, with the USA and the European Community countries importing almost 40% of all traded steel, compared with less than 10% of the much smaller total in the early 1950s. For the future, industrialized countries are likely to find their exports reducing and imports increasing, with growing self-sufficiency and expansion of exports by developing countries. This prospect is illustrated very clearly in the case of the Japanese producers, who, having regard to the much stronger international value of the yen, are expecting a

SEQ 0212 JOB IOM8627-013-011 PAGE-0212 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

212

Hatfield Memorial Lectures Vol. III

Fig. 4 Share of construction in GDP, 1960–1980.

Fig. 5 Steel intensity, 1965–1986 (apparent steel consumption/GDP).

reduction in exports from 30–35 Mt to about 20 Mt over the next few years. In contrast, imports into Japan are rising rapidly, although, of course, from an extremely low base. The continuing decline of the dollar will put increasing pressure on the net trade position of the European and Japanese industries, although the US industry should benefit and could even re-emerge as an exporter. The question is, then, against this world steel background in terms of demand into the future and the challenge of competitive materials, how far will future steel requirements be met from Europe in liquid steel, or even in finished steel terms?

SEQ 0213 JOB IOM8627-013-011 PAGE-0213 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future?

213

Fig. 6 Developing countries’ consumption, 1976/86.

Let us look first at Third World countries and then at the advantages that they may be held to enjoy. Over the past 10 years or so developing countries have installed over 50 Mt of new capacity while their demand has grown by little more than 25 Mt. That means that, although developing countries are still net importers as a whole, the position is changing rapidly. Figure 6 demonstrates this. The reasons why new producers put down capacity, apart from a general feeling that an independent country needs to assert itself economically in this way, are very varied. In the case, for example, of Brazil, the prime attraction is of building an industry on local raw materials with the potential of a large market, although the latter has still to be realized on a really major scale. A country such as Indonesia, also, has based itself on a local resource: in their case natural gas. On the other hand, South Korea bases its expansion of steel activity essentially on local market demand. As we all know, the most successful steel industry of the past 40 years has been that of Japan, enjoying virtually no natural resources but with a strong market, a skilled workforce, and an entrepreneurial tradition. In contrast. India and Australia, countries with very differing overall economic status but enjoying enormous natural indigenous resources, have both managed to do little more in the steel field than substitute for imports behind strong protective barriers.

SEQ 0214 JOB IOM8627-013-011 PAGE-0214 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

214

Hatfield Memorial Lectures Vol. III

This picture underlines the fact that it is hard to predict who will make a sustained long term entry into the world steel business. In some cases, apparent advantages like low labour costs are offset by such problems as a low level of educational skills. Again, most steel producers incur about one-third of their costs in purchasing raw materials, with about two-thirds being devoted to domestic labour, goods, and services. Inadequacy of quality in this second large element can outweigh any advantages on the raw materials side. A country like South Korea is fortunate in certain key respects in that it combines low wages with a technically skilled workforce, while the quality of the locally purchased goods and services is excellent. The overall approach to general industrial development in any country has a major impact on the prospects for establishment of a viable steel industry. Few steel companies have succeeded where steel development has been isolated. This is essentially because it is difficult in such circumstances to develop a pool of skilled labour; to secure supplies of purchased goods; and to maintain and develop plant as technology changes. Third World countries have the ability nowadays to purchase the most recent technology developed beyond the pilot stage. The less developed countries need the technology and expertise of the industrialized world in order to exploit their potential as steel producers, and in general the industrialized countries have been willing to provide such knowhow and assistance on a commercial basis. It is sometimes suggested that the industrial countries of the OECD should agree not to help less developed countries to enhance and improve their steel industries in this way. However, that is not a realistic or practical approach: at best such an embargo on the spread of technology would make for only temporary delay. In the longer run, relative economics and market forces would still prevail. Moreover, East-West economic rivalry looms behind this prospect, with the certainty that the USSR would seek so far as possible to fill gaps left by the West. A competitive advantage of less developed countries is their lower standard of living and lower wages. Typically, the average cost of employment overall for a steel worker in Latin America is perhaps 25–30% of that of his counterpart in the UK. That is reflected not only in direct labour costs of steel production, but also indirectly in the costs of indigenous raw materials and services. Table 2 gives indications of relative remuneration, including benefits and employers’ social security outgoings, in 1984, the most recent year for which this country coverage is available. Reduction of the gap in labour costs between industrialized and developing countries could come about through an increase in real wage levels in the latter. That in turn depends largely upon the rate of economic growth in developing countries and there has been little evidence that the gap between richer and poorer countries is closing. On the other hand the possibility exists that the real wages of workpeople in industrialized countries’ steel industries could fall. However, while there have been cuts in real wages in the United States, this has essentially done no more than bring them somewhat nearer to those of other industrialized countries. A general downward movement seems extremely

SEQ 0215 JOB IOM8627-013-011 PAGE-0215 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future? Table 2

215

Hourly remuneration costs for production workers in the iron and steel industry: 1984

USA Japan UK Continental EEC Brazil Mexico Taiwan Republic of Korea

US dollars

Relative to USA. %

20.3 11.0 7.2 10.1 1.7 2.3 2.1 2.1

100 54 36 49 8 11 11 10

unlikely. Therefore, the probability is that the developing countries will maintain their wage cost advantage as far as the eye can see. With an industry as capital intensive as steel, money for investment is a major issue everywhere, and especially in the less developed countries. However, experience in recent years suggests that, at the political level, industrialized countries are willing to make a certain amount of finance available, either directly from governments in the form of aid, or through soft loans, generally provided through the banking system, which in some Western countries is quite closely associated with government. Frequently the provision of such finance is linked to a political desire to support the export efforts of heavy engineering contractors, especially when these businesses are located in relatively depressed areas. Despite debt problems in some developing countries and widespread concern about the overexposure of banks, which seems sure to be underlined by the turmoil in world financial markets this autumn, the likelihood appears to be that, in one way or another, a measure of financial assistance to developing countries in the establishment and expansion of steel capacity will continue. Over the past quarter century the rate of industrial and infrastructure development in many Third World countries has been rapid. Understandably, this has been the main priority, often with relatively little consideration given to environmental aspects. In many such countries legislation exists on environmental protection – sometimes recent – but its enforcement can be less than vigorous, such that it seems likely that older works will continue to operate without incurring the cost of the controls now required in the industrialized world. That position is changing to some extent; for example, the phase III expansion of Usiminas in Brazil, from 2.5 to 3.5 Mt/year, involved expenditure of $70m on environmental control systems, which represented 10% of the total investment cost for the plant. Nevertheless, it is very likely that, overall, environmental measures will, for the foreseeable future, press less hard in developing countries than in industrialized ones, where ever more demanding standards are insisted on.

SEQ 0216 JOB IOM8627-013-011 PAGE-0216 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

216

Hatfield Memorial Lectures Vol. III Table 3 Apparent steel consumption, developing countries, Mt Average 1977–83

1985

1990

1995

Asia Latin America Africa Middle East

39 31 10 16

47 28 10 16

60 35 11 12

68 44 12 13

Total

96

101

118

137

Let us now turn to what might reasonably be judged the relative disadvantages that Third World countries are likely to face in regard to the development of steel production. First, the fact that their markets for steel are, in comparison with the position achieved in the industrial countries, undeveloped, or, at best, are developing only slowly. That is sure to change to some extent over coming years and there are grounds for concern that developing countries may be about to repeat the mistake of overestimating likely demand growth which the industrialized world made a decade ago, with the disastrous consequences for market balance and for the profitability of the steel industry which we are still, in many cases, experiencing. Table 3 shows apparent consumption forecasts for developing countries made at last year’s International Iron and Steel Institute annual conference. Reports suggest that many developing countries have ambitions well in excess of these levels. The overall prospect for political stability in these countries remains a potential problem. Actual severe political difficulties can result in disruption of production and failure of demand, but even a perceived serious risk of it can have a detrimental effect through discouragement of foreign capital. Clearly it would be wrong to imply that all, or even most developing countries carry a risk of political instability, but the record does, without question, show that the risk is greater there than in the industrialized world. The most extreme example of recent years has been Iran, where the economy was developing on quite a stable and flourishing basis, including a significant steel industry, when, in an extremely short space of time, a regime and system took over whose priorities, in Western terms, appear more akin to our Middle Ages than to the present day. That was exceptional, but lesser upheavals do take place, and will probably continue into the future. Although wage rates are normally lower in developing countries, labour productivity is also inferior. At present the balance of these two factors probably leaves the developing countries with manpower costs/tonne overall not much more than half those prevailing in the developed world. Moreover, lower wage levels do not create the pressures for improved manpower productivity experienced in the developed world, although there is growing awareness that gross overmanning results in management problems, which show themselves in poor technical performance: in terms of yield, availability, and, above all,

SEQ 0217 JOB IOM8627-013-011 PAGE-0217 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future? Table 4

217

Assessed labour costs in certain countries: 1985 (hot rolled products)

UK France FRG USA Japan Republic of Korea Brazil

Unit labour cost US $/t

Labour cost as share of total cost. %

34 54 50 86 41 18 29

15 23 22 25 16 8 10

quality. Table 4 shows assessments of comparative labour costs at the hot rolled product stage based on work by Paine Webber, the US stock market steel analysts, a couple of years ago. Extreme movements in the dollar have presented severe challenges to the methodology which the originators have not yet resolved for any later years. Very important is the rather indefinable quality which one might describe as an entrepreneurial tradition, or the relative absence of it. If one asks oneself how it came about that the Japanese successfully made such an enormous expansion in steel from about 1960, then one of the reasons surely was that, in contrast with many countries which have subsequently sought to develop steel, the Japanese had not only an appreciable basis of experience in the steel industry itself, but also a very vigorous and flourishing business background and tradition of long standing. In many countries such traditions, at least so far as operating business on a large scale is concerned, are absent. Much more specific is the absence in developing countries of a research and development, and technical background. Steel is coming under increasing pressure from other materials, and while this may not pose serious problems in relatively unsophisticated domestic markets, it is a very important factor in the international market and will become more so. In order to meet such competitive challenges it is essential to carry out research into improved methods of production so as to reduce costs and improve quality; and into product properties to protect existing markets and to develop new ones. Moreover, a research facility provides specialist expertise in key areas, as well as equipment that can be put at the service of many plants which individually would not justify its cost or fully utilize it. Research and development departments are also an important stimulus to development, and they can provide an input of technically trained and experienced people for production and commercial activities. None of these advantages can be generated overnight. Research departments cannot be created in isolation: they depend not only on graduates, but also on technicians and others with varying degrees of technical education, and skills. That implies an educational system which, at all levels, has an appreciation of industry’s needs and personal links with it. Europe has not been good at that everywhere, but in some places, and Sheffield is an excellent example, an effective and mutually beneficial relationship has long existed. It is hard to see anything comparable developing in newer producing countries in the short term.

SEQ 0218 JOB IOM8627-013-011 PAGE-0218 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

218 Hatfield Memorial Lectures Vol. III Let us stay for the moment with disadvantages as we turn to the European scene. I suggest that the disadvantages which face the European steel industry are in considerable measure associated with the word ‘tradition’. Tradition, that is, not only in the sense of the social climate, though that is important, but also in terms of industrial inheritance. In the latter area, the problems are associated especially with the division of operating sites deriving from the scale of the technology of earlier periods and with multiplicity of ownership in those days. Correcting the cost disadvantages of this inheritance is, of course, one of the major thrusts of the restructuring in Europe which we have been engaged in for several years now, and to which I will refer in my conclusion. The disadvantage arising from multiple sites is compounded where those sites are inland ones, and very serious disadvantages arise where works are remote both from a deep water port and from principal markets. Table 5 showing iron ore imports and exports by the main regions of the world illustrates the flows in this key raw material. The disadvantages in the social sphere deriving from the traditional character of most of our societies in Europe are more difficult to describe in precise terms. They are associated, first, with the pull of communities, which makes for a relative lack of labour mobility. This is reinforced, certainly in the UK, by longstanding housing policies, especially as concerns the private rented sector. Quite closely linked with this is the notion of adherence to traditional working practices, which have been an especially strong force in the UK, and which we are only now overcoming in part and with great difficulty. This problem is probably exacerbated by the relative absence of what one might call consensus management in much of Europe. Relationships within businesses in Europe have been structured and potentially confrontational, although I do believe that we have made significant and very welcome progress in this respect over recent years. I have already referred to the relatively high wages rates in the industrialized world. Partly responsible for this are the expectations created by the consumer affluent society, which we have generally known in recent times. While this has been one of the engines of continuing growth, it nevertheless carries disadvantages on the cost front. These are reinforced where there is a political background which also increases social and employment costs more generally. Of a more measurable quality are the disadvantages which Europe, as well as other industrialized areas, suffers in terms of pollution control costs. The trend within the European Community is towards ever more stringent environmental legislation, sometimes based on inadequate technical arguments. The Federal Republic of Germany is often, rightly, perceived as the trend setter in this regard, and has been influential in the definition of community law. In 1986, BSC made estimates of the cost of air pollution control to meet all legislative requirements at a typical UK integrated steelworks producing about 2.5 Mt of steel/year. We reckoned the investment costs to be about £75m on that basis, and operating costs of control measures to add about £8/t of liquid steel. Finally, there is the involvement of governments, in one way or another, which, indeed, Third World countries also experience to a high degree.

SEQ 0219 JOB IOM8627-013-011 PAGE-0219 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future?

219

Table 5 World iron ore trade: 1986 Mt Region

Import

Export

EEC Japan USSR and East Europe Other Asia North America Other West Europe China Latin America Middle East Africa Asia Oceania

128.1 124.4 54.5 22.5 21.9 9.7 9.5 3.0 2.8 ... ... ...

6.8 ... 43.9 ... 37.3 20.9 ... 111.3 ... 35.7 32.7 87.8

Let there be no doubt in anyone’s mind that, in relation to the severity of the crisis which European steel has faced and the dramatic character of the restructuring needed for survival, government intervention, financial and otherwise, has been inevitable and essential. Nevertheless, there are major disadvantages associated with what one might term the natural non-commerciality of governments. Rightly and inevitably, governments are concerned with social issues in all their many dimensions. Moreover, no responsible employer arrives at his decisions without having proper regard to such considerations. Nevertheless, when non-commercial considerations become paramount, as they often do where governments are involved, whether through complete ownership or through more indirect means, competitiveness inevitably suffers, with the longer term costs associated with that. Now What does the European steel industry specifically have going for it? First, and most important, the existence of very developed and sophisticated domestic markets. In recent times, when the biggest industrial powers of the West have been referred to, it has tended to be in terms of the USA, Japan, and Germany. Now Europe’s complex industrial structure calls, long term, for a cost, quality and delivery effective steel industry to support it. If the European Community’s current programme for moving towards an integrated and coherent internal market of 300 million people operating in a stable political and social environment could make substantial progress, the prospects for the competitive development of its economic activities, including steel, would be greatly strengthened. Customers demand a high and consistent quality of product delivered when and where it is wanted. They demand also that any current technological improvements be incorporated into their purchases immediately, and are progressively moving towards ‘just in time’ delivery, with all its associated quality control implications. Increased involvement between the research and development activities of supplier and consumer

SEQ 0220 JOB IOM8627-013-011 PAGE-0220 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

220

Hatfield Memorial Lectures Vol. III

can lead to tailored products, developed for particular applications. If a steel plant is put down in advance of broader industrial development, as often happens in the Third World, the crucial sales links such as we in Europe can and do develop with our key customers are lacking. It is almost certain that commodity markets will continue to represent a large part of the steel business, and in that area price is the overriding factor. European producers are likely to experience considerable pressure from developing countries in these sectors, but important and lucrative closely customer linked niches which, in total, are substantial, could potentially remain largely the preserve of the European producers themselves. This factor imparts great importance to the finishing end of the steel business, which is nearest the customer. Many recent and current technical developments are in that area, embracing, for example, the increased use of zinc and other protective coatings for corrosion resistance, including in the motor car industry. Investment in plant in Europe is likely, on these general grounds, to be concentrated much more at the finishing end of the operation than has often been the case in the past. A specific challenge in the markets where the European industry with its developed customer links potentially has the ability to move quickly is the development of composite materials. Major steel strip markets have been threatened by the emergence and proliferation of plastic-fibre composite materials for at least the last quarter century. Initially these materials were introduced into commodity markets, where their lightness and freedom from atmospheric corrosion provided an early foothold. As a result of recent development, more sophisticated composites can now offer a level and consistency of technical performance approaching that of strip steel. The steel industry must respond with its own product and process innovations. When combined with high strength steel substrates, the variety of durability and aesthetic appeal conferred by metallic (predominantly zinc based) coatings are further augmented by additional strength and lightness through the potential for downgauging. I have already referred to the beneficial tradition of close liaison between the steel industry and the educational establishment in Europe. A complex set of relationships exists which make for a wide variety of education and training facilities at local and national level. These are reinforced by professional organizations representing academic and industrial interests, which also confer membership representing recognized levels of professional competence and experience. Moreover, the steel industry has always been an open one in technical terms, both nationally and internationally. This is particularly true of Europe where plant management researchers and academics have traditionally been in close touch. The mechanisms for ECSC and EEC research, and for demonstration plant programmes under both the Paris and the Rome treaties which underlie the European Communities, have reinforced these relationships. Beside these traditions in education and in research and development lie a business flair, and long experience in organizing large steel production. Ability in these respects varies between different countries, but overall in Europe it is relatively high. Certainly

SEQ 0221 JOB IOM8627-013-011 PAGE-0221 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future?

221

levels of personal and business initiative and flexibility compare very favourably indeed with those generally evident in, for example, Eastern Europe, while in developing countries there is simply, for the most part, less background and experience in these fields. A crucial factor is the ability to adapt to changing circumstances, which has always been integral to success in economic and industrial activity. There can be little doubt that the capitalist system such as we broadly have it in Europe generates pressures to make changes, without which businesses cannot survive. More corporatist political and economic forms, such as many Third World countries and very obviously Eastern Europe experience, do not achieve this. Table 6 shows, by way of illustration, reductions in employment in steel in certain industrialized countries between 1974 and 1986. Figure 7 graphs an assessment of the development in labour productivity in a few key Western countries over the period 1972–86 which shows marked convergence. At this point one’s mind inevitably turns to the UK over the last two decades. No one could deny that the challenge with which we in the UK have been confronted over the past few years, largely as a result of the financial management and policies of the present government, has driven us to achieve far greater efficiency here. As is well known, there is a good story to tell in that regard for British Steel, but I will resist the temptation to tell it. I have tried to present to you some of the actual and potential advantages and disadvantages experienced by the steel industries of the Third World and of the industrialized world respectively. I should now venture an answer to the question that forms the title of the lecture: ‘European steel: what future?’.

Fig. 7 Labour productivity movements in certain countries, 1972–1986.

The factors I have already reviewed argue that many developing countries are now strong competitively, and will become more so in the future. In summary, low wages and low or nonexistent social charges make for much lower fixed costs than among traditional producers, while developing countries often enjoy also lower variable costs of production

SEQ 0222 JOB IOM8627-013-011 PAGE-0222 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

222

Hatfield Memorial Lectures Vol. III Table 6

USA UK France Spain FRG Italy Japan Total OECD area (estimate)

Average employment in steelmaking 1974 and 1986, 000s 1974

1986

% change

512 194 158 89 96 232 459

175 57 68 51 66 143 340

–66 –71 –57 –43 –31 –38 –26

1950

1110

–43

through availability of cheaper raw materials and energy. FOB costs of steel exported from such well placed developing countries can at present typically be 20% lower than for equivalent European producers in terms of total operating costs, excluding depreciation and interest. The main advantages which we in Europe have are a large and developed domestic market with proximity to customers, and certain business, technical, and educational traditions and experience. Can these advantages counteract those from which many Third World countries benefit? The challenges we shall meet in the future will be intense without doubt, but in my judgment they can be successfully met, on condition that Europe exploits to the very fullest the key advantages we do have. That means first and foremost that the European steel industry must complete the restructuring process it has been engaged in for so long now; and that very soon. Continuing excess capacity resulting from the stagnation of demand following the first oil crisis, as against the expansion of capacity through investment put in train during the boom years of the early 1970s, lies, as everyone must now surely understand, at the very heart of the problems faced by the European steel industry. It is essentially this which ongoing restructuring efforts are seeking to correct. Figure 8 is a series of pie charts showing reductions of hot rolling capacity in EEC countries to date, amounting to around 35 Mt since 1980 out of a total of 168 Mt then, i.e. just over 20%. You are well aware of the present stage reached in this process and of the report recently submitted to the European Commission by the so called ‘Three Wise Men’ which has confirmed the apparent intractability of the problem. The Industry Council of Ministers of the European Communities is meeting today [8 December] to examine the results of this latest work and to consider the future position. At present the European Community steel industry is gravely handicapped by its slowness to complete essential restructuring. Furthermore, the Commission-sponsored Article 58 compulsory quota system, which was introduced in order to provide the industry with breathing space while fundamental restructuring took place, now appears to be seen in some quarters as an end in itself. Worse, the industry has become intensely selfregarding and absorbed with its own problems. Indeed, the whole ECSC steel market

SEQ 0223 JOB IOM8627-013-011 PAGE-0223 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future?

223

support system has become so elaborate and complex that it constitutes a distraction from those key tasks focused on meeting customers’ changing requirements to which the industry should urgently be directing its energies in order to ensure its future. There is no future in the European steel industry seeking to continue artificial market support measures internally on an indefinite basis as an alternative to tackling the fundamental restructuring. Nor, on the other hand, is there any future in trying to live permanently behind protective barriers against third country competition, tempting course though that may appear to be, except in specific and defined circumstances as provided under the GATT to counter subsidized or dumped imports which threaten to undermine the European industry by those means. If wholesale protection of steel were to be resorted to and our consumers denied access in the long term to genuinely competitive offers from the developing world and elsewhere outside the community, the thrust of external competition would simply move downstream, as overseas steel consuming businesses competed successfully against ours. In that case, the last state for steel would be worse than the first. Beyond the fundamental restructuring drive which must be completed, the European steel industry will need to continue to work very hard to improve its technical performance and lower its cost base in every way. Concentration of operations on fewer and more efficient plants of itself gives significant improvement in energy usage and yield, which can be further built on by judicious capital investment. Specific energy consumption in European steelmaking has improved appreciably in recent years. It is lower than in most Third World countries and centrally planned economy countries. However, comparison with the Japanese situation shows that there remains room for further improvement, although the rate at which this takes place will inevitably depend upon the ratio of fuel costs to other costs, particularly capital. Moreover, account has to be taken of the fact that energy consumption in each country is influenced by the tonnages of steel produced by the integrated and the scrap based electric routes respectively. Table 7 indicates the net energy requirements of the steel industry of a range of industrialized countries in 1985 and 1986; and the change in each since 1980. So far as Europe’s advantage in terms of experienced human resources is concerned, a highly skilled, educated, and adaptable workforce is a very major asset in a changing world. The future of the European steel industry is likely to hinge as much on continuation of a high level of technical education and of effective training as on economic factors. The British steel industry has given high priority to training and management development ever since the Second World War, and we are paying ever more attention and committing more resources to this area. In terms of further major changes that the European steel industry may have to make for survival in the future, two possibilities are, in my judgement, the most likely to arise. The first is associated with the point I have already made that, in the future, in view of the need for the European industry to optimize its market advantage and proximity to customers, investment may be somewhat concentrated at the finishing end. This consid-

SEQ 0224 JOB IOM8627-013-011 PAGE-0224 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

224

Hatfield Memorial Lectures Vol. III

Fig. 8 EEC reductions in rolled steel capacity, 1980–87 Table 7

Net steel industry energy requirements GJ/tcs*

% change

Year

1980

1985

1986

1980–85

1980–86

France USA FRG UK Canada Italy Japan

24.30 25.80 22.38 23.40 22.83 18.98 19.47

22.40 22.30 21.58 21.20 22.10 17.70 17.66

21.00 ... 21.73 21.00 23.15 18.47 17.64

–7.8 –13.6 –3.6 –9.4 –3.2 –6.7 –9.3

–13.6 ... –2.9 –10.3 +1.4 –2.7 –9.4

*tcs = tonne of crude steel.

eration raises the possibility of multinational efforts, by way of more liquid steel being produced in developing countries for onward processing in Europe and elsewhere into customers’ finished product requirements. These notions have, of course, been around for 20 years and more now, and have so far not been implemented in most cases. Nevertheless, they must remain a serious possibility for wider adoption in the longer term future, despite the fact that at present there are few multinational steel producing companies to facilitate such a development, though this is very unlikely to happen on a wholesale basis. With low labour costs and access to indigenous raw materials, it ought to be possible to manufacture semifinished steel in certain less developed countries and transfer them to Europe for finishing on an economically attractive basis. However, a number of considerations would need to be taken very firmly into account, and particularly assurance of quality of the semifinished steel, which in many cases would require an injection of

SEQ 0225 JOB IOM8627-013-011 PAGE-0225 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

European Steel: What Future?

225

technology from Europe; and the potential of political risk where there is dependence on supply of the primary steel from overseas. The second main area of change is likely to concern competition from other materials. Not in all, but in many parts of Europe, the steel industry remains very self-contained, in the sense that it is not directly linked to downstream activities, many companies having no diversified operations into other sectors. For the longer term future there must be a strong possibility that European companies will not only concentrate on operation of sophisticated finishing and processing lines, but also develop the distribution and service centres dimension of activity, including the kind of further processing that can be undertaken on that basis. Moreover, the steel industry may come to regard itself increasingly as a materials industry participating in such fields as composite materials, to which I have already referred. The existence in the European Community of the Paris Treaty, which governs exclusively the coal industry and the steel industry down to hot rolled products and a limited range of cold rolled products, has reinforced the tendency for the steel industry to regard itself as something different and special. Without doubt some of the distinctive features of the Paris Treaty have proved very useful in operation, but for the longer term it would be healthier for steel to become like any other industry subject to the broader European Community regime under the Rome Treaty. The Paris Treaty in any case expires on 22 July 2002. The ability of the European steel industry to adapt to future challenges, including particularly those from developing countries, will depend upon its having commercial freedom. That implies movement away from the posture of state dependence or close linkage, which has obtained in varying degrees and at different times in many parts of Europe for most of the period since the First European War. In turn, that implies in principle that private ownership of European steel will need to be the norm in the future. In summary, quicker development in the Third World must be assumed, with producers there continuing to enlarge their share of the world market. However, despite disadvantages in access to raw materials and in labour costs, European steel can survive long term, provided that it maximizes its advantages in terms of the market, and in R&D, in technical education, and in business experience. Given the will to drive forward with imagination, and provided that it adapts effectively to the inevitable oncoming challenges, I hold to an optimistic view of the future of the European steel industry.

SEQ 0226 JOB IOM8627-013-011 PAGE-0226 207-226 35 SCHOL REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0302 JOB IOM8627-014-008 PAGE-0227 227-246 BOM REVISED 22DEC04 AT 00:16 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

THE FORTY-SEVENTH HATFIELD MEMORIAL LECTURE

Iron, the Hidden Element – the Role of Iron and Steel in the Twentieth Century R. Boom At the time the lecture was given Professor Rob Boom was Director R&D, Strategy & Competence, Corus Research, Development & Technology, Ijmuiden, the Netherlands. He held the NIMR chair in Primary Metals Production at Delft University of Technology. The lecture was delivered in Sheffield on 7th December 1999.

Alexandre Gustave Eiffel, Henry Bessemer, Andrew Carnegic, Alfred Krupp: all well known to the general public at the end of the nineteenth century, and all involved in some way with iron or steel. Where are their late twentieth century equivalents? Most of us would struggle to name spontaneously four figures strongly related to steel whose names are known to the man and woman in the street. People are hardly aware of the fact that the availability of iron and steel is crucial to almost every application, every feature, that shapes the quality of modern life. Iron and steel are so basic to society that they are taken for granted and hidden – sometimes literally, but more often figuratively.

THE WORLD 100 YEARS AGO Economic and social changes at the end of the nineteenth century were marked by developments that can be attributed to the steam engine, in its stationary form (in factories) or moving on railroads, canals, rivers, or oceans. In most ‘developed’ countries the railway network had to a large extent been completed. (Note that the French use the expression chemin de fer, the Germans use Eisenbahn, and the Italians ferrovia – all meaning ‘road of iron’. In contrast, the British use railway, the Americans railroad, and the Dutch spoorweg, thereby denying, or hiding, the fact that rails are made of steel.) The railway infrastructures included some fine examples of advanced construction technology in iron and steel, like the Forth Bridge in Scotland, with a length of 2.46 km and two 521 m spans, then the longest bridge in the world. Small scale domestic production of goods was substituted by mass production in mechanised factories. A striking example of this is found in textiles. Cities were growing fast, notably by the addition of thousands of small houses for factory workers. A labourer’s life was of low quality. There were no sewage systems and hardly any facilities for health and personal care. Heating was inadequate. Oil for heating was just being introduced at

227

SEQ 0261 JOB IOM8627-014-006 PAGE-0228 227-246 BOM REVISED 20DEC04 AT 22:14 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

228

Hatfield Memorial Lectures Vol. III

Fig. 1

Iron house originally exhibited at the Paris World Exhibition. 1887–89, now on display in the city museum, Yekaterinburg.

the end of the nineteenth century; coal and wood were the fuels that really mattered. Food was expensive and of low nutritional value. There was a great need for mechanisation of agriculture to produce the amounts of food required for human and animal consumption. Transport from the farms to the cities had to be arranged. But the first signs of a New World appeared, some of which were shown at the famous Paris World Exhibition of 1887–89. For example, the cast and forged iron house, a demonstration of the best of iron technology from Russia (Fig. 1). The 300 m high Eiffel tower stole the show; note, however, that the Eiffel Tower is made of iron, not steel. The tower was a demonstration by the engineer Eiffel of the feasibility of using iron for large constructions, a proof to the world of new possibilities. British engineers such as Telford and Brunel had realised more complicated work; in fact, the Eiffel tower was a monument to the past rather than to the future. It was intended to survive the exhibition for only a couple of years, but today it is still the symbol of Paris.

FROM IRON TO STEEL Henry Bessemer was a great inventor; he triggered the steel revolution in Great Britain at the end of the nineteenth century with his converter process. By blowing air through low phosphorus iron in a refractory lined vessel, low carbon steel could be produced in large quantities in short production times. Subsequently, Sidney Gilchrist Thomas adapted this process to convert high phosphorus iron, very important to European producers. The open hearth process, developed by Friedrich and Wilhelm Siemens and Emile Martin,

SEQ 0229 JOB IOM8627-014-005 PAGE-0229 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 229 was a next step in the large scale production of steel. The material, production processes, and facilities were there; steel was ready to play a major role in the twentieth century. What was the importance of Bessemer’s and Siemens’ findings? Steel, a low carbon iron alloy, is ductile; it is ‘tougher’ and ‘stronger’ than cast iron; and it can be made in large quantities at an economic price far exceeding the possibilities of forged iron. This contributed greatly to the victory of steel over iron; whereas in the second half of the nineteenth century iron and steel were applied to a comparable extent, the twentieth century saw applications mainly of steel. The Forth Bridge, for instance, was one of the very first large bridges in steel. In fact the Iron Age turned silently into the Steel Age, unnoticed by the majority of mankind.

THE EARLY TWENTIETH CENTURY During the early decades of the twentieth century, steelplants and steel companies were established throughout the world. In 1901 United States Steel Co. was founded by Andrew Carnegie. US Steel contained Carnegie’s interests in numerous small, mainly Pittsburgh based, steel companies and was introduced to the New York Stock Exchange in that year, the first steel producing company to be so listed. US Steel soon became the largest integrated steel producer in the world, owning iron ore mines, coal mines, railways, and shipping companies. Its power and success became so overwhelming that the US authorities started an antitrust investigation in 1910. In that year the company controlled 80% of the US steel market. By the time the investigations and the court hearings were finished, 10 years later, the situation had drastically changed and the market share of US Steel had dropped to 40%. The judge did not accept the charges, because of this drop in market share and because monopoly price setting was not proven. Carnegie, being a businessman rather than an inventor, made a lot of money out of steel. He used it wisely and with great charity to support science and technology and to reward those who had showed heroic behaviour in saving others. Carnegie lived in two centuries, but was mainly active in the nineteenth. Also in 1901, Yawata Steel was founded as a state owned company on the island of Kyushu in southern Japan, and it produced the first iron ever made by the blast furnace route in Japan. Kyushu coal and local ironsand were used as raw materials for iron and steel production. Yawata Steel is one of the predecessors of Nippon Steel, now one of the world’s largest steel companies. In 1903 Friedrich Krupp GmbH was founded in Essen, Germany. A steel producer in the heart of the Ruhr area, famous for foundry products such as cannons, the family business, founded about 100 years previously and considerably expanded by Alfred Krupp, now became a publicly quoted company under his son, Friedrich. Krupp rapidly introduced Bessemer technology1 (Fig. 2). This happened in a country that was industrialising at a high pace (Fig. 3). In terms of iron production, it can be seen that the UK, the creator of the Industrial Revolution, led

SEQ 0230 JOB IOM8627-014-005 PAGE-0230 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

230

Hatfield Memorial Lectures Vol. III

Fig. 2 Bessemer steelworks of Krupp in Essen (c. 1910).

industrialisation in the nineteenth century. France was picking up slowly, and Germany was rapidly increasing iron ore production. Coal and ore were available in Silesia, the Ruhr area, and in Alsace and Lorraine, which had been acquired from France in 1871. In Germany the Second Industrial Revolution began, triggered by coal, iron, and steel and by the chemical and electromechanical industries. Steel permits equipment for large scale generation and use of electrical energy to be built (to be exact: copper and iron). Without steel there can be no economic production of electricity: although iron does not conduct electricity very well, it guides the bulk of electrical power – transformers and generators need electrical steels. Electricity was easily transportable and could take over the role of steam engines as power sources. Germany led this new technology. There was also the development of sewage systems and other water ducts using iron or steel pipes. Transport of gas for domestic lighting and cooking was done through 1 inch steel ‘gas pipes’: steel as distributor of health and welfare for all classes of society. The first metal carriages or motor cars were designed and built. Henry Ford founded Ford Motor Co. in 1903. Mass production of automobiles started with the success of the Model T Ford in 1908. Henry Ford was elected businessman of the twentieth century by Fortune magazine. Without large quantities of steel of reliable quality, available at low prices, he could never have achieved this. I select him as the first name connected to steel in our century.

SEQ 0266 JOB IOM8627-014-007 PAGE-0231 227-246 BOM REVISED 20DEC04 AT 22:25 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 231

Fig. 3

Comparison of iron ore production per capita in the UK, Germany, and France (1810–1910).

Large ships were being built in steel. We all recall the ‘Titanic’, one of three sister ships built around 1910 for the White Star Line to compete with the already famous ‘Lusitania’ and ‘Mauretania’ of the Cunard Line – ships that set the measure for a whole branch of industry. The UK led the world with its famous shipyards such as Vickers in Newcastle and John Brown on Clydeside. The recent movie about the tragic maiden voyage of the ‘Titanic’ focused the attention of the public on matters such as steel quality and joining technology. The other ships were also unlucky; both the ‘Lusitania’ and the ‘Mauretania’ were later victims of war.

WAR: A BOOST TO TECHNOLOGY History has shown that large scale wars generally provide a boost to technology – long before the word even existed. The First World War was the first in which steel was available in large quantities. It was used for weapon systems, such as Dicke Bertha (Big Bertha) made by Krupp, a cannon with a range of 100 km that was used inter alia to fire on the steel city of Li`ege. It was named after Bertha Krupp, only child of Friedrich. Tall warships, made from Heavy armoured plate, were constructed with heavy artillery aboard, introducing the term ‘battleship’. Armoured cavalry wagons with steel caterpillar tracks were introduced and became known as military tanks. Historians claim that none of these developments significantly affected the course of the First World War. No, the real tragedy of the tremendous losses of young soldiers is to be attributed to the machine gun and the thousands of kilometres of barbed wire at the front lines – hidden weapons made from steel. An irony of history, given the theme under consideration, is that the decisive battles on the Belgian front were fought along the river Yzer, the Flemish word for iron.

SEQ 0232 JOB IOM8627-014-005 PAGE-0232 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

232

Hatfield Memorial Lectures Vol. III

Fig. 4 Hoogovens site near Ijmuiden (c. 1925).

The First World War made it evident that steel had become essential to national economies. This was felt particularly in the Netherlands, which because of its belligerent status had been isolated and deprived of foreign strategic supplies (any material in which it was not self-sufficient, including all metals) from 1914 to 1918. The experience boosted private initiatives to found a steel industry in the Netherlands. There had been attempts before, but during the nineteenth century the prevailing opinion was that such activities were out of place in a traditional trading nation without the necessary sources of iron ore. In fact, it needed a war to make people realise that the country could no longer do without a steel industry.

HOOGOVENS Just before the end of the war, in 1918, Koninklijke Nederlandsche Hoogovens en Staalfabrieken NV was founded, a company that later became known as Hoogovens, the Dutch word for blast furnaces.2 At the chosen site near Ijmuiden (Fig. 4), a small port on the North Sea coast some 25 km west of Amsterdam (chosen after that city had provided some additional capital), the necessary raw materials were not readily available. Neither was skilled labour: the people in the area were strawberry or bulb growers or worked at sea as fishermen or sailors. Nevertheless, these people were hired and trained. Technology was imported from Germany; German Meisters, experienced ironmakers from the Ruhr area, trained the personnel.

SEQ 0233 JOB IOM8627-014-005 PAGE-0233 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 233 Blast furnace ironmaking started in early 1924. Despite the lack of experience, Hoogovens soon built up a reputation as a producer of high quality pig iron and became renowned for its marketing policy. The annual production capacity of over 0.25 Mt of pig iron, though quite modest by present standards, far exceeded the needs of the local market. So, right from the beginning the intention was to export a substantial part of the iron produced. This was a brave approach indeed, given the import and tariff barriers that existed then in Europe. But the management found a way out of this problem, setting up an international network of agents who sold directly to foundries and steelworks. By doing so, Hoogovens gained an advantage over its foreign competitors; whereas they sold only via traders and had no direct contact with their customers. Hoogovens supplied exactly what it found its customers wanted. Thanks to its favourable geographical location it could sell at very competitive prices. In fact, Hoogovens’ marketing expertise proved to be its greatest asset. Whereas, during the crisis of the 1930s, iron and steel producers all over the world were forced to cut production and even to withdraw from export markets, Hoogovens managed to maintain a constant production level. In that period it was the world’s largest exporter of pig iron – though at rock bottom prices. The modest income from that business proved sufficient to keep the company running.

INTERWAR PERIOD The first decade after the First World War saw strong growth of industrial production in almost all industrialised or industrialising countries. The motor industry and the machine industry expanded. Tremendous growth in applications of electricity called for a corresponding infrastructure, whereas the growth of all kinds of transport made numerous civil works necessary: roads, bridges, ports, and so on. From a technical point of view, the shipbuilding industry reached maturity. Its further sophistication became a driving force for the development of new steel grades. Large, fast ocean liners could now be built for the intensive Atlantic passenger trade. These subsequently became genuine prestige objects, competing for the Blue Riband, the award for the fastest transatlantic crossing. In the UK, for example, the ‘Queen Mary’ and the ‘Queen Elizabeth’ were built to designs that allowed the British to regain the Blue Riband from competitors such as the French, the Americans, and the Italians. In the USA, the first household appliances came onto the market. The first washing machines appeared in 1920, soon followed by refrigerators. They became very popular; the American kitchen, with many pieces of equipment made out of steel, became a dream for millions all over the world. Consumers, not just governments, were asking for steel products. On the other side of the globe the USSR, created after the 1917 revolution, first had to struggle with a civil war. After consolidation and after the death of Lenin, a plan for large scale industrialisation was developed. It was launched by Joseph Vissarionovitsch

SEQ 0234 JOB IOM8627-014-005 PAGE-0234 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

234

Hatfield Memorial Lectures Vol. III

Djugashvili, who changed his name first into Koba and then into Stalin (‘man of steel’), my second nomination as a name of the twentieth century strongly related to steel. Stalin called for industrialisation in 1925 as the base for his economic policy. The goal was socialism, the method was planning, the instrument was the five year plan. Stalin was almost obsessed by the role of heavy industry, in particular by Big Steel. He was in a hurry; the USSR was on the ‘March for metal’. He ordered the construction of a huge steelplant at Magnitogorsk, the ‘magnetic mountain’ in the Urals. Here, surrounded by empty steppe, iron ore resources were available. The region, however, was hardly suitable for human life, but this was of minor concern to the Communist authorities. Active construction started in 1928. Unskilled labourers were moved to the site from almost everywhere. Forced labour by convicts was used. Both men and women worked on all aspects of the construction, including heavy labour. The Soviets wanted the best technology, so they had the plant designed by the US engineering company Frey, which had also designed the largest steelworks at that time, US Steel’s Gary Works, built on the marshes of South Chicago. Once the giant project had come on stream, the Americans were asked to leave and Magnitogorsk was declared a closed town, as was the old metal industry centre Yekaterinburg. This centre of mining and metals industry since the early eighteenth century was renamed Sverdlovsk, and was meant to be anther nucleus of the Soviet metals industry. Foreigners were no longer allowed to visit those cities. The Magnitogorsk plant became the biggest steel producing site in the world. Stephen Kotkin, in his book ‘Steeltown, USSR’, wrote:3 It was one of the symbols of the October Revolution itself; of the rise of backward peasant Russia to the status of an industrial and military superpower. In this way steel became almost a synonym for advance, for presumed or expected prosperity. Malenky, in his book ‘The metallurgical combine of the future’, wrote:4 Metal is not produced simply for its own usage . . . Metal draws all industry along with it, all spheres of human life . . . Metal is the basis of modern civilisation. At the same time steel became a route to power. Politicians quickly realised that control of the essential industries – starting with the iron and steel industry and, on its tail, the military industry – paved the way to gaining control of the whole country. In the meantime, the Western economies had collapsed following the Wall Street Crash in 1929. The steel industry suffered badly and passed into a difficult period. It was a labour intensive industry and the effects on employment were dramatic. However, the crisis produced only a slackening in technological development, which accelerated again during the Second World War. In Germany, the National Socialists attacked the employment problem by rearmament. This is reflected in the steep rise of crude steel production, shown in Fig. 5;5 output at the outbreak of war in 1939 was more than four times that in 1932.

SEQ 0270 JOB IOM8627-014-007 PAGE-0235 227-246 BOM REVISED 20DEC04 AT 22:25 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 235

Fig. 5

Production of crude steel in Germany (1870–2000).5

SECOND WORLD WAR AND ITS AFTERMATH During the Second World War steel was a matter of life and death for the combatants. The Atlantic lifeline between the USA and the UK was of paramount importance to Britain. Submarine action resulted in vast losses of allied merchant shipping, which had to be replaced. To this end the USA designed an assembly line to build ‘Liberty’ cargo and T2 tanker ships. From 1942, over 2500 Liberty ships were built; it took about 60 days to produce one ship. This high rate was possible because of technological developments that allowed welded joints to replace the usual plate and rivet construction. The steel industry was able to supply enough material with the right properties to produce this immense number of ships, steel thus creating the aorta between the USA and the UK. However, some ‘Liberty’ ships were lost without enemy action; in 19 reported cases they even broke in two. Metallurgical investigations showed that crack formation in the steel plates was the cause. In the traditional rivet and plate construction this would not be disastrous, but in a fully welded construction it was. (Incidentally, the joints in the ‘Titanic’ – mentioned earlier – were of the rivet and plate type, making crack formation an unlikely reason for its sinking after hitting the iceberg.) Impact tests were designed to map the brittle behaviour of steel grades. New grades were developed for low temperature conditions. The lesson was learnt the hard way, but in the end physical metallurgy provided the solution. The success of the Russian tanks in the battles of Stalingrad and Minsk was largely attributed to the steel produced at Magnitogorsk. The Second World War brought considerable technical advances. It procured for the USA the position of undoubted economic world leader. However, many countries that had been occupied, or that had used their last resources to defeat the enemy, emerged in a desolate state, hindering reconstruction and political stability. Winston Churchill warned early on against the upcoming threat of Communism. In a telegram dated 12 May 1945 to US President Harry Truman, he used the expression ‘Iron Curtain’ for the first time.

SEQ 0236 JOB IOM8627-014-005 PAGE-0236 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

236

Hatfield Memorial Lectures Vol. III

Subsequently, in his famous speech at Westminster University in Fulton, MI, he said: From Stettin to the Baltic to Trieste in the Adriatic, an iron curtain has descended across the Continent . . . In front of the iron curtain which lies across Europe are other causes for anxiety . . . The Communist parties or fifth columns constitute a growing challenge and peril to Christian civilisation . . . Through his popularisation of the term ‘Iron Curtain’ and through his iron will to defend freedom and democracy, his name is also closely connected with iron in the twentieth century. In 1947 George Marshall, the US Secretary of State, announced an extensive European Recovery Program with a target of making impoverished European countries selfsupporting again within four years. Major focal points were the rationalisation of agriculture and restoration of production in essential sectors of the economy. In principle, Marshall Aid was coupled to cooperation between countries to create a stronghold against Communism, but this proved to be ‘a bridge too far’ at that time. Whereas the USSR rejected the offer and forced countries under its influence to do the same, most Western European countries reacted favourably. The Marshall Plan was launched at a time when the Dutch government was changing its industrialisation policy, shifting the focus from import substitution to restoring the balance of payments and creating employment. This implied stimulating capital intensive industries. The government – contrary to its traditional point of view – was willing to participate in the Hoogovens expansion plan, if the capital required came from Marshall Aid.6 Though the plans met with opposition from (more industrially advanced) neighbours like France and the UK, financial support under the Marshall Plan was granted at the end of 1949. Mainly US equipment was installed. HM Queen Juliana officially inaugurated the complete set of new facilities (hot and cold strip mill and tinning lines) in 1953 (Fig. 6). In fact, this marked the completion, 35 years after the launch of the company, of the original plan of its founders: an integrated steelplant. From the Dutch point of view Marshall Aid was crucial in shaping the future of our steel industry. For Germany the change in policy was of utmost importance. The Allied Occupation Force had put a dismantling programme in action, especially in the British sector where about 90% of the German steel industry was located. The first intention was to limit German steel production to a mere 10 Mt/year, but this was abandoned as the Cold War intensified. A major impetus for the Marshall Plan was the US Government’s wish to counteract the threat of Communism in Europe. Subsequently, the same policy was adopted towards its former enemies in Asia. Initially, the USA intended to restrict the development of Japan. Under the ‘reparation policy’, steel production would be lowered to 2–3 Mt/year. The Korean crisis produced a change of policy: the USA sent engineers to teach the Japanese steelmaking.7 The Americans, most of them from US Steel, were excellent teachers – and the Japanese were even more eager students. The first stage consisted of the introduction of new technology. In the second stage steel manufacturing was

SEQ 0237 JOB IOM8627-014-005 PAGE-0237 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 237

Fig. 6 Hoogovens No. 1 strip mill (1953).

improved with discipline and patience, applying the ‘total quality’ approach developed in the USA by Deming. This resulted in such an advance (in the third stage) that other countries began to import Japanese technology. In Asia, Korea and China especially adopted Japanese technology. In Europe, The Netherlands, Italy, and France were leaders in importing Japanese steel manufacturing technology, soon followed by the UK. This advanced technology made possible higher productivity per man hour, thus greatly reducing the numbers employed in the industry (Fig. 7). In the 1980s, the five big Japanese steel manufacturers took financial interests in almost all US steel producers and began to export new technology to the USA. In the 1950s, the French Foreign Minister Robert Schuman, another twentieth century man of steel, launched an initiative for European cooperation. His aim was to stimulate ‘world peace’ – la paix mondiale – by which he meant preventing new conflicts between France and Germany. The Treaty of Paris was signed in 1951 and its first manifestation, the European Coal and Steel Community (ECSC), was formed the next year: steel in the role of joining different nations and different cultures. ECSC formed the backbone of the later EEC and of today’s EU. The member states of ECSC contributed through a levy per tonne of steel produced. Part of the funding raised was used for social measures and for regional reconstruction of the industry; part was used by ECSC to organise a common R&D plan (Fig. 8). This has led to the present high technological level of the iron and steel industry in ECSC, benefiting customers, stakeholders, and also

SEQ 0238 JOB IOM8627-014-005 PAGE-0238 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

238

Hatfield Memorial Lectures Vol. III

Fig. 7 Production and employment in Japanese steel industry; 1970=100% (Ref. 8).

the environment and energy conservation. A network of specialists, researchers, and managers from industry, RTOs, and academia has been established.

THE END OF THE IRON AGE? According to Sze and Feldman, the Iron Age ended in 1968 (Fig. 9). In that year, the number of scientific publications having iron as the subject was equalled by the number of publications about silicon, since when silicon has forged ahead – clearly the glory days of iron belong to the past as far as scientific journals are concerned! Was it coincidental that around this presumed end of the Iron Age the UK Labour Government nationalised the British iron and steel industry? British Steel Corp. (BSC) was formed by incorporating the 14 largest steel companies in the UK. The highest production level reached was 26Mt in 1970, the total maximum capacity being 33 Mt/year. The energy crises of the 1970s led to rationalisation and reduction of the labour force. In 1988 British Steel was privatised, under a Conservative Goverment led by Margaret Thatcher, the ‘Iron Lady’ and my final selection as a name of the twentieth century strongly related to steel. Subsequently privatisation also took place in Italy and France. In 1972 a challenging and alarming report appeared, ‘Limits to growth – a report for the Club of Rome project on the predicament of mankind’, by Dennis Meadows, a professor at the Massachusetts Institute of Technology and leader of the System Dynamics Group.10 His group developed a mathematical model to predict the future availability of

SEQ 0239 JOB IOM8627-014-005 PAGE-0239 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 239

Fig. 8 European Coal and Steel Community R&D expenditure under Steel RTD programme (1955–99): source ECSC.

Fig. 9 The silicon age (1968– ): after Sze and Feldman9.

resources. Energy, food, water, space, and other vital resources would, they inferred, become scarce in the near future. The concept of exponential growth was explained and its intrinsic danger was made clear. This danger had to be turned into a global challenge. In this report, steel consumption was used as a measure of national prosperity; a clear relationship between gross national product and steel consumption per capita was established for the USA (Fig. 10). Over 80 years, the consumption of steel in the USA rose from 50 to 700 kg/head, a factor of 14, and gross national product per head from US$800 to US$3600, a factor of 4.5 (note the saturation effect). Taking this observation a little further, steel consumption was considered as a measure for the stage of development of a nation. Developing countries such as China and India were found at the lower end of the curve, whereas the former FRG, Sweden, and the

SEQ 0303 JOB IOM8627-014-009 PAGE-0240 227-246 BOM REVISED 22DEC04 AT 00:26 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

240

Hatfield Memorial Lectures Vol. III

Fig. 10 Relationship between GNP and steel consumption for USA (1890–1969): after Meadows10.

Fig. 11 International comparison of relationship between GNP and steel consumption in 1968: after Meadows10.

USA were at the upper end (Fig. 11). With steel consumption so strongly related to prosperity, the future availability of iron sources would be of paramount importance to developing countries. Here the Club of Rome was not alarming in its conclusions. Steel was not considered a big problem; iron ore is abundantly available. Moreover, the mathematical formulae of Meadows did not incorporate much recycling. Only the recycling potential of steel, more or less securing future availability, was mentioned. Resources of other metals such as tin, however, were predicted to be completely exhausted within a few decades.

SEQ 0241 JOB IOM8627-014-005 PAGE-0241 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 241 In 1973, just a few months after the publication of the Club of Rome report, the world was faced with an energy crisis, followed by a second one in 1979. A sudden rupture of all trends was experienced. Nations in control of natural oil and gas sources proved able to exert control over the industrial world simply by disrupting the supply of these fossil fuels. At first, the steel industry profited from this dramatic oil crisis as new sources were sought. Steel provided the means and boundary conditions for drilling and dragging, for exploration and exploitation, for distilling and refining, for transport, storage, and distribution. Drilling in the North Sea provided prosperity to Scotland and Norway where offshore supply bases were set up. Exploitation of natural gas generated prosperity for the Netherlands. Distilling and refining initially brought employment. Storage in parks of huge steel tanks provided strategic reserves. Mammoth tankers, built of steel in Japan and later in Korea, were used to transport crude oil. Transport of sour gas was made possible by development of special steel grades for large diameter pipes. The steel industry was able to deliver pipelines for Alaska and Siberia that could withstand severe Arctic conditions. Steel wherever you look, hidden and yet visible to those who realise its importance. The alarming words of the Club of Rome and the two energy crises made clear that energy intensive industries like the iron and steel industry faced a great challenge. The introduction of new technologies such as continuous casting of slabs, blooms, and billets, direct rolling of hot slabs in the rolling mill, and maximising use of production gases and surplus heat has brought tremendous energy savings. Since 1975, the average energy consumption per tonne of steel for the European steel producers that are members of Eurofer has been reduced by almost 40%. The best technology available gives even higher energy reduction figures for individual plants. When the effects of the energy crisis hit the European steel industry, the economic and social roles of the ECSC again became apparent. A production quota system was announced and controlled by the ECSC authorities. After long deliberations a state of emergency was declared and a reduction of production capacity agreed on in 1983. In 1986 the reductions in production capacity had reached 13% for the former FRG, almost 20% for Belgium, roughly 20% for Italy, about 23% for France and the Netherlands, and 24% for the UK. This had an impact on employment (Fig. 12); in the EU the number of jobs in the iron and steel industry fell by more than 60% between 1975 and 1993. The labour force in the UK suffered most. The emergency measures for the steel industry in the ECSC were withdrawn in 1988. The ECSC Treaty will expire in 2002. Given the success of collaboration in R&D, the European steel industry is seeking a way to continue the shared research in iron and steel to improve its position in the marketplace.

CHANGING MATERIALS MARKET Although steel is still the major construction material, its relative role in the materials market place is changing incessantly. It is clear from Fig. 13, showing saturation in global

SEQ 0277 JOB IOM8627-014-007 PAGE-0242 227-246 BOM REVISED 20DEC04 AT 22:25 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

242

Hatfield Memorial Lectures Vol. III

Fig. 12 International comparison of employment in steel industry (1975–95); 1975=100% (Ref. 8).

crude steel production, that steel is not following the growth in world population and thus steel consumption per capita is falling. However, this reflects increased materials competition and technological improvements in the manufacture and application of steel. Far less crude steel is needed to manufacture products that have high customer demands. Certainly plastics have become a threat to steel, but to form plastic products steel moulds are needed: steel shaping the competition. Aluminium has become another serious competitor, notably because of its light weight and visual presence in high tech applications such as aircraft and aerospace vehicles. But no aerospace production is possible without the steel machining equipment, the cranes and heavy lifting equipment in the factories, and the steel frameworks of the large hangar type buildings, let alone the transport to the launch platform and the launching equipment itself. Aluminium wings keep aeroplanes airborne, but steel landing gear guarantees a safe landing. However, these steel items do not share the glamour of the other metals. Steel is fighting back and has recently shown its great innovative potential in lightweighting of vehicles. A group of 35 steel manufacturers joined forces under the umbrella of the

SEQ 0243 JOB IOM8627-014-005 PAGE-0243 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 243

Fig. 13 World crude steel production in the twentieth century: source IISI.

Fig. 14 Part of Kurushima Kaikyo bridges linking Japanese islands Hunshu and Shikoko.

International Iron and Steel Institute in the Ultra Light Steel Automotive Body project. By combining innovative holistic design and the improved properties of newly developed steel grades, it was proven that the weight of the body in white could be reduced by 25%. Stiffness, crash resistance, and other required properties were equal to those of conventional designs or better. A further striking example of how the development of new steel grades has enabled advances in technology is found in civil construction and notably in bridges. Today’s record

SEQ 0244 JOB IOM8627-014-005 PAGE-0244 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

244 Hatfield Memorial Lectures Vol. III holder for single span suspension bridges is the 3910 m long Akashi Strait Bridge, one of three links between the Japanese islands Honshu and Shikoko. It has a central span of 1990 m carried by two 1.1 m diameter cable bundles. But perhaps the greatest achievement of all is the middle link, which is a sequence of three linked suspension bridges (Fig. 14). In June 1999 British Steel plc and Koninklijke Hoogovens NV announced a merger to form Corus Group plc, based in London. Corus is a splendid example of the role of steel in the materials market; not on its own, but together with other metals, establishing a profitable position in a challenging global marketplace. With a new market approach, a sound international manufacturing base, and a highly capable product application focused R&D organisation, the company will set the tone in the materials market of the next century. No longer is tonnage the important parameter; customer focus, market share, profitability, and shareholder values are decisive. This approach differs fundamentally from the path inventors and steelmaking pioneers were able to adopt, as does the technical, economic, and social environment in which industrial operations are taking place. From being a motor and undisputed leader of economic development, steel has become an essential part of daily life in the knowledge oriented society at the turn of the twenty-first century. In the same way, from being an almost universal construction material steel has become a member of a large family of materials, each with its own specific merits. But despite the introduction of many other materials, steel remains an indispensable basis of modern society, indispensable to such an extent that it almost seems ‘natural’ and that many of us never actively think of how far it has penetrated daily life, including arts and leisure. And so it undoubtedly will remain for many years to come.

ACKNOWLEDGEMENTS Support from my brother Stefan and my son Joeri, both graduates in history from the University of Amsterdam, is gratefully acknowledged. Thanks also to Dick Hamels, Manager Communication and Publicity at Ijmuiden, who provided great assistance, and to the Corus companies, The Institute of Materials, the International Iron and Steel Institute, and ECSC for supplying data and illustrative material on the iron and steel industry.

REFERENCES 1. C. BODSWORTH (ed.): ‘Sir Henry Bessemer: father of the steel industry’; 1998, London, loM Communications Ltd. 2. W. NIEUIVENHUYS: ‘Company on the move: seventy-five years of Hoogovens’; 1993, Ijmuiden, Hoogovens Groep. 3. S. KOTKIN: ‘Steeltown, USSR: Soviet society in the Gorbachev era’; 1991, Berkeley, CA, University of California Press.

SEQ 0245 JOB IOM8627-014-005 PAGE-0245 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Iron, the Hidden Element 245 4. A. MALENKY: ‘The metallurgical combine of the future’; 1932, Moscow. (Quoted in S. KOTKIN: ‘Magnetic mountain: Stalinism as a civilisation’; 1995, Berkeley/Los Angeles, CA, University of California Press.) 5. H. UEBBING: ‘Stahl schreibt Geschichte: 125 Jahre Wirtschaftsvereinigung Stahl’; 1999, Dusseldorf, Verlag Stahleisen. 6. D. F. M. F. JACOBS: ‘Gereguleerd staal’ (‘Regulated steel’), doctoral thesis, University of Nijrnegen, The Netherlands, 1988. 7. E. ABE and Y. SUZUKI (eds.): ‘Changing patterns of international rivalry’; 1991, Tokyo, University of Tokyo Press. 8. S. MIONOV: in ‘The steel industry in the new millennium’, Vol. 2, (ed. S. Ranieri and E. Gibellieri), 17–28; 1998, London, loM Communications Ltd. 9. F. W. SARIS: FOM, personal communication. 10. D. MEADOWS: ‘Limits to growth – a report for the Club of Rome project on the predicament of mankind’; 1972, Cambridge, MA, Massachusetts Institute of Technology.

SEQ 0246 JOB IOM8627-014-005 PAGE-0246 227-246 BOM REVISED 19DEC04 AT 21:02 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0300 JOB IOM8627-901-005 PAGE-0247 END MATTER REVISED 21DEC04 AT 23:47 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Author Index Barnes, R. S., 63–87 Boom, R., 227–245 Cartwright, W. F., 183–199 Coche, L., 89–95 Colclough, T. P., 167–181 Cottrell, A., 201–205 Edington, J., 101–122 Finniston, H. M., 25–43 Flemings, M. C., 97–99 Pettifor, M. J., 123–142 Pl¨ockinger, E., 45–62 Richardson, F. D., 3–24 Scholey, R., 207–225 Sellars, C. M., 143–164

247

SEQ 0006 JOB IOM8627-901-007 PAGE-0248 END MATTER REVISED 06FEB05 AT 21:39 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

SEQ 0732 JOB IOM8627-400-008 PAGE-0249 S.INDEX REVISED 06FEB05 AT 20:44 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Subject Index Ajax process, 38, 128 aluminium subhalide process, 15 thermal extraction, 16 asymmetric beam, 135 attributes and properties, 144

Corex process, 104 Corus Bi-steel, 137 Corus Group plc, 244 corrosion resistance, 42 counter-gravity casting, 98 deformation, during rolling, 134, 157 Deming, 237 dendritic growth, 98 direct strip casting, 129 Dorman, Arthur, 35

basic oxygen processes, 126 basic steelmaking, 6 beam blank casting, 110 Bessemer, Henry, 4, 12, 228 Blaenavon works, 5 blast furnace, 70, 103, 126, 168, 233 injection into, 126 blast protection, buildings, 137 Blue Riband, 233 BOF slag, uses of, 107 Bohr, Niels, 123 BOS process, 104, 183, 190 bridges, 136, 243 British Steel Corporation, 114, 238, 244 Brown, John, 231 Brunel, 228 buildings, design and construction of, 117, 140

earthquake protection, 137 EEC, 219, 232, 237 Eiffel Tower, 228 electric are furnace, 105, 126, 128, 193 electric arc steelmaking, 105, 194 electroslag refining, 39 electroslag remelting, 45, 48 energy, 76, 108, 193, 202, 223, 230 environment, 74, 107, 218 ESR process, 48 steel, mechanical properties of, 56 European Coal and Steel Community, 61, 63, 65, 73, 237, 241 exports, 211

cans, 115 capital costs, 196 Carnegic, Andrew, 229 Challenor, George, 4 coatings, protective, 220 commodity markets, for steel, 220 composite materials, 220, 225 construction knowledge based, 140 materials and systems, 117 share in GDP, 211 continuous casting, 39, 98, 129, 131 continuous steelmaking, 40, 93 controlled rolling, 90

Ferrolite, 120 fire protection, 118, 136 Ford, Henry, 230 Forth bridge, 227, 229 fracture mechanics, 29 frequency distribution, 145 fuel-oxygen-scrap processes, 38 Genetic Algorithm Optimiser, 160 Guest Keen Nettlefolds and Guest Keen Baldwin, 171 hairline cracks, 29

249

SEQ 0728 JOB IOM8627-400-006 PAGE-0250 S.INDEX REVISED 06FEB05 AT 19:12 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

250

Hatfield Memorial Lectures Vol. III

Hall, 4 Heroult, 4 Hoogovens, 232 hot torsion test, 54, 90 hot working definition of, 143 of ESR material, 54 hybrid modelling, 160, 162 hydrogen reduction process, 9, 18 hydrometallurgy, 17 information technology, 111 ingots properties of, 46 ESR, structure in, 54 investment casting, 97 Iron Age, 238 Iron Curtain, 235 ironmaking, 35, 233 iron ore sources of 67, 72, 75 trade in, 218 ‘just in time’ delivery, 219 knowledge based construction, 140 knowledge based modelling, 150, 160 Krupp, Alfred, 229, 231 labour costs, 216 labour productivity, 221 ladle refining, 129 LD and LDAC processes, 11 Lancashire Steel Corporation, 171 laser cutting and welding, 114, 138 Liberty ships, 131, 138, 235 lost foam casting, 97 magnesium, thermal extraction, 14 Magnitogorsk, 234, 235 markets for materials, 101, 225, 242 for steel, 66, 126, 128, 216 Marshall Plan, 236 Martin, 228

Meadows, Dennis, 238 metals, consumption of, 124, 202 metallurgy, anatomy of, 7, 22 metallurgy in modelling, 147 microalloyed steel, 148 minimills, 131, 183 modelling, mathematical, 98, 133, 143 aims of, 145 methodology, 149, 159 near net shape casting, 99, 130 Nippon Steel, 229 nuclear energy, 81 open hearth furnace, 38, 128, 168, 177 overheads, 195 oxygen steelmaking, 12, 104 packaging materials and design, 115 Paris Treaty, 208, 237 pig iron, production, 170, 175, 233 pilot plants, 32, 93 phosphoric slag, 6 plant, capacity of, 186 powder, production of, 19 process innovation, 126 profitability, 184, 244 protective coatings, 220 raw materials, 75 cost of, 189 sources, 67 recrystallization, 90 behaviour, 147, 160 time, 152 recycling, 108 research, 25, 63, 95, 179, 217 application of, 89, 94 committees, 26, 98 organisation, 82, 89 strategy, 84 topics for, 73 Richard Thomas & Co Ltd, 172 rolling controlled, 90, 132

SEQ 0729 JOB IOM8627-400-006 PAGE-0251 S.INDEX REVISED 06FEB05 AT 19:12 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS

Subject Index mills, 168, 178 simulation of, 92, 134, 148, 155 Rome Treaty, 208 Schuman, Robert, 237 scrap, price of, 183, 189 section rolling, 133 segregation, 98 semi-solid casting, 98 Sheffield, 207, 217 Sherritt-Gordon Process, 17 ships, 231 Siemens, 228 sinter, production of, 176 slag-metal reactions, ESR, 49 Snelus, George, 6 solidification research, 98 sound-deadened steel sheet, 121 Stalin, 234 steel consumption of, 102, 124, 208, 213, 239 employment in, 214, 241 energy, 76, 223 investment in, 215 markets for, 66, 101, 126, 128, 219 plants, 171 prices of, 35 processes, 103 product development and use, 72, 113, 188 production of, 33, 124, 128, 169, 173, 174, 176 trade in, 211 utilization, 72 wages and competition, 214 steel industry, location and pattern, 68, 71 steelmaking, 37, 176 steelworking, 40 Stewarts and Lloyds, 171

251

strain induced precipitation, 152, 153 strip casting, 110, 131 strip products, 208, 220 sub-models, 150 subhalide process, for aluminium, 15 Summers, John and Sons, 172 Swinden, Dr Thomas, 25 Telford, Thomas, 228 temperature during rolling, 134 Thatcher, Margaret, 238 thermomechanical controlled processing, 132, 143 thin slab casting, 109, 131 Third World steel, 215, 221 Thomas, Richard & Co, 172 Thomas, Sydney Gilchrist, 4, 23, 228 Thring, 40 transport, cost of, 189 Ultra Light Steel Automotive Body, 243 United Steel Companies, 172 US Steel Co., 229 V process, 97 vacuum melting and degassing, 39, 40, 47 vacuum metallurgy, 46, 47, 59 Vickers, 231 virtual microstructures, 161 wire drawing, 135 wire production, 135, 171 welding techniques, 138 work hardening, 147 Worner, Howard, 40 Yawata Steel, 229 zinc-lead blast furnace, 8 zinc smelting, 9

SEQ 0730 JOB IOM8627-400-007 PAGE-0252 S.INDEX REVISED 06FEB05 AT 19:15 BY TF DEPTH: 62.04 PICAS WIDTH 45.03 PICAS