Burrs – Analysis, Control and Removal
Jan C. Aurich · David Dornfeld Editors
Burrs – Analysis, Control and Removal Proceedings of the CIRP International Conference on Burrs, 2nd–3rd April, 2009, University of Kaiserslautern, Germany
123
Editors Prof. Jan C. Aurich University of Kaiserslautern Institute for Manufacturing Technology and Production Systems - FBK Gottlieb-Daimler-Straße 67663 Kaiserslautern Germany
[email protected] Prof. David Dornfeld University of California at Berkeley Department of Mechanical Engineering 5100A Etcheverry Hall Berkeley, CA 94720-1740 USA
[email protected]
ISBN 978-3-642-00567-1 e-ISBN 978-3-642-00568-8 DOI 10.1007/978-3-642-00568-8 Springer Heidelberg Dordrecht London New York Library of Congress Control Number: 2009926960 © Springer-Verlag Berlin Heidelberg 2010 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: eStudio Calamar S.L. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
Preface
The International Conference on Analysis, Control and Removal of Burrs, held at the University of Kaiserslautern, Germany, marks the end of the three-year cycle of the CIRP working group on Burrs. During these years, researchers as well as experts from industry have contributed and gathered knowledge on research and industrial practice in the area of burrs, edge conditions and part cleanliness. A CIRP round robin has been carried out with participants in many countries and three continents to compare available methods on burr measurement. This conference now brings together the members of the working group as well as experts from industry and academia from all over the world to discuss the state of the art in research as well as industrial applications in the area of burrs in one focused conference with a special workshop character. The conference shall provide a forum for the intense exchange of concepts and methods, the dissemination of technological breakthroughs, and for discussions of future directions of research and development. The conference program covers a broad spectrum of topics ranging from standards for burr description and classification and the mechanics of burr formation over modeling and simulation of the underlying mechanisms to burr control and deburring strategies. Our special appreciation goes to the invited keynote speakers: Prof. Dornfeld from UC Berkeley, USA, one of the leading experts in the field, will give an overview of the state of research in analysis, control and removal of burrs. Prof. Biermann from TU Dortmund, Germany, will draw on his scientific and industrial experience to present burr reduction and control strategies. Mr. Berger from Daimler, Germany, and Dr. Martinsen from RTIM, Norway, will give an insight into the industrial importance of burr issues and part cleanliness. Our special gratitude goes to the International Program Committee and Local Organizing Committee members for their wonderful efforts in reviewing papers, handling papers, and preparing the technical and social program. Without their effort and dedication the conference would not have been possible. We would also like to extend our sincere appreciation to the paper authors for their excellent contributions to the conference. The authors who are willing to share their most recent findings in basic research as well as in industrial application, both in presentations and in many discussions in and around the technical sessions, represent the dominant factor in the success of this conference. Finally, the organizations and companies who contributed to the financial support of the conference, even in financially difficult times, deserve our great respect for their contribution. We would also like to thank the companies who opened their facilities for the industrial tour. They allowed giving this conference the final touch of strong interaction between research institutions and industry.
v
vi
Preface
On behalf of the organizing committee, I wish all of you a successful conference, with exciting technical sessions, fruitful discussions and a perfect get together of our research community. Jan C. Aurich Kaiserslautern, Germany
David Dornfeld Berkeley, CA, USA
Contents
Keynotes A Review of Burr Formation in Machining . . . . . . . . . . . . . . . . . . . . . . D. Dornfeld and S. Min
3
Burr Minimization Strategies in Machining Operations . . . . . . . . . . . . . . . D. Biermann and M. Heilmann
13
Burr Formation and Avoidance for Robust Circular Blade Sawing of Thin Walled Extruded Aluminum Profiles . . . . . . . . . . . . . . . . . . . . . . . . . . K. Martinsen and G. Ringen
21
Mechanics, Modeling and Simulation of Burr Formation Burr and Cap Formation by Orbital Drilling of Aluminum . . . . . . . . . . . . . E. Brinksmeier and S. Fangmann
31
Cutting Force Model for Analysis of Burr Formation in Drilling Process . . . . . . T. Matsumura and J. Leopold
47
Burr Formation in Microstructuring Processes . . . . . . . . . . . . . . . . . . . . B. Denkena, L. de Leon, and J. Kästner
55
Analytical Modeling and Experimental Investigation of Burr Formation in Grinding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H. Sudermann, I.G. Reichenbach and J.C. Aurich Developing a Process Model for Abrasive Flow Machining . . . . . . . . . . . . . . E. Uhlmann, V. Mihotovic, H. Szulczynski, and M. Kretzschmar Modeling and Simulation of Burr Formation: State-of-the-Art and Future Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. Leopold and R. Wohlgemuth
63 73
79
Burr and Chip Formation Mechanisms Interfacial Burr Formation in Drilling of Stacked Aerospace Materials . . . . . . S.N. Melkote, T.R. Newton, C. Hellstern, J.B. Morehouse, and S. Turner
89
Burr Formation in Drilling Intersecting Holes . . . . . . . . . . . . . . . . . . . . L. Leitz, V. Franke, and J.C. Aurich
99
Chip Breakage Prediction by a Web-based Expert System . . . . . . . . . . . . . . F. Klocke, D. Lung, and C. Essig
107
vii
viii
Contents
Parameters with Influence on Burr Formation Size Effects in Drilling Burr Formation . . . . . . . . . . . . . . . . . . . . . . . . R. Neugebauer, G. Schmidt, and M. Dix Burr Formation and Surface Characteristics in Micro-End Milling of Titanium Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G.M. Schueler, J. Engmann, T. Marx, R. Haberland, and J.C. Aurich
117
129
Influence of Minimum Quantity Lubrication on Burr Formation in Milling . . . . U. Heisel, M. Schaal, and G. Wolf
139
Burr Formation and Removal at Profile Grinding of Riblet Structures . . . . . . . B. Denkena, L. de Leon, and B. Wang
147
Burr Measurement Burr Measurement System for Drilled Hole at Inclined Exit Surface . . . . . . . . H.P. Hoang and S.L. Ko
157
Burr Measurement: A Round Robin Test Comparing Different Methods . . . . . V. Franke, L. Leitz, and J.C. Aurich
167
Deburring Processes – Fundamentals Deburring with CO2 Snow Blasting . . . . . . . . . . . . . . . . . . . . . . . . . . E. Uhlmann, M. Kretzschmar, F. Elbing, and V. Mihotovic
181
A Study on Deburring Inconel 718 Using Water Jet Technology . . . . . . . . . . . F. Boud, J. Folkes, N. Tantra, S. Kannan, and I.W. Wright
189
Ice Blasting – An Innovative Concept for the Problem-Oriented Deburring of Workpieces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Karpuschewski and M. Petzel
197
Deburring Processes – Applications Study of Internal Deburring of Capillary Tubes with Multiple Laser-machined Slits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . H. Yamaguchi and J. Kang Robotic Deburring Based on On-line Burr Measurement . . . . . . . . . . . . . . L. Liao, F. Xi, and S. Engin Deburring Machine for Round Billets – Equipment for Efficient Removal of Burrs from Billets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . M. Schnabl
205 213
221
Removal and Cleanability Formulation of the Chip Cleanability Mechanics from Fluid Transport . . . . . . S. Garg, D. Dornfeld, and K. Berger Burr Minimization and Removal by Micro Milling Strategies or Micro Peening Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Kienzler, M. Deuchert, and V. Schulze
229
237
Assessment of Deburring Costs in Industrial Case Studies . . . . . . . . . . . . . . P.J. Arrazola
245
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
253
A Short View on CIRP
CIRP was founded in 1951 with the aim to address scientifically, through international cooperation, issues related to modern production science and technology. The International Academy for Production Engineering takes its abbreviated name from the French acronym of College International pour la Recherche en Productique (CIRP) and includes ca. 550 members from 41 countries. The number of members is intentionally kept limited, so as to facilitate informal scientific information exchange and personal contacts. CIRP has some 170 Fellows and Honorary Fellows who are internationally recognized scientists elected to be CIRP members for life. Due to the limited number of CIRP Fellows, the election of a Fellow is a lengthy, rigorous process ensuring the highest possible academic standards. CIRP includes some 130 associate members, well known scientists, with high potential, elected typically for a period of three years with the possibility of renewal. A number of Associate members eventually become Fellows. Some Associated members may also belong to fields related to Manufacturing. CIRP, although an academic organization, encourages the participation of industry in its activities. There are ca. 140 corporate members who follow the research work of the academic members of CIRP, and very often contribute to the information exchange within CIRP by presenting their views on industrial needs and perspectives. Invited members, particularly from countries not yet involved in CIRP, are also included in the CIRP community. In a recent development, there is work under way to establish a CIRP Network of young scientists active in manufacturing research. CIRP aims in general at: • Promoting scientific research, related to • • • •
manufacturing processes, production equipment and automation, manufacturing systems, and product design and manufacturing.
• Promoting cooperative research among the members of the Academy and creating opportunities for informal contacts among CIRP members at large. • Promoting the industrial application of the fundamental research work and simultaneously receiving feed back from industry, related to industrial needs and their evolution. • Organizing an annual General Assembly with keynote and paper sessions and meetings of the Scientific and Technical Committees, publishing papers, reports, annals and other technical information, organizing and sponsoring international conferences. CIRP is organized along the lines of a number of Scientific Technical Committees (STCs) and Working Groups (WGs), covering many areas. CIRP organizes annually a General Assembly and the so called January Meetings. In the General Assembly (GA), which lasts ix
x
for a week, there is an intensive technical program with over 140 technical paper presentations from different fields of manufacturing, a number of keynote papers, at the opening of the conference, as well as technical work within the STCs. In parallel, there is a social program, aiming at making the culture of the General Assembly site known to the members and also at creating an informal environment for information exchange among the members. The January meetings are always organized in Paris, and last three days. Moreover CIRP organizes, through its membership, a number of conferences, notably the Manufacturing Systems Seminar and a number of other conferences with relevant topics. CIRP members also organize a variety of conferences, under the sponsorship of CIRP. The main publications of CIRP are the CIRP Annals under ISI standards with two volumes; Volume I, with refereed papers presented in the GA by Fellows, Associate, Corporate and Invited members and Volume II with refereed keynote papers. There are also CIRP proceedings, including round table discussions, technical reports, special issues, reports and internal communications, proceedings of CIRP conferences, dictionaries of production engineering etc. A Newsletter is also published twice a year. Currently the CIRP Annals are published by Elsevier, while Springer Verlag publishes the Dictionaries of Production Engineering. There are under development one or more journals, complementing the work published in the CIRP Annals. The CIRP organization includes besides the President, who is elected annually, the Council and a number of other committees ensuring a continuous improvement of the CIRP organization and reflecting the changing needs of manufacturing science and technology. CIRP has its headquarters in Paris, staffed by permanent personnel and welcomes potential corporate members and interested parties in CIRP publication and activities in general. For further information please contact: CIRP Secretariat, 9 rue Mayran, 75009 PARIS, France Phone: ++33 1 45 26 21 80, Fax: ++33 1 45 26 92 15 e-mail:
[email protected] http://www.cirp.net/secretariat/secretariat.html
A Short View on CIRP
Organization
Conference Chairman Aurich, J.C., University of Kaiserslautern Co-Chairman Dornfeld, D., University of Berkeley Heisel, U., University of Stuttgart Organizing Committee Aurich, J.C., University of Kaiserslautern Franke, V., University of Kaiserslautern Herzenstiel, P., University of Kaiserslautern Leitz, L., University of Kaiserslautern Mannweiler, C., University of Kaiserslautern Schleret, R., University of Kaiserslautern Tuncay, S., University of Kaiserslautern International Scientific Committee Abele, E. (Germany) Altena, H. (Netherlands) Arrazola, P. (Spain) Bouzakis, K. (Greece) Brinksmeier, E. (Germany) Byrne, G. (Ireland) Childs, T. (UK) Denkena, B. (Germany) Heisel, U. (Germany) Ko, S. (Korea) Leopold, J. (Germany) Min, S. (USA) Teti, R. (Italy) Weinert, K. (Germany) Secretariat R. Schleret, S. Tuncay c/o Institute for Manufacturing Technology and Production Systems – FBK University of Kaiserslautern xi
xii
Gottlieb-Daimler-Straße, Gebäude 42 67663 Kaiserslautern Germany Phone: +49 631 205 2618 Fax: +49 631 205 3238 e-mail:
[email protected] url: www.fbk-kl.de
Organization
We Thank Our Sponsors
Adam Opel GmbH Opelkreisel 1-9 67663 Kaiserslautern
Deutsche Forschungsgemeinschaft (DFG) Kennedyallee 40 53175 Bonn
Land Rheinland Pfalz Staatskanzlei Rheinland-Pfalz Peter-Altmeier-Allee 1 55116 Mainz
Wipotec GmbH Adam-Hoffmann-Str. 26 67657 Kaiserslautern
xiii
Contributors
P.J. Arrazola Manufacturing Department, Faculty of Engineering, Mondragon University, 20500 Mondragon, Spain,
[email protected] J.C. Aurich Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany K. Berger Daimler AG, Material Technology Department, 70546 Stuttgart, Germany D. Biermann Institute of Machining Technology, Technische Universität Dortmund, Baroper Str. 301, 44227 Dortmund, Germany,
[email protected] F. Boud Department of Mechanical Materials and Manufacturing Engineering, University of Nottingham, Nottingham, NG7 2RD, UK,
[email protected] E. Brinksmeier Foundation Institute for Materials Science, Badgasteiner Str. 3, 28359 Bremen, Germany L. de Leon Institute of Production Engineering and Machine Tools, Leibniz Universität Hannover, An der Universität 2, 30823 Garbsen, Germany B. Denkena Institute of Production Engineering and Machine Tools, Leibniz Universität Hannover, An der Universität 2, 30823 Garbsen, Germany M. Deuchert Universität Karlsruhe (TH), Institut für Produktionstechnik, 76131 Karlsruhe, Germany M. Dix Institute for Machine Tools and Production Processes, Chemnitz, Germany,
[email protected] D. Dornfeld Mechanical Engineering Department, University of California, Berkeley, CA 94720-1740, USA F. Elbing CryoSnow GmbH, Zitadellenweg 20e, 13599 Berlin, Germany S. Engin Pratt & Whitney Canada Corp., 1000 Marie-Victorin, Longueuil, Quebec, J4G 1A1, Canada J. Engmann Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany C. Essig WZL, Laboratory for Machine Tools and Production Engineering, Aachen University, 52056 Aachen, Germany S. Fangmann Foundation Institute for Materials Science, Badgasteiner Str. 3, 28359 Bremen, Germany,
[email protected] J. Folkes Department of Mechanical Materials and Manufacturing Engineering, University of Nottingham, Nottingham, NG7 2RD, UK
xv
xvi
V. Franke Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany,
[email protected] S. Garg Mechanical Engineering Department, University of California, Berkeley, CA 94720-1740, USA,
[email protected] R. Haberland Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany M. Heilmann Institute of Machining Technology, Technische Universität Dortmund, Baroper Str. 301, 44227 Dortmund, Germany U. Heisel Institute for Machine Tools, Universität Stuttgart, P.O. Box 106037, 70049 Stuttgart, Germany,
[email protected] C. Hellstern George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, 801 Ferst Drive, Atlanta, GA 30332, USA H.P. Hoang Department of Mechanical Design and Production Engineering, Konkuk University, 1 Hwayang dong, Gwangjin gu, 143-701, Seoul, Korea J. Kang Department of Mechanical and Aerospace Engineering, University of Florida, P.O. Box 32611, Gainesville, FL 32611-6300, USA,
[email protected] S. Kannan Rolls-Royce plc, Derby, DE24 8BJ, UK B. Karpuschewski Institute of Manufacturing Technology and Quality Management, Otto-von-Guericke-University Magdeburg, P.O. Box 4120, 39016 Magdeburg, Germany,
[email protected] J. Kästner Institute of Production Engineering and Machine Tools, Leibniz Universität Hannover, An der Universität 2, 30823 Garbsen, Germany,
[email protected] A. Kienzler Institut für Werkstoffkunde I, Universität Karlsruhe (TH), 76131 Karlsruhe, Germany,
[email protected] F. Klocke WZL, Laboratory for Machine Tools and Production Engineering, Aachen University, 52056 Aachen, Germany,
[email protected] S.L. Ko Department of Mechanical Design and Production Engineering, Konkuk University, 1 Hwayang dong, Gwangjin gu, 143-701, Seoul, Korea,
[email protected] M. Kretzschmar Institute for Machine Tools and Factory Management (IWF), Technical University Berlin, Office PTZ 1, Pascalstr. 8-9, 10587 Berlin, Germany,
[email protected] L. Leitz Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany,
[email protected] J. Leopold Fraunhofer Institute for Machine Tools and Forming Technology, IWU Chemnitz, Reichenhainer Str. 88, 09126 Chemnitz, Germany,
[email protected] L. Liao Department of Aerospace Engineering, Ryerson University, 350 Victoria Street, Toronto, M5B 2K3, Canada D. Lung WZL, Laboratory for Machine Tools and Production Engineering, Aachen University, 52056 Aachen, Germany K. Martinsen RTIM AS, P.O. Box 2831, Raufoss, Norway, kristian.martinsen@rtim. raufoss.com T. Marx Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany
Contributors
Contributors
xvii
T. Matsumura Department of Mechanical Engineering, Tokyo Denki University, 2-2 Kanda Nishiki-cho, Chiyoda-ku, Tokyo 101-8457, Japan,
[email protected] S.N. Melkote George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, 801 Ferst Drive, Atlanta, GA 30332, USA,
[email protected] V. Mihotovic Institute for Machine Tools and Factory Management (IWF), Technische Universität Berlin, Office PTZ 1, Pascalstr. 8-9, 10587 Berlin, Germany,
[email protected] S. Min Laboratory for Manufacturing and Sustainability, Mechanical Department, University of California, Berkeley, CA 94720-1740, USA
Engineering
J.B. Morehouse Georgia Institute of Technology, Manufacturing Research Center, 813 Ferst Drive, Atlanta, GA 30332, USA R. Neugebauer Fraunhofer Institute for Machine Tools and Forming Technology, Chemnitz, Germany T.R. Newton George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, 801 Ferst Drive, Atlanta, GA 30332, USA M. Petzel Institute of Manufacturing Technology and Quality Management, Otto-vonGuericke-University Magdeburg, P.O. Box 4120, 39016 Magdeburg, Germany I.G. Reichenbach Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany G. Ringen RTIM AS, P.O. Box 2831, Raufoss, Norway M. Schaal Institute for Machine Tools, Universität Stuttgart, P.O. Box 106037, 70049 Stuttgart, Germany G. Schmidt Fraunhofer Institute for Machine Tools and Forming Technology, Chemnitz, Germany M. Schnabl framag Industrieanlagenbau GmbH, Neukirchnerstrasse 9, 4873 Frankenburg, Austria,
[email protected] G.M. Schueler Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany,
[email protected] V. Schulze Institut für Werkstoffkunde I; Institut für Produktionstechnik, Universität Karlsruhe (TH), 76131 Karlsruhe, Germany H. Sudermann Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany,
[email protected] H. Szulczynski Robert Bosch GmbH, Diesel Systems – Manufacturing Department 7, Engineering Nozzle BaP/MOE7, 96045 Bamberg, Germany N. Tantra Department of Mechanical Materials and Manufacturing Engineering, University of Nottingham, Nottingham, NG7 2RD, UK S. Turner Lockheed Martin Aeronautics Corporation, 86 South Cobb Drive, Marietta, GA 30063, USA E. Uhlmann Institute for Machine Tools and Factory Management (IWF), Technische Universität Berlin, Office PTZ 1, Pascalstr. 8-9, 10587 Berlin, Germany B. Wang Institute of Production Engineering and Machine Tools, Leibniz Universität Hannover, An der Universität 2, 30823 Garbsen, Germany,
[email protected]
xviii
R. Wohlgemuth TBZ-PARIV GmbH, Bernsdorfer Strasse 210-212, 09126 Chemnitz, Germany G. Wolf Institute for Machine Tools, Universität Stuttgart, P.O. Box 106037, 70049 Stuttgart, Germany I.W. Wright Rolls-Royce plc, Derby, DE24 8BJ, UK F. Xi Department of Aerospace Engineering, Ryerson University, 350 Victoria Street, Toronto, M5B 2K3, Canada,
[email protected] H. Yamaguchi Department of Mechanical and Aerospace Engineering, University of Florida, P.O. Box 32611, Gainesville, FL 32611-6300, USA,
[email protected]
Contributors
Keynotes
A Review of Burr Formation in Machining D. Dornfeld and S. Min
Abstract One of the major concerns in deburring technology is centered on how to predict the size and shape of burrs to insure uniform removal and, if this is possible, how to design the process or product in advance to minimize or control the burr size. This paper reviews some of the research done over the past several years on this important topic. The paper includes a discussion of burrs in conventional machining, process planning for burr minimization as well as micromachining applications.
understanding the creation of burrs and, specially, the material influence, data bases describing cutting conditions for optimal edge quality, and design rules for burr prevention as well as standard terminology for describing edge features and burrs. Ultimately, engineering software tools must be available so that design and manufacturing engineers can use this knowledge interactively in their tasks to yield a mechanical part whose design and production is optimized for burr prevention along with the other critical specifications. A review of the background to burr control is given first.
Keywords Burr · Machining · Size effects
1 Introduction 1.1 Motivation There has been a steadily increasing emphasis on enhanced quality of machined workpieces while at the same time reducing the cost per piece. Accompanying this is the decreasing size and increasing complexity of workpieces. This has put continual pressure on improvements in the machining process in terms of new processes, new tooling and tool materials, and new machine tools. Fundamental to this continual improvement is understanding edge finishing of machined components, especially burrs. Deburring, like inspection, is a non-productive operation and, as such, should be eliminated or minimized to the greatest extent possible. An understanding of the fundamentals of burr formation leads us to procedures for preventing or, at least, minimizing, burr formation. This depends on analytical models of burr formation, studies of tool/workpiece interaction for
D. Dornfeld (), S. Min Laboratory for Manufacturing and Sustainability, Mechanical Engineering Department, University of California, Berkeley, CA 94720-1740, USA e-mail:
[email protected] url: http://lmas.berkeley.edu
1.2 Introduction and Background Burrs in machined workpieces are complex and troublesome problems. They require additional finishing operations (deburring) and complicate assembly as well as risk damage to the part. Handling parts with burrs is a challenge for workers. In a perfect world we’d like to avoid, or at least minimize, burrs by careful choice of tools, machining parameters and tool path or work material and part design. The fact is that most burrs can be prevented or minimized with process control. Research and interest has been focused on problems associated with generation of burrs from machining for sometime but the focus has traditionally been on deburring processes. Understanding the burr formation process is critical to burr prevention. The level of scientific knowledge on burr formation is just in the early stages of development, see Fig. 1. The critical information, associating details of the part performance and functionality with requirements for edge condition, is still not well understood. Standards and specifications are only now being developed for this led by the German automotive and mechanical parts industries, see Berger, [1]. To effectively address burr prevention, the entire “process chain” from design to manufacturing must be considered, Fig. 2. Here we see the importance of integrating all the elements affecting burrs, from the part design, including material selection, to the machining process.
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_1, © Springer-Verlag Berlin Heidelberg 2010
3
4
D. Dornfeld and S. Min
Fig. 1 State of knowledge in burr formation
Fig. 2 Five level integration required for burr minimization, Dornfeld and Lee [2]
Burr formation affects workpiece accuracy and quality in several ways; dimensional distortion on part edge, challenges to assembly and handling caused by burrs in sensitive locations on the workpiece and damage done to the work subsurface from the deformation associated with burr formation. A typical burr formed on a metal component due to the exit of a cutting edge can range in shape and size from small and uniform (as in a “knife burr”) to rather large, nonuniform in shape and many millimeters in length. A number of things are clear from close inspection of burr images. There is substantial subsurface damage and deformation associated with a burr, the shape is quite complex and, hence, the description of a burr can be quite complex, and the presence of a burr can cause problems in manufacturing. More recently the problem of burrs and other machining and manufacturing related debris causing problems in the smooth functioning of precision mechanical devices has been addressed by a number
of researchers. This just adds additional importance to the understanding of burr formations. In fact, the range of burrs found in machining practice is quite wide, specially when the full range of processes from drilling to grinding is considered. To emphasize the point, Fig. 3 shows typical drilling burrs and their classification in stainless steel as an indication of the potential variation. Burrs in milling and turning exhibit wide variation as well. The costs associated with removing these burrs is substantial. The typical cost as a percentage of manufacturing cost varies up to 30% for high precision components such as aircraft engines, etc. In automotive components, the total amount of deburring cost for a part of medium complexity is in the range of 15–20% of manufacturing expenses. Industrial practice has shown that the actual investment in deburring systems increases with part complexity and precision.
A Review of Burr Formation in Machining
5
• inspection strategies for burr detection and characterization including specialized burr sensors • development of specifications and standards for burr description and measurement
Fig. 3 Three typical burrs in drilling stainless steel, Dornfeld and Lee [2]
A better strategy is to attempt to prevent of minimize, or prevent, burrs from occurring in the first place. This has two immediate benefits in that, first, it eliminates the additional cost of deburring the component and the likelihood of damage during the deburring process and, second, in the case burrs cannot be eliminated it improves the effectiveness of any deburring strategy due to reduced and more standard burr size and shape. This requires a comprehensive approach to burr prevention and minimization consisting of a number of components. To minimize or prevent burr formation requires that all stages of manufacturing from the design of the component through process planning and production be integrated so that the potential part features and material constraints, tooling and process sequences and process variables be considered from a perspective of the potential for creation of burrs on the workpiece, as seen in Fig. 2. That is, the inputs (process, material, tools, workpiece geometry, fixturing, etc.) must be considered along with the part functionality (part performance, fit and assembly requirements) as well as any expected or required deburring processes. This is most successful when clear standards and classifications are available, edge tolerances can be specified and the relationship between the edge quality and part functionality is clearly understood as shown in Fig. 1. This is not generally the case. The successful implementation of integrated burr control methodologies is necessary to overcome the limitations of burr issues in machining. The future development of comprehensive integrated strategies for burr minimization and prevention will depend on:
Specialized tooling for deburring is not discussed in this paper although that is an important area and is covered to some extent by commercial organizations today.
2 Process-Based Solutions 2.1 Introduction The models, databases and strategies mentioned above must be linked to the process of interest to be most effective. There are substantial differences between burr formation in drilling, milling and grinding, for example. In drilling, feed rate usually plays an important role in the development of drilling burrs. In addition, the drill geometry can affect the size and shape of the burr formed as well as prevent burr formation in some cases. Analytical models are increasingly supplemented with finite element method (FEM) models of the drilling process to predict effects of drill geometry, process parameters and workpiece characteristics on size and shape of the burr. Applications to aerospace component manufacturing, specially multi-layer structures and composite materials, is a primary area of focus for FEM drilling process modeling. In addition, the problem of burr formation in intersecting holes in precision components is well suited to analytical approaches for parameter selection and tool design. These approaches are also applicable to milling but less so due to the complexity of the milling process. Grinding geometry is typically more straight forward but the multiple abrasive “tools” with complex shapes complicates the analysis.
2.2 Milling • the continued development of predictive models with competent databases, including “expert data bases” for process specification • simulation models of burr formation capable of indicating the interaction and dependencies of key process parameters for burrs at all scales • strategies for burr reduction linked to computer aided design (CAD) systems for product design and process planning (and close coordination with CAD/CAM resource suppliers)
Since milling (specially face milling) figures so prominently in the manufacture of so many parts, for example, automotive engines and transmission components, it has been a major focus for burr reduction and prevention for many years. In milling, the kinematics of tool exits from the workpiece are a dominant factor in burr formation and, as a result, substantial success has been realized by adjusting the tool path over the workpiece, Fig. 4. The principal criteria in tool path determination have been:
6
D. Dornfeld and S. Min
θ
Fig. 4 Tool path strategies for minimizing and preventing burrs in face milling
• avoiding exits of inserts (or always machining on to the part edge) • sequencing of process steps to create any burrs on a last, less significant edge • control of exit order sequence (EOS) by tool geometry and path variation • maintaining uniform tool chip loads over critical features • lift and re-contact of milling cutter for some features where maneuverability is limited • avoiding “push exits” (those with long cutter path/edge contact length)
of the hole). The burr types illustrated in Fig. 3 are created by a sequence of events starting when the drill action first deforms the material on the exit surface of the workpiece through creation of the hole, Fig. 5. When intersecting holes are drilled, the specific orientation of the axis of the intersecting holes will have a tremendous effect on the location and creation of burrs around the perimeter of the holes. Figure 6 shows a schematic of burr formation in intersecting holes. Since the “exit angle” of the drill varies around the circumference of the hole intersection, the potential for burr formation will vary. This means that intersection geometry as well as tool geometries optimized to minimize adverse burr formation conditions can be effective in minimizing burr formation. Burr formation in intersecting holes shows high dependence on angular position under the same cutting conditions. Large exit angles, as seen in Fig. 6, yield small burrs. There is also a strong dependence on inclination angle (that is the degree of inclination of the intersecting hole from perpendicular.) Research shows that an inclination angle of 45◦ reduces burr formation. Further, research on drilling and intersecting hole challenges, Min [3], shows that the kinematics of edge exit sequence relative to the instantaneous geometry relationship between drill cutting edge and hole edge geometry can
While these criteria are often difficult to apply in all situations they have shown dramatic reductions in burr formation with the corresponding increases in tool life (tools are often changed when burr size reaches a specification limit) and reductions in deburring costs. In all circumstances cycle time constraints must be met with any redesigned tool paths as do surface finish and form criteria. With burr expert data bases for different materials and process parameters and the software for tool path planning, the possibility of designers being able to simulate the likely scenario of machining a component and any resulting problems with burrs as part of a conventional CAM software program is becoming a reality. These software systems must also be comprehensive enough to include other process steps and constraints so that other critical specifications are not compromised.
2.3 Drilling Burr formation in drilling is primarily dependent upon the tool geometry and tool/work orientation (that is, whether the hole axis is orthogonal or not to the plane of the exit surface
Fig. 5 Sequence of burr formation in hole drilling for uniform burr with cap
A Review of Burr Formation in Machining
7 Burr shapes on the workpiece in surface grinding 3 2
2
1
entrance burr
2
side burr
3
exit burr
1
Fig. 6 Schematic of burr formation in intersecting holes
predict the burr formation potential (basically when the vector sum of tool rotation and feed are in a “forward” direction relative to the feed motion). Hence, drill designs that minimize this forward vector for as much of the hole circumference as possible could be effective in burr minimization. Further, holes in multilayer materials offer additional challenges. This is specially true in aerospace applications where structures are often composed of “sandwich” configurations of metal, composite and sealant as found in advanced aerospace structures. Burr formation here is challenging as interlayer burrs often need to be removed before final assembly. Finite element analysis of these types of specific situations often offers increased understanding of the problems. When drilling multilayer material structures, the fixturing often plays an important role in determining the size and location of burrs. The gap that occurs between sheets during drilling provides space for burr formation at the interface of the two material sheets, see, for example, Newton et al. [4] and Choi et al. [5].
2.4 Grinding Research on burr formation in grinding is less well developed in terms of literature. Grinding burrs are complicated by the specialized removal mechanisms seen in grinding. Figure 7 shows the basic configuration of burrs in surface grinding, from [6], and a closeup of a an exit burr in grinding, sometimes referred to as a “Karpu” burr after the distinctive moustache of Professor B. Karpuschewski in Germany. Aurich et al. [6] summarized the results of early grinding burr research as follows: • superabrasive grinding leads to a significantly higher degree of burr formation • burr length and burr height are more influenced by the workpiece material than by the cutting parameters (for superabrasive grinding) • hybrid wheels (that is, combination of superabrasive and conventional features) seems to produce similar or
300 μm μm
Fig. 7 Typical burr shapes in surface grinding and closeup of “Karpu” burr, [6]
slightly less burr as the conventional grinding wheel; the abrasive has the dominant influence on burr formation
3 Examples of Application of Burr Minimization Strategies 3.1 Tool Path Planning in Milling One of the most successful areas of application of burr minimization strategies is in tool path planning for face milling. To a great extent, burr formation in milling can be prevented by adjusting the path of the milling cutter over the workpiece face. Specific cases have been evaluated in automotive engine manufacturing with major automobile companies. This can be extended to optimization of the process to insure that surface quality, including flatness, specifications are met or exceeded. Usually, burr size is used to indicate state of tool wear and when burr size increases beyond a predefined level the tool/insert is changed. Reduction in burr size (or increased number of parts produced before the burr size exceeds limits) directly relates to increased tool life and the accompanying reduction in tooling costs and tool change costs.
8
D. Dornfeld and S. Min
Figure 8 shows a conventional tool path for face milling a surface on a cast AlSi alloy automotive engine block. The presence of substantial burrs at critical locations required frequent tool changes as well as additional deburring operations. The optimized tool path using the criteria described above is shown in Fig. 9 and, in Fig. 10, shows the resulting burr free workpiece. Although the tool path is substantially longer in this example, it was possible to increase the feedrate without loss of surface finish to maintain the required 5 s cycle time for the process. The tool life (as a result of dramatically reduced burr formation) was increased by a factor of 3 with substantial resulting savings per machine/year. Other examples of these kinds of tool path planning improvements are available, [7–11]. All rely on a geometric
Fig. 10 Workpiece resulting from optimized tool path; Tool path length: old path – 209 mm, new path – 524 mm, cycle time (with increased feedrate) remains at 5 s
definition of part geometry, quantitative description of tool path, or tool exit/edge geometry or sequential geometrical relationship between tool cutting edges and part geometry on tool exit from the work linked to data on burr formation potential as a function of material properties (ductility and composition, for example).
Fig. 8 Conventional tool path for face milling engine block face and resulting burrs at key locations
Fig. 9 Modified tool path for part in Fig. 8
3.2 Drilling – Burr Control Chart Burr minimization and prevention in drilling is strongly related to process conditions (feedrate and speed, for example) and drill geometry. It is possible to represent the reasonable ranges of operating conditions for drilling by use of a “burr control chart” derived from experimental data on burr formation for varying speeds and feeds. This can be normalized to cover a range of drill diameters and, importantly, can be used across similar materials (carbon steels, for example). Data shows the likelihood of creating one of three standard burrs, as shown in Fig. 3, namely, small uniform (Type I), large uniform (Type II) and crown burr (Type III). Figure 11 below shows a typical burr control chart for 304L stainless steel. Continuous lines delineate different burr types. Type I is preferred. Burr height scales with distance from the origin. This burr control chart can be integrated with an expert system allowing queries of likelihood of burr formation to be shown on the control chart when information on drill diameter, speed, feed, etc. are input. Typical burr sizes expected are shown. An interesting example applying this burr control approach was described by Min and Dornfeld [12] for an
A Review of Burr Formation in Machining
Fig. 11 Drilling burr control chart for 304L stainless steel material showing normalized speed, s (vertical axis) vs. normalized feed, f (horizontal axis), d is drill diameter. Minimized burr conditions are indicated in crosshatched region
9
the process operation location on the Drilling Burr Control Chart (DBCC) (that is, in the “speed vs feed” domain), the DBES suggested new process parameters and burr size prediction, Fig. 12, top. In order to minimize burr formation, new cutting conditions that meet cycle time constraint were chosen and its location on the DBCC is shown in Fig. 12, bottom. The DBES did not have data for the workpiece material for the part (AISI 5046). But from the perspective of burr formation, the behavior of AISI 5046 is similar to low alloy steel, AISI 4118, which is available in the DBES. In general, alloy steels behave in a similar fashion in terms of burr formation. Hence, burr sizes were estimated from the DBCC for AISI 4118 and then, scaled linearly using a scale factor for AISI 5046. The scale factor, Sf , was obtained using measurements of the burrs on holes machined with new drills using current cutting conditions as the ratio of estimated burr sizes for 4118 to measured burrs size for 5046.
3.3 Burrs in Precision Machining As described above, various problems such as surface defects, poor edge finish, and burrs in conventional machining have plagued conventional manufacturing for some time. These problems are also significant in micromachining and require much more attention because, in many cases, inherent material characteristics or limitations in part geometry do not allow some of the solutions used in macromachining. Figure 13 shows some typical defects in micromilling including burrs on the edges of the slots.
Fig. 12 Application of burr control chart in automotive application, [12]; top – DBES output for original machining conditions; bottom – suggested conditions with parameters adjusted for cycle time constraints
automotive application. Using the Drilling Burr Expert System© (DBES) [3], cutting conditions were tested in order to obtain small burrs in a situation previously creating unacceptably large burrs leading to shorter tool life. By moving
Fig. 13 Typical surface defects in micromilling (work material: NiTi shape memory alloy), from Weinert et al. [13]
10
NiTi work material is used for many medical applications, such as surgical implants, and micromilling is commonly used to fabricate these products. This material is very ductile and easily work hardens during machining causing adhesion and high burr formation. Additionally, high ductility causes adverse chip formation, long and continuously snarled chips. At the micro level, these chips interfere with tool engagement and burrs and contribute to poor surface quality of finished parts. As part of a study on the burr formation in micromachining Lee and Dornfeld [14] conducted micro-slot milling experiments on aluminum and copper and found various standard burr types depending on location and work geometry. Interestingly, these burr shapes were similar to those found in macromachining in terms of formation mechanisms and influence of cutting parameters. One major difference found was that the influence of tool run-out on burr formation was significant in micro-slot milling. Min et al. [15] conducted micro-fly cutting and microdrilling experiments on single crystal and polycrystalline OFHC copper in order to understand the effects of crystal orientation, cutting speed, and grain boundaries on surface roughness, chip formation, and burr formation. Certain crystallographic orientations were found to yield rougher surface finish, as well as significant burrs and breakout at the tool exit edge. The <100> and <110> direction of machining on the workpieces exhibited the greatest amount of variation in formation of burrs and breakout at the exit edge and in chip topology as a function of the angular orientation of the workpiece. This corresponded to a variation in the interaction between the tool and the active slip systems. They also conducted slot milling experiments on the same material and found a strong dependency of top burr formation on slip systems of each crystal orientation except (100) workpiece. Bissacco et al. [16] found that top burrs are relatively large in micromilling due to the size effect. When the ratio of the depth of cut to the cutting edge radius is small, high biaxial compressive stress pushes material toward the free surface and generates large top burrs. Ahn and Lim [17], Ahn et al. [8] proposed a burr formation model in a microgrooving operation based on a side shear plane and an extended deformation area which is caused by the tool edge radius effect. The material near the cutting edge experiences the side shear deformation due to hydrostatic pressure. Aluminum and OFHC generated larger burrs than brass, and thus it was concluded that the thickness of the burr is proportional to the ductility of the material. Further work by Schaller et al. [19] showed that when fabricating microgrooves in brass, burr formation can be drastically reduced by coating the surface with cyanacrylate. Sugawara and Inagaki [20] investigated the effect of drill diameter and crystal structure on burr formation in microdrilling. They utilized both single crystal and polycrystalline iron
D. Dornfeld and S. Min
(a) 10 mm/min, 7000 rpm
(b) 5 mm/min, 6000 rpm
(c) 5 mm/min, 7000 rpm
(d) 10 mm/min, 8000 rpm
Fig. 14 Microdrilling burr formation (250 μm diameter); (a) burr within grain boundary, (b) burr across grain boundary, (c) burr over small grain, (d) grain boundary follows burr topology, from Min et al. [15]
with a thickness between 0.06 and 2.5 mm and high speed twist drills with diameters from 0.06 to 2.5 mm. In general they confirmed that burr size is reduced and cutting ability increased as drill size decreases. Min et al. [15] found that grain orientation affected burr formation in drilling of polycrystalline copper, Fig. 14. A single material may produce a ductile-like cutting mode in one grain and brittle-like cutting in another, indicating that favorable and non-favorable cutting orientations for good surface and edge condition exist as a function of crystallographic orientation. Additional micro-drilling research indicated that the effects seen at a larger, macro, scale (transition from uniform to crown burr with feed increase and basic similarity in burr shapes) also holds for microdrilling [21]. This means the drilling burr control chart concept could be applied at this scale also.
4 Summary and Conclusions Although edge finishing in machined components is a constant challenge in precision manufacturing of mechanical components, there are a number of strategies, built on competent process models and extensive data bases, that can substantially minimize or eliminate burrs. These strategies, some illustrated above, can be incorporated in the software relied
A Review of Burr Formation in Machining
upon by design and manufacturing engineers in their normal activities to insure that the conditions which can lead to burr formation can be avoided while insuring that production efficiency is maintained. This is part of the development of the “digital factory.” Recent experience indicates that the basis for this process optimization may also yield increases in throughput due to decreases in cycle time thanks to optimum part orientation on the machine during machining. In situations where burrs cannot definitely be eliminated there is the possibility, using these tools, to at least control their size over a range of conditions so that commercial deburring techniques are more reliably implemented – techniques such as abrasive filament brushes, for example. Finally, the inclusion of design rules for burr minimization will allow the design engineers to reduce the likelihood of edge defects at the most effective stage – during product design. Future work on burr prevention must focus more on tool design. The potential for substantial improvement, especially in drilling, will depend on analysis of drilling burr formation with the objective of optimization of tool drill design. It may be some time before it is possible to prevent all burr formation during the machining of mechanical components. But, in the meantime, there is much that can be accomplished towards that goal using the techniques and systems discussed in this paper to produce parts with higher edge precision.
11
6.
7.
8.
9.
10.
11.
12.
13.
14. Acknowledgments This work is supported by the Industrial Affiliates of the Laboratory for Manufacturing and Sustainability (LMAS)/CODEF, Machine Tool Technology Research Foundation (MTTRF) and NSF GRANT DMI-20062085 from Manufacturing Machines and Equipment Program. Additional details on this work can be found at lmas.berkeley.edu.
15.
16.
References 1. Berger, K., “An overview of status and trends in the automotive industry,” Proc. 7th Int’l Conference on Burr Formation and Surface Finishing, University of California, Berkeley, June, 2004. 2. Dornfeld, D., and Lee, D., Precision Manufacturing, 2008, Springer, New York, p. 649. 3. Min, S., Studies on Modeling and Optimization of Drilling Burr Formation. Ph.D. Dissertation, Dept. of Mech. Eng., University of California at Berkeley, 2001. 4. Newton, T., Morehouse, J., Melkote, S., and Turner, S., “An Experimental Study of Interfacial Burr Formation in Drilling of Stacked Aluminum Sheets,” Trans. NAMRI/SME, Vol. 36, 2008. 5. Choi, J., Min, S., and Dornfeld, D., “Finite element modeling of burr formation in drilling of a multi-layered material,” Proc. 7th
17. 18.
19.
20.
21.
Int’l Conference on Burr Formation and Surface Finishing, University of California, Berkeley, June, 2004. Aurich, J., Sudermann, H., and Braun, O., “Burr Formation in Grinding – Investigation of the Influence of Different Grinding Wheels,” CIRP Paris Meeting, January, 2004. Narayanaswami, R., and Dornfeld, D.A., “Design and Process Planning Strategies for Burr Minimization and Deburring,” Trans. North American Manufacturing Research Institute, SME, Vol. 22, 1994, pp. 313–322. Narayanaswami, R., and Dornfeld, D.A., “Burr Minimization in Face Milling: A Geometrical Approach,” Trans. ASME, J. Engineering for Industry, Vol. 119, No. 2, 1997, pp. 170–177. Dornfeld, D., Wright, P., Wang, F., Sheng, P. Stori, J., Sundarajaran, V., Krishnan, N., and Chu, C., “Multi-agent Process Planning for a Networked Machining Service,” Trans. North American Manufacturing Research Institute, 1999, pp. 191–196. Chu, C.H., and Dornfeld, D., “Linking Tool Paths Generated with Different Offset Distances for Edge Quality Enhancement in Planar Milling,” Proc. Inst. Mech. Engs., Part B. J. Engineering Manufacturing, Vol. 218, 2004, pp. 721–730. Avila, M., and Dornfeld, D.A., “Exit Order Sequence Burr Formation Prediction Algorithm Based on Rectangular Coordinates,” Trans. North American Manufacturing Research Institute, Vol. 34, 2006, pp. 301–307. Min, S., and Dornfeld, D., “Application of Four Levels of Drilling Burr Minimization Strategies,” Proc. 2nd Asia-Pacific Forum on Precision Surface Finishing and Deburring Technology, Seoul Korea, July, 2002, pp. 95–104. Weinert, K., Kahnis, P., Petzoldt, V., and Peters, C., “Micromilling of Steel and NiTi SMA,” 55th CIRP General Assembly, STC-C Technical Meeting, Antalya, Turkey, 2005. Lee, K., and Dornfeld, D., “An Experimental Study on Burr Formation in Micro-milling Aluminum and Copper,” Trans. NAMRI/SME, Vol. 30, 2002, pp. 1–8. Min., S., Lee, D., De Grave, A., De Oliveira, C., Lin, J., and Dornfeld, D., “Surface and Edge Quality Variation in Precision of Single Crystal and Polycrystalline Materials,” Proc. 7th Int’l Conference on Burr Formation and Surface Finishing, University of California, Berkeley, June, 2004. Bissacco, G., Hansen, H., and De Chiffre, L., “Micromilling of Hardened Tool Steel for Mould Making Applications,” J. Materials Processing Technology, Vol. 167, No. 2–3, 2005, pp. 201–201. Ahn, J., and Lim, H., “Side Burr Generation Model in MicroGrooving,” Proc. ASPE, Vol. 16, 1997, pp. 215–219. Ahn, J., Lim, H., and Son, S., “Burr and Shape Distortion in Microgrooving of Non-Ferrous Metals Using a Diamond Tool,” KSME International Journal, Vol. 14, No. 11, 2000, pp. 1244– 1249. Schaller, T., Bohn, L., Meyer, J., and Schubert, K., “Microstructure Grooves with a Width of Less than 50 Microns Cut with Ground Hard Metal Micro End Mills,” Precision Engineering, Vol. 23, No. 4, 1999, pp. 229–235. Sugawara, A., and Inagaki, K., “Effect of Workpiece Structure on Burr Formation in Micro-Drilling,” Precision Engineering, Vol. 4, No. 1, 1982, pp. 9–14. Lee, K., and Dornfeld, D., “Micro-Burr Formation and Minimization Through Process Control,” Precision Engineering, Vol. 29, 2005, pp. 246–252.
Burr Minimization Strategies in Machining Operations D. Biermann and M. Heilmann
Abstract Reducing burr formation in machining operations is of vital importance as they can decrease the functionality of components and can cause injuries. Nowadays, additional processes for deburring are often necessary. To avoid deburring, the modification of machining processes is a promising approach. Here, different parameters have a significant influence on burr formation. For example, the use of alternative machining processes or the reduction of the workpiece temperature near the edge of the workpiece shows high potential for burr reduction. This temperature reduction causes a change in material properties which decreases burr formation. In this paper, methods for burr minimization in various cutting processes are presented. Burr reduction strategies for turning, drilling and milling of different materials are presented. Keywords Machining · Burr minimization strategies
1 Introduction Components used in technical applications must fulfill specific criteria to ensure the functionality of the product. Here, in particular, the adherence to the tolerances for form, dimension and position of the component are of great concern. Another important factor is the shape of the workpiece edge. Material overhangs, burrs, should be avoided. These deviations in the desired shape of workpieces can reduce the functionality of the component and can cause injuries occuring while handling of the workpieces. These evaluation factors are influenced by the kind of production processes used. Adjusting these processes, taking into account the demands placed on the components, can lead to higher workpiece
D. Biermann (), M. Heilmann Institute of Machining Technology, Technische Universität Dortmund, Baroper Str. 301, 44227 Dortmund, Germany e-mail:
[email protected] url: www.isf.de
quality. Here, specific processes and process strategies can cause favorable technological and economical results. Within the scope of this study different burr reduction strategies in machining operations are presented.
2 Burr Formation and Reduction Strategy in Turning Austenitic-Ferritic Stainless Steel Stainless steels, especially with an austenitic or austeniticferritic microstructure, are distinguished by their suitable applicative nature due to their good combination of high chemical resistance and good mechanical properties. These properties are influenced by the kind, quantity and nature of their alloying elements. Moreover, they are also dependent on the heat treatment used, the previous history of the material and the resulting microstructure. However, the suitable usage properties of stainless steels, especially with austenitic or austenitic-ferritic microstructure, pose challenges in machining which are expressed in high adhesion affinity up to high cutting speed ranges, high thermal loads as well as in a hardening of the material [1, 2]. Furthermore, the high toughness leads to an unpropitious chip breakage and increased burr formation. In turning stainless steel, burr formation is of great importance because it influences not only the quality and handling of the workpiece but also the tool wear. A direct connection between notch wear occurring at turning inserts and burr formation is discernible. Figure 1 shows the principle coherences between burr formation and notch wear separated into three steps. The individual process steps are illustrated during the turning of a stainless steel with austenitic-ferritic microstructure. The linear contact zone of workpiece material and insert at the end of the contact area causes a furrowing load. In machining materials that have a tendency towards work hardening, the effect described is reinforced because of the increasing hardness of the material. These loads can develop notch wear at the insert, although the cutting material and the
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_2, © Springer-Verlag Berlin Heidelberg 2010
13
14
D. Biermann and M. Heilmann
Fig. 1 Phases of burr formation and the development of notch wear [3, 4]
coating show a high wear resistance at that time. In this phase a plastically deformed area at the main cutting edge at the end of the contact zone between material and insert is created. A burr appears on the workpiece side but its growth is minor. In the second step, the burr causes the cobalt binder-phase to wear off. This results in fatigue of the cemented carbide. Therefore, there is an interaction between burr formation and the formation of notch wear as well as an increase of both. The consequence of this effect is a progressive burr growth. The high values of notch wear and the loads induced by the burrs lead to a significant damage of the insert. The mechanisms described cause a plastically deformed area at the end of the contact zone of the major cutting edge. The initiation and the development of burr formation depend significantly on the properties of the work material, the tool and the cutting parameters. Due to high fracture toughness and low heat conductivity, the burr formation is distinctive in machining stainless steels with austenitic and austenitic-ferritic microstructure [4]. The effects described can be reduced by the use of a process cooling. In this research an innovative cooling concept – the use of a carbon-dioxide cooling – was applied to overcome the challenges occurring in turning stainless steel like tool wear, chip and burr formation. In Fig. 2 the potential of the carbon dioxide cooling in minimizing burr formation and tool wear is presented in comparison to conventional cooling lubrication concepts. The results show a distinct effect of the cooling lubrication concept on burr formation and notch wear, with the
other aspects kept constant. When using dry machining and minimum quantity lubrication (MQL) concept in machining tests, a medium-sized wavy burr with a stable burr root was created. Here, a wide notch wear having a moderate depth occurs. This kind of notch wear is created by high process temperatures in dry machining and causes a larger plastically deformed area at the insert. Furthermore, dispense with any lubricant leads to adhesion and the appearance of material deposits at the notch of the insert. The reduced temperature in experiments run with emulsion used as the coolant leads to a smaller notch wear area. This wear form benefits burr formation. At first, a homogeneous burr is created, which is cut into parts when a specific value for burr thickness is reached. Compared to burr formation when using conventional cooling concepts, the application of carbon dioxide as cooling medium leads to negligible burr formation. As in experiments with emulsion as coolant, a small notch is created. But the intensive cooling leads to an effect on the material properties in the area at the end of the contact zone. The reduction of flow stress is minimized by the cooling and this leads to a lower formability. In addition, the kinetic energy of the snow blast can produce a deformation and breakage of the burr. Other benefits when using carbon dioxide as cooling medium in turning stainless steel are a reduction of tool wear, especially adhesion and notch wear, as well as an improved chip breakage. In conclusion, the use of carbon dioxide as a cooling medium is an appropriate method for improving the machinability of stainless steel.
Burr Minimization Strategies in Machining Operations
15
Fig. 2 Influence of the cooling concept on burr formation and tool wear in turning stainless steel
3 CO2 -Process Cooling for Burr Reduction in Face Milling of Aluminum Alloys For the fulfillment of their functionality components with functional surfaces like seal faces make special demands on the properties of these surfaces. In particular, in this con-
text, the surface quality and the workpiece edge shape are of main importance. To satisfy these workpiece requirements, it is necessary to modify the process. With respect to burr reduction at the component edge, different burr reduction strategies and additional separate processes for deburring were developed. A new, promising, strategy is to cool the edge of the component with dry ice snow while machining.
Fig. 3 Influence of the cooling strategy and the feed rate on the burr formation [5]
16
Because of the intensive cooling of the workpiece edge with carbon dioxide snow having a minimum temperature of Tmin = –79◦ C, the decrease in the flow stress caused by the process is minimized and the corresponding increase in the formability of the material is avoided. In comparison to the conventional process under dry conditions the material becomes more brittle and this results in a reduced burr formation. In Fig. 3 burr formation when face milling an aluminum alloy with dry ice is compared to the burr formation when dry machining. Here, different cooling strategies are compared. In this context a pre-cooling of the workpiece with CO2 and an in-process cooling using two nozzles with different cooling power where the CO2 -nozzles were connected to the feed motion of the machine tool were carried out. Compared with the burr formation appearing in experiments under dry conditions, a pre-cooling of the part produces a slight reduction in burr formation. A definite reduction in burr formation is achieved by in-process cooling, especially when using the nozzle with the lower cooling power. This can be explained by the higher exit velocity of the CO2 -snow that results in a higher mechanical load on the edge of the workpiece. A quantitative measurement of the burr volume confirmed the results presented. In comparison to dry machining, the burr volume when face milling with carbon dioxide cooling can be reduced by about 30–40% [5]. Varying the feedrate in experiments conducted using the most beneficial cooling strategy, significantly influences the burr volume. With increasing feed rate the burr volume can be reduced. Using high feed rates the compressive stresses on the material near the edge of the workpiece is higher than when using lower feed rates. The deformation of material begins at a major wall thickness and the heating up of the edge of the workpiece by the process heat is lower than when using lower feed rates. In combination with the cooling of the material it reacts more brittle, a reduction in burr formation can be achieved.
D. Biermann and M. Heilmann
frequently accompanied by problems such as tool wear or poor machining results, it is necessary to examine various processes. Due to the rather expensive production of these composite materials the bore holes must be manufactured with high quality, high reproducibility and process reliability. Furthermore, machining of drill holes must be performed without damaging the peripheral zone. Defects which often occur when drilling composite materials are, for example delamination, fraying, fiber cracks, fiber pullouts or fiber bundle pull-outs. Mechanical overloading is mainly responsible for these defects, but thermal loads also play a role [6]. The defects occurring in machining fiber reinforced polymer plastics are presented in Fig. 4. After drilling, mainly fiber pull-outs, fiber cracks as well as fiber bundle disruptions can be detected. It is possible
4 Delamination Reduction Strategies in Machining Fiber-Reinforced Plastics Modern polymer composites reinforced with carbon or glass fibers are used for lightweight structures because of their low density and high strength as well. These materials are often applied in the aviation, aerospace, defense and automotive industries. Although these materials are manufactured close to the final shape, it is often necessary to machine drill holes. Direct machining of drill holes has some advantages compared to the application of metal inserts during the fabrication process, for example, production costs and lower weight. Since the cutting of fiber-reinforced composites is
Fig. 4 Defects in drilling fiber-reinforced polymer plastics [6]
Burr Minimization Strategies in Machining Operations
17
Fig. 5 Axial cutting forces and their effect on delamination [6]
to reduce the defects shown in Fig. 4 by adapting the drill tool geometry to the machining task and by substituting the drilling process with an alternative method. To avoid high axial forces that favors the development of the defects shown, orbital milling offers possibilities. Further advantages are the higher flexibility, suitable chip breakage and removal. The comparison of the two competing processes for hole manufacturing is given in Fig. 5. In the research performed, a clear relationship between the axial forces and the defects at the tool exit of the workpiece is visible. The high potential of orbital milling is based on its lower axial forces occuring when the holes are machined. The formation of delaminations originating from a pulling apart of the fibers when the tool exits the drill hole is significantly reduced. This kind of damage leads to a loss in strength of the holes produced. The different characteristics of the defects which emerge in the peripheral zone in both processes can clearly be identified. The damages occuring when using twist drills, appear, to a minor degree after orbital milling. As a result, it is obvious that orbital milling represents a good alternative to the conventional drilling operations [6].
5 Burr Formation and Reduction in Machining Thin-Walled Light-Metal Frame Structures In multiple applications light metal profiles are used. Main application areas of light metal profiles are construction, automotive and locomotive industries, for example, for body parts in locomotives. Machining of these profiles is necessary, for example, for fixation elements. Here, different challenges can occur. Main aspects are vibrations and deformations during the machining process. Furthermore, depending on the process, massive burr formation can occur. To avoid these problems it is necessary to choose an appropriate process. As when machining fiber-reinforced polymer plastics, orbital milling offers good possibilities when cutting lightmetal profiles. Figure 6 illustrates the potential of orbital milling concerning burr reduction in comparison to a conventional drilling process. In this comparison a drill was used that has shown good process behavior and in relationship to other drilling tools low burr formation. In drilling it is obvious that burr
18
D. Biermann and M. Heilmann
Fig. 6 Comparison of drilling and orbital milling when machining light weight metal frame structures [7]
formation is unavoidable, even under favorable process conditions. In contrast to the results in drilling, orbital milling produces bore holes with minor burr formation. Only at the tool entry point into the material a slight burr occurred. The lower burr formation in orbital milling, compared to that when drilling, is a result of the lower axial force. With higher axial forces a stronger deformation of the material at the tool exit side occurs, which ends in a shearing out of the bore cap. This shearing out effects a larger burr formation. Both processes are characterized by specific properties. Using drilling, there is a short operating time and low machine tool requirements. On the other hand drilling provides a low flexibility with regard to bore hole diameter, and the machining results are poor. In comparison, the orbital milling operations allow flexible production of different drill-hole diameters and further functional elements using just one tool. This generally allows the reduction of the non-productive time of a machining process and represents an improvement of the productivity. Furthermore, remarkably improved machining qualities can be achieved in comparative tests. The discontinuous cut results in an advantageous chip formation and removal, as well as in a significantly lower burr formation [7, 8].
6 Burr Formation in Micro-milling of NiTi Alloys Because of their properties materials like NiTi Shape Memory Alloys (SMA) offer high potential in special applications. A main field of application is in medical technology, because the material has versatile properties, like biocompatibility, superelasticity and shape memory behavior [9, 10]. These favorable usage properties are offset by properties that complicate the machinability of the material. These properties are the tendency towards adhesion and work hardening. The main properties of the material used are presented in Table 1. Due to the possibility of machining parts with a complex shape, micro-milling is of great importance for the production of micro parts, for example in medical applications. Here, an individual adaption of the cutting process to the machining task is necessary due to the fact that a linear scaling of the milling process is not feasible. Therefore, groove milling was carried out with end mills with a diameter of d = 0.4 mm. In Fig. 7, the influence of the cutting parameters, depth of cut ap and width of cut ae , is presented.
Burr Minimization Strategies in Machining Operations Table 1 Material properties of β-NiTi alloy [11]
Fig. 7 Influence of the depth of cut and width of cut on burr formation in micro-milling of NiTi [11]
19
20
Experiments conducted using values for depth of cut and width of cut, ap = ae = 10 μm, that have produced advantageous process behavior in dry machining reveal a significant burr along the groove edge when using MQL. The surface of the groove is very homogeneous, without any material deposits. Doubling the values for both parameters leads to a reduction in burr formation. Disadvantageous are the feed marks at the surface of the groove. The higher force occurring in this experiment founded in the higher parameter values results in a higher radial push away of the tool and that causes marks at the surface of the groove at every lateral position. A further increase in depth and width of cut leads to an intensification of this effect. By microscopy, the longitudinal marks can clearly be seen. In addition, a further increase in burr formation occurs. Despite the increase in the material removal rate, the tool wear does not reach a critical level [11, 12].
D. Biermann and M. Heilmann
workpiece temperature on burr formation and the workpiece quality in face milling of aluminum alloys will be worked out. The measurement of thermal and mechanical loads will allow conclusions concerning the material properties while the machining process. Using this data, burr reduction strategies founded on in-process workpiece cooling will be developed. Furthermore, the influence of the process parameters which are relevant for burr reduction will also be analyzed. Concerning micro-milling, the main parameters on burr formation in five-axis milling of NiTi-alloys will be figured out. In addition to the main parameters influencing burr formation in conventional three-axis micro-milling processes the effects of the tool inclination angle will be investigated. Acknowledgements The authors acknowledge funding by the German Research Foundation (DFG) for several research projects.
References 7 Conclusion and Outlook This paper has shown that choosing the appropriate process and strategy has a high potential for reducing burr formation. It is possible to transfer similar approaches for burr reduction to machining tasks that display parallel properties. For example, the decrease in the flow stress by process temperatures should be minimized using an efficient cooling in different machining operations and workpiece materials. As presented, the burr formation can be reduced when turning stainless steel and in face milling of aluminum alloys using this approach. With regard to bore hole production the axial force is mainly responsible for burr formation. To reduce burr formation a decrease in axial forces when using alternative machining processes leads to beneficial results for materials with varying properties. In this context, orbital milling allows a reduction in burr formation when machining aluminum frame structures as well as in reduced delamination when machining fiber-reinforced plastics. In micromachining, presented on the example of micro-milling of NiTi-alloys, the cutting parameters, especially the depth of cut and width of cut are main influencing variables on burr formation. Here, an adaption of the cutting parameters taking account of the workpiece quality can lead to favorable machining results. However, the practicability of alternative processes or strategies depends on the machining task and the boundary conditions of the process. The reduction of burr formation in machining processes is of high relevance, because deburring causes additional processes and is associated with additional costs. For burr reduction, the knowledge of the relevant processes of burr formation must be known. Therefore, the influence of the
1. Agrawal, S., Chakrabarti, A., Chattopadhyay, K. A. B., 1995, A Study on the Machining of Cast Austenitic Stainless-Steels with Carbide Tools. Journal of Materials Processing Technology, 52: 610–620 2. Schoß, V., 2001, Martensitische Umwandlung und Ermüdung austenitischer Edelstähle – Gefügeveränderungen und Möglichkeiten der Früherkennung von Ermüdungsschäden, Dissertation Technische Universität Bergakademie Freiberg 3. Buschka, M., 2002, Formgedächtnistechnik – Prozessgestaltung beim Drehen und Bohren von NiTi-Formgedächtnislegierungen, Dissertation Universität Dortmund 4. Hesterberg, S., 2006, Trockenbearbeitung nichtrostender Stähle. Prozessauslegung für das Drehen und Bohren mit Wendeschneidplatten, Dissertation Universität Dortmund 5. Weinert, K., Kersting, M., 2007, Schneesturm vermindert Gratbildung. Technica, 3: 30–32 6. Weinert, K., Kempmann, C., 2005, Comparing Drilling and Circular Milling for the Drill Hole Manufacture of Fiber Reinforced Composites, Production Engineering – Production Engineering, Annals of the German Academic Society for Production Engineering, Vol. XII(2): 1–4 7. Weinert, K., Hammer, N., 2004, Zirkularfräsen von Bohrungen im Leichtbau. WB Werkstatt und Betrieb, Industrielle Metallbearbeitung, 10: 54–58 8. Hammer, M., 2007, Spanende Bearbeitung endlos stahlverstärkter Aluminiummatrixstrangpressprofile, Dissertation Universität Dortmund 9. Kahn, H., Huff, M. A., Heuer, A. H., 1998, The TiNi-shapememory alloy and itds applications for MEMS, Journal of Micromechanics and Microengineering, 8: 212–221 10. Machado, L. G., Savi, M. A., 2003, Medical applications of shape memory alloys, Brazilian Journal of Medical and Biological Research, 86: 683–691 11. Petzold, V. 2006, Formgedächtnistechnik – Tiefbohren und Mikrofräsen von NiTi, Dissertation Universität Dortmund 12. Weinert, K., Petzold, V., 2006, Micromachining of NiTi Shape Memory Alloys, Production Engineering – Production Engineering, Annals of the German Academic Society for Production Engineering, Vol. XIII, 2: 43–46
Burr Formation and Avoidance for Robust Circular Blade Sawing of Thin Walled Extruded Aluminum Profiles K. Martinsen and G. Ringen
Abstract Thin walled extruded aluminum profiles are usually cut by sawing processes to appropriate length and angles. Burrs can be a problem for the robustness of this process, as well as for handling and function of the parts. Deburring processes after sawing (manually or automatic) will be an additional cost. This paper shows analysis of bur formation after circular blade sawing of thin walled extruded aluminum profiles and the effects of changing process parameters such as cutting speed and saw tooth geometry. The paper shows how substantial improvement could be achieved through small investments on new sawing blades and saw blade motors with higher rotation speed. Keywords Burr analysis and control · Sawing · Aluminum profiles
machining burrs classification: Cutting edge and direction of formation. The burrs in this paper are mainly exit burrs form the minor cutting edge of the saw tooth. As Aurich et al. [2] writes, are burr formation impossible to avoid in cutting operations. The size and shape of the burrs, however, are dependent on several parameters on the work piece, cutting tool, kinematics and machine tool. Despite the problems caused by burrs and that burrs are impossible to avoid completely, not many studies have been made on burr and burr formations. Hashimura et al. [3] being one of the few who have med Finite Element analysis on burr formation. Heisel et al. [4] showing a model for predicting burr dimension in short-hole drilling. Toropov and Ko [5] describes a analytic model of burr formation in orthogonal cutting.
1 Introduction Extruded aluminum profiles are popular semi-finished products usable for further assembly and processing into many different products in many industrial sections. Typical examples are bumper beams or space frames in automotive industry and window or door frames in constructions industry. The extrusion process usually extrudes the profiles in very long lengths and these must be cut into shorter lengths according to the application of the profile and circular sawing is one of the most common processes. This paper focuses on burr formation in circular blade sawing. Burr and chip formation are one of the major obstacles in achieving a robust sawing process. Nakayama and Arai [1] calls burrs together with chips the most bothersome foes to productivity and atomization in machine part production. They introduce two systems for
1.1 Sawing Process Parameters The circular sawing process for cutting of thin walled aluminum profiles is described in references [5–9]. According to these as well the process parameters should typically be within the areas as described in Table 1.
1.1.1 Cutting Speed The literature recommends high cutting speed for aluminum, from 3000 and up to 4500 m/min and in certain cases up to 5500 m/min. Too high cutting speed may lead to overheating
Table 1 Typical process parameters for circular blade sawing of thinwalled (< 8 mm) aluminum profiles K. Martinsen (), G. Ringen RTIM AS, P.O. Box 2831, Raufoss, Norway e-mail:
[email protected] url: www.rtim.com
Cutting speed Rake angle [y] Clearance angle [α] Feed rate per. tooth
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_3, © Springer-Verlag Berlin Heidelberg 2010
[m/min] [degree] [degree] [mm/tooth]
2500–4500 5–10 7–13 0.01–0.06
21
22
and thereby affinity of material to the saw teeth. Too low cutting speed in the other end often requires increased feed rate in order to keep up the productivity with increased burr formations, rougher surfaces and deformation of the work piece as side effects. Large saw blade diameters are affordable according to achieve high cutting speed but could also lead to instability, unevenness and distortion. Smaller diameters may in turn lead to rougher surfaces and increased burr formation. Diameter decisions must also include aluminum profile dimension, sawing efficiency, motor effect and available space in the sawing frame.
1.1.2 Rake Angle Rake angle [y], shown in Fig. 1, is recommended positive for softer materials as aluminum. A positive rake angle gives reduced cutting force and thereby less deformation and burr formation. On the negative such geometry weakens the saw tooth, and for hard metal (HM) saw teeth it is a compromise between tool life and reduced cutting force. The literature recommends up to ten degree rake angle for aluminum and in general the softer the material the higher the rake angle.
1.1.3 Clearance Angle
K. Martinsen and G. Ringen
1.1.4 Feed Rate per Tooth The literature recommends a broad range at feed rate per tooth where 0.01–0.02 mm/tooth are suggested for cases where reduction of burr formation is important, but up to 0.06 mm/tooth are recommended for certain cases. It is also recommended to keep the feed rate per. tooth constant, that means when vf changes the rotational speed has to adjust as well given in [2]. Too high feed rate per. tooth led to increased burr formation and deformation of the work piece. Number of saw teeth is a trade off between degree of surface fineness and avoidance of affinition of material to the tool.
1.1.5 Tooth Geometry Figure 2 shows three types of saw tooth geometry (A, B and C), where type A is recommended for sawing of massive aluminum billets and thick-walled profiles. These profiles can be found in [10]. Type B and C are recommended for thinwalled profiles, where the limit between thick and thin is assumed to be around four millimeters or little above. Type B is assumed to be best suited in order to reduce burr formation. The inclination of the teeth will unfortunately wear out the saw blade faster than more traditional once and will after a while contribute to an increased burr formation. For cutting of thin-walled aluminum profiles it is therefore a trade off between type B and C when it comes to life time economy and reduced burr formation.
A high clearance angle [α] is important to achieve good chip removal and fine surfaces since less material will adhere to the saw tooth. But as for the rake angle a too large angle will reduce the saw tooth strength and the saw blade life time. The literature recommends a clearance angle between 7 and 13 degree, but it has to be seen in combination with the rake angle.
Fig. 1 Basic tooth angles
Fig. 2 Saw tooth geometry (cutting edge view)
Burr Formation and Avoidance for Robust Circular Blade Sawing of Thin Walled Extruded Aluminum Profiles
1.1.6 Machine Requirements Sawing of aluminum requires machines with the ability to perform high cutting speed with good effect and dynamic stability. Especially for aluminum alloys in the 6xxx and 7xxx categories, in soft condition, this requirement must be met. High rotational speed requires high motor effect, and on a general basis it is recommended five times the effect of comparable saws designed for steel cutting. Distortion is another important feature to control when improving sawing effectiveness. Two conditions which could affect burr formation negatively are either if the cut-off table is not aligned with the saw blade or if the feed direction of the saw blade is not aligned with the cut-off table. Vibrations are also considered to worsen burr formation, and typical solutions in order to reduce or avoid such are often to adjust rotational speed away from the machine’s own frequencies. Robot sawing is often more exposed to vibrations due to less stiffness than dedicated sawing machines.
23
aluminum profiles before cutting is in the category F, which is a soft condition. In the second sawing process aluminum profiles are cut both before and after the forming process by robot saws. Also here the temper conditions are relatively soft since the precutting is done in the same condition as after the extrusion process and the post-cutting is done after solution treatment, quenching and forming but before age hardening. In these cutting processes most of the cuts are done in an inclined angle so the cross section of the cut increase compare to a perpendicular once. The fine sawing process is used to a diversity of aluminum products where smooth work piece surface texture is important.
2 Experimental Setup The test equipments used to perform the tests in this analysis are three different circular blade saws in their original environment at the case company.
1.1.7 Sawing of Aluminum Profiles The case company has many sawing processes but they can mainly be categorized into: • Sawing of aluminum profiles after extrusion • Sawing of aluminum profiles before assembly • Fine sawing In the sawing of aluminum profiles after extrusion, the saw cuts 40 m long profiles after the extrusion process into more suitable lengths for further downstream processes. The sawing process is done by dedicated sawing machines where several profiles are fixed side by side and cut off simultaneously by a rotating circular saw blade. The temper condition of the
2.1 Experiments It was run three experiments, which are denoted experiments from one to three, according to type of saw as seen in the header row in Tables 2, 3 and 4. In general, all the three experiments are run with the original setup with a new or re-sharpened blade, the new blade and unchanged feed rate and the new blade with a higher feed rate. The three tables summarizes the critical parameters for both the saw blade and the sawing process, and of importance to notice is the tooth width difference between the original and the new saw blade (Fig. 3).
Table 2 Experiment 1 Experiment 1, Peris 450 manual saw Parameters Rake angle Clearance angle Diameter Tooth width Blade width Number of teeth Rotational speed Feed rate Center boss diameter Weight Cutting speed Feed rate per tooth
y α D B b z N vf Db g Vc sz
angle angle mm mm mm number 1/min mm/min mm kg m/min mm/tooth
Original saw blade
New blade, unchanged feed rate
New blade, high feed rate
5 13 450 4.4 3.2 108 2875 1360 97.4 18 4062 0.00438
5 13 450 2.8 2.4 108 2875 1360 100 18 4062 0.00439
5 13 450 2.8 2.4 108 2875 3400 100 18 4062 0.01095
24
K. Martinsen and G. Ringen Table 3 Experiment 2 Experiment 2, Emmegi Model 550 Auto Parameters Rake angle Clearance angle Diameter Tooth width Blade width Number of teeth Rotational speed Feed rate Center boss diameter Weight Cutting speed Feed rate per tooth
y α D B b z N vf Db g Vc sz
angle angle mm mm mm number 1/min mm/min mm kg m/min mm/tooth
Original saw blade
New blade, unchanged feed rate
New blade, high feed rate
5 13 550 4.4 3.4 108 2870 652.8 100 18 4956 0.00211
5 13 550 3.5 3.0 108 2870 652.8 100 18 4956 0.00211
5 13 550 3.5 3.0 108 2870 858.9 100 18 4956 0.00277
Original saw blade
New blade, unchanged feed rate
New blade, high feed rate
5 13 550 4.4 3.4 108 1450 2975.4 100 18 2504 0.01900
5 13 550 3.7 2.7 120 1450 3306.0 100 18 2504 0.01900
5 13 550 3.7 2.7 120 1450 3967.2 100 18 2504 0.02280
Table 4 Experiment 3 Experiment 3, Robot saw Parameters Rake angle Clearance angle Diameter Tooth width Blade width Number of teeth Rotational speed Feed rate Center boss diameter Weight Cutting speed Feed rate per tooth
y α D B b z N vf Db g Vc sz
angle angle mm mm mm number 1/min mm/min mm kg m/min mm/tooth
Fig. 3 Tooth width difference between old blade (1) and new one (2), (cutting edge view)
In Fig. 4, Fig. 5 and Fig. 6 below are pictures of the physical parts according to the three experiments shown (the cutting length of the parts along the black arrows is 100 mm). The black arrows on the pictures show which edges of the parts that are analyzed regarding to burr formation. These locations are chosen in order to give the best representative picture of the burr for these parts.
Fig. 4 Part experiment 1
Burr Formation and Avoidance for Robust Circular Blade Sawing of Thin Walled Extruded Aluminum Profiles
3.1 Results from Experiment 1
Fig. 5 Part experiment 2
Fig. 7 Experiment 1, original saw blade (N = 2875 and vf = 1360)
Fig. 6 Part experiment 3
Fig. 8 Experiment 1, new blade (N = 2875 and vf = 1360)
3 Results The results from the three experiments and the three states within each experiment, original saw blade, new saw blade with unchanged feed rate and new saw blade with high feed rate, are shown in Figs. 7–15.
Fig. 9 Experiment 1, new blade (N = 2875 and vf = 3400)
25
26
K. Martinsen and G. Ringen
3.2 Results from Experiment 2
3.3 Results from Experiment 3
Fig. 10 Experiment 2, original saw blade (N = 2870 and vf = 652.8)
Fig. 13 Experiment 3, original saw blade (N = 1450 and vf = 2975.4)
Fig. 11 Experiment 2, new blade (N = 2870 and vf = 652.8)
Fig. 14 Experiment 3, new blade (N = 1450 and vf = 3306)
Fig. 12 Experiment 2, new blade (N = 2870 and vf = 858.9)
Fig. 15 Experiment 3, new blade (N = 1450 and vf = 3967.2)
Burr Formation and Avoidance for Robust Circular Blade Sawing of Thin Walled Extruded Aluminum Profiles
27
4 Discussion of Results
4.4 Possible Sources of Errors
4.1 Tooth Profile
The burrs in this analysis was measured by cutting out a sample from the profile, mould this into a plastic compound and then grind and polish this before looking at the polished surface in a microscope. The microscope picture shows a intersection of the burr and a part of the work piece. This is the same method as used for metallographic analysis. Since the burrs are not uniformly shaped along the profile, the microscope pictures will one be samples of the burr. Emphasis was put on getting the most typical burr intersection, but this is still only a sample and not the most extreme nor the average of the burrs on the profile. The intersection should be nearly perpendicular to the burr, but the process of cutting, molding and grinding would means this will have some offset.
The experiments shows that saw blades with a thinner saw blade and tooth width (kerf) profile results in considerably less burr formation given an appropriate feed rate. The tooth width on the new blades is respectively reduced by 36.4%, 20.5% and 15.9% compared to the original blades in the three experiments. A thinner saw blade gives reduced cutting force, which is assumed to impact the burr formation positively in addition to improve the cutting economy for the sawing process. By reducing tooth width the blade width itself has to be reduced in order to remove chips properly, and this could affect the stability of the cutting process. Less area to transfer force from the saw motor to the cutting process may lead to vibrations and a more unstable process. This was not a problem during these experiments.
4.5 Further Work 4.2 Feed Rate Another feature of the experiments is that the new blades with original feed rate result in less burr formation than the tests with the new blades and high feed rate. It should be noticed that the feed rates in use during these experiments are low according to the recommendations shown in Table 1. In general the results are well in accordance with theory that claims that in order to reduce burr formation high cutting speed and low feed rate are preferred process parameters. As seen from experiment number three, the burr formation with the new blade and original feed rate gives less satisfying results compared to the previous two experiments. This could be explained by the relatively low cutting speed and higher feed rate than the others in addition to the inclined exit cut which is likely to increase burr formation compared to a straight cut.
4.3 Other Process Parameters Other process parameters which could impact the process efficiency and burr formation are rake angle, clearance angle, saw tooth geometry, blade diameter, number of saw teeth, tool material, aluminum profile clamping, stability and straightness of saw and saw table etc. All these parameters were the same during the experiments except from the number of saw teeth between the original and new saw blade in experiment three. An interesting additional finding during the experiments was that the peak noise level for the new saw blades was observed to drop between 4 and 7 dB compared to the original once.
For further research it has to be performed analysis of the long term performance of these new blades according to wear and at which point they should be changed or re-sharpened in order to keep the burr formation at an acceptable level.
5 Conclusion This paper shows the results from a series of experiments in three different circular blade saws used for cutting thinwalled aluminum profiles. The main objectives was to investigate how process parameters and saw blade geometry affects the burr formation and thereby the process robustness. The results show that thinner kerfs give less deformation impact and thus less burrs. Secondly, an appropriate combination of cutting speed and feed rate is important. High cutting speed and low feed rate gives satisfying results in the performed experiments. Acknowledgements We would like to thank the Norwegian Research Council for funding this research in the NORMAN Centre for Research Based Innovation.
References 1. Nakayama, K., Arai, M., 1987, Burr Formation in Metal Cutting, Annals of the CIRP, 36(1): 33–36. 2. Aurich, J.C., Sudermann, H., Bil, H., 2005, Characterisation of Burr Formation in Grinding and Prospects for Modelling, Annals of the CIRP, 54(1):313–316.
28 3. Hashimura, M., Ueda, K., Dornfeld, D.A., Manabe, K., 1995, Analysis of Three Dimensional Burr Formation in Oblique Cutting, Annals of the CIRP, 44(1):27–30. 4. Heisel, U., Luik, M., Eisseler, R., Schaal, M., 2005, Prediction of Parameters for the Burr Dimensions in Short-Hole Drilling, Annals of the CIRP, 4(1):79–82. 5. Toropov, A.A., Ko, S.L., 2006, A New Burr Formation Model for Orthogonal Cutting of Ductile Materials, Annals of the CIRP, 55(1):55–58. 6. Edstrøm, C.M., 1975, Circular Sawing of Aluminium – Quality Control of Cut, Proceedings of Prod. Enhancement in Cutting and Forming Operation, Jan 29–31.
K. Martinsen and G. Ringen 7. Henning K., Bleher, S., 1988, Produktivitätsorietiertes Kreissägen von Aluminium Knetlegierungen, Werkstatt und Betrieb 121. 8. Johne, P., 1984, Bandsägen und Kreissägen zum Aluminiumtrennen, Werkstatt und Betrieb 117. 9. Johne, P., 1984, Handbuch der Aluminiumzerspanung, Aluminium-Verlag, Düsseldorf. 10. DIN Norm 1837 og 1838.
Mechanics, Modeling and Simulation of Burr Formation
Burr and Cap Formation by Orbital Drilling of Aluminum E. Brinksmeier and S. Fangmann
Abstract In orbital drilling the tool (special end mill) moves relative to the workpiece on a helical course. Because of the three dimensional tool path and the superimposed rotary cutting motion a complex machining motion results which determines the contact conditions of the tool. In the aircraft industry this process is used for cutting composite materials (CFRP/Aluminum) with automated drill and rivet machines, for example in the manufacturing of flap tracks and vertical tail planes. A key problem in the industrial manufacturing of closed structures is the cap and the burr formation on the bore exit side. All caps and chips must be able to remove by vacuum, in addition a minimum burr height is required. The objective of this study is to describe the effects of orbital drilling on cap and burr formation in the exit composite material, primed clad aluminum 2024. The influence on cap and burr formation with different tool geometries, different coatings, different cutting parameters, tool wear and minimum quantity lubrication were examined. Keywords Orbital drilling · Burr formation · Cap formation
1 Introduction Cap and burr formation can generate great problems in the manufacturing of closed structures, e.g., flap tracks in the aircraft industry, because a contamination of the internal space with caps and chips is not acceptable. In this case there is the risk of corrosion and electric breakdown [1]. If it is not possible to drill with small caps and burrs, two more working steps in the assembly are necessary for the removal of the burr and the cleaning of the closed structure. This requires a
E. Brinksmeier, S. Fangmann () Foundation Institute for Materials Science, Badgasteiner Str. 3, 28359 Bremen, Germany e-mail:
[email protected] url: www.iwt-bremen.de/ft
large part of the manufacturing costs. Cap and burr formation result from the deformation of the workpiece material near the bore exit, caused by the forces of the cutting process [2]. Former investigations show, that burr formation is a complex interaction of various factors. Primarily, burr formation depends on the three factors: tool, workpiece and process parameters, which can be subdivided further [3]. In Fig. 1 the main influences on cap and burr formation for the orbital drilling process are shown. Between the influences there are many interactions, however. If there is no ideal primer, coating or workpiece material used, the tool wear and many other resulting parameters are hard to reproduce and thus are not constant. Therefore it is impossible to reproduce a cap and a burr to one hundred per cent. For cap and burr formation only tendencies can be predicted so far [2–8]. As an alternative to conventional drilling, orbital drilling offers many advantages. The kinematics of orbital drilling allow different working steps for the production of different borehole geometries with one tool. Orbital drilling can optimize the drilling process in different forms. Cylindrical holes can be drilled independent of the tool diameter and without changing the tool. This makes it possible to drill complex tapered holes and perform finishing operations using the same cylindrical tool and settings. Another significant advantage of the orbital drilling process is the diameter variability without tool change. This fact offers the possibility for a dynamic correction of the bore diameter during the drilling process. This can be used for example, for the com-
Influences on the cap and burr formation in orbital drilling
Tool
Workpiece
Process parameter
Geometry Material Coating Wear
Material Yield stress Temperature Primer Coating
Cutting speed Depth setting per orbital path Feed velocity axial and tangential Lubrication Cooling
Fig. 1 Influencing factors on cap and burr formation [2, 3]
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_4, © Springer-Verlag Berlin Heidelberg 2010
31
32
E. Brinksmeier and S. Fangmann
Fig. 2 Burr formation on the hole exit side after conventional drilling of clad primed aluminum [1]
pensation of diameter deviations caused by different material properties in a CFRP/aluminum compound structure or arising tool wear [9]. More advantages of orbital drilling are little delamination in CFRP, good chip transportation and cutting fluid accessibility. Characteristics of orbital drilling are low feed forces and a cutting edge which is not permanently engaged [10–12].
vf,a vc vf,t ap
= = = =
feed velocity, axial cutting speed feed velocity, tangential depth setting per orbital path vf,a
vc
vf,t
2 Target of the Orbital Drilling Investigation Figure 2 shows a typical problem in the industrial manufacturing of closed structures in the aircraft industry after conventional drilling in CFRP/aluminum: The cap and burr formation on the bore exit side in aluminum. The objective of this study is to assess the effects of orbital drilling on cap and burr formation in aluminum 2024 T351 and primed clad aluminum 2024 T351 (see Sect. 5.1). To evaluate possible ways for low burr formation and a small cap formation on the bore exit side different investigations with varied parameters and tool geometries in a bore diameter range between 4.2 mm up to 6 mm have been carried out. The experiments were carried out in two steps: Orbital drilling in aluminum 2024 T351 and as worse case in primed clad aluminum 2024 T351. A further aim was to achieve good surface qualities, maximum tool life and high accuracy of the borehole.
ap
Fig. 3 Orbital drilling kinematics
A
B
C tool
drilling
3 Orbital Drilling
2 1
milling aluminium
In orbital drilling (helical drill-milling) the borehole is generated by a milling tool which executes a helical path into the workpiece (Fig. 3). The bore diameter is adjusted by the radius of the helix path in the NC-program [10, 13, 14]. In Fig. 4 the first cuts of an orbital drilling process are shown. Pictured is a 3 mm solid (cemented) carbide tool orbital drilling a 4.2 mm borehole. In position A the arrow 1 illustrates the spindle rotation and arrow 2 the orbital rotation. In picture B of Fig. 4 the two cutting zones of the orbital
4 mm
Fig. 4 First cut in the orbital drilling of a primed clad aluminum 2024 workpiece
drilling process are pictured. There is a difference between drilling and milling. The ratio between drilling and milling plays a decisive role for the reached borehole qualities [13]. This ratio can be adjusted by the ratio between bore diameter and tool diameter [14].
Burr and Cap Formation by Orbital Drilling of Aluminum
33
3.1 Periphery and Front Cutting Zone
URi = Ri . 2π DB
The orbital drilling process is characterized by a simultaneous movement of the tool on a circular path and a superimposed movement in the axial direction. The superposition of these two movements forms a helical course. After a closer inspection of the geometrical conditions between the tool and the workpiece this course produces two different engagement zones into the workpiece volume. Figure 5 shows in two views the volume which is removed by the front and the peripheral cutting edges over one bore hole depth. In the first view of the front cutting edge volume in Fig. 5 the small disc pictures an approximated front cutting zone which is added over one borehole depth. The dark volume (V2 ) shows the front cut and the light volume the peripheral cut (V1 ). In addition a specific consideration of the ratio between the front cut and the peripheral cut over the total bore volume can be made. Figure 6 shows the developed (projected onto the plane) cutting plane for the derivation of the total ratio for one orbital revolution with the arbitrary inspection radius (Ri ). The total ratio (G) between peripheral cut and front cut (milling/drilling) in orbital drilling gives a constant value at every point of the helical course after the first cut. This ratio is independent of the axial feed velocity, the spindle speed
first view front cutting edge volume
peripheral cutting edge volume
drilling and milling volume
second view front cutting edge volume
peripheral cutting edge volume
A1
ap
A2i
A1i DW
Fig. 6 Cutting plane for the front and peripheral cut in different views
or the orbital rotational speed; it is solely given by the ratio between bore diameter (DB ) and tool diameter (DW ). The total ratio (G) can be calculated as follows [14]:
G=
V1 DB 2 − DW 2 = V2 DW 2
(1)
3.2 Machine Tool Data The investigations have been carried out on a Schmid Orbital Drilling Unit (ODU). The main technical data are a maximum spindle speed of ns = 30,000 RPM, a maximum orbital rotational speed of no = 400 RPM, an eccentric adjustment of the spindle of r = 5 mm, a tool fitting HSK 63, minimum quantity lubrication (internal, external) and a main drive with P = 39 kW (Fig. 7). Furthermore a force, torque, impact sound and oscillation measurement system is mounted on this machine.
drilling and milling volume
Fig. 5 Cutting volume for the front and peripheral cut in different views [14]
A2
Fig. 7 Schmid Orbital Drilling Unit (ODU)
34
E. Brinksmeier and S. Fangmann
4 Cap Formation on the Bore Exit Side For the investigation a special experimental setup is used. The bore exit side is monitored with a digital video camera and the feed forces are measured with a three component dynamometer. During the orbital drilling process the forces and the cap formation at the moment of the tool exit on the bore exit side are recorded. A specified range of tools with different geometries and coatings is tested with different process parameters.
4.1 Cap Formation on the Bore Exit Side in Aluminum 2024 T351
Fig. 9 Cap formation on the bore exit side, closed cap
In the first step of the investigation the effects of different orbital drilling process parameters in aluminum 2024 are shown. All boreholes were drilled without lubrication, in one working step and with one tool geometry on the ODU. There are three different forms of cap formation. Figure 8 illustrates the formation of a sickle cap. In Fig. 9 the formation of a closed cap is shown. It could be delicate to remove the cap through the borehole during the orbital drilling process. The third form of cap formation is pictured in Fig. 10. A cap which is not detached from the component is the worst case for a closed structure assembly. Summarizing the three forms of cap formation in aluminum 2024 T351, the sickle cap is the best one for the manufacturing of closed structures because by this the removal of the chip is ensured as far as possible. In aluminum 2024 there are two possibilities to great a sickle cap by orbital drilling with a conventional end mill. On one hand an adaption of the process parameters (high depth of cuts ap and high axial feed
1 mm
Fig. 8 Cap formation on the bore exit side, sickle cap
1 mm
1 mm
Fig. 10 Cap formation on the bore exit side, not detached cap
rates vf,a ) may be successful. The other alternative is to adjust the ratio between the hole and the tool diameter especially to the application (D/d ≥ 1.45). However, the influences of these process parameters on the other requirements like the accuracy of the borehole, the surface qualities, etc. must be considered. Figure 11 shows in the first three pictures the visualization of the tool exit. In the first step a completely closed cap similar in size to the tool diameter is produced in the range of the front cutting edges (red pictured). The rest of the material is cut with the peripheral cutting edges. In the visualization no elastic or plastic deformation is considered. The next two pictures in Fig. 11 illustrate the real process. On the workpiece surface the exit area surface curves first elastic, then plastic due to the high axial feed force. A cap is formed, which is only adhered at the edge within a thin range. The cap takes up the feed force only at the border-edge. The rest material is not affected by the axial feed force under the cap. In the range of the front cutting edges there is no more chip formation only shaping. At the border of the cap in the range of
Burr and Cap Formation by Orbital Drilling of Aluminum
front cut
35
peripheral cut
cap e no
ns
Fig. 11 Cap formation on the bore exit side; visualization and picture of the tool exit
the transition between workpiece and cap, the cap is pressed upward in a plastic deformation procedure and separated by the peripheral cut from the workpiece. The burr remaining at the workepiece is now transformed completely by the peripheral cutting edges into the final burr.
4.2 Feed Rate Influence on the Cap Formation in Aluminum 2024 T351
vf,a = 60 mm/min and in the right picture vf,a = 180 mm/min is used. The best cap for removing through the bore is produced with a high axial feed rate vf,a and a high depth of cut ap . But the holes drilled with high axial feed rates and depths of cut are characterized by a low accuracy, less than IT 8. The main reason for the diameter deviation is the high tool deflection due to friction at the tool front in combination with the orbital movement. The holes drilled with the low feed rates achieve the best results in accuracy.
Figure 12 illustrates the high influence of the axial feed rate on the cap formation. Both holes are drilled with a spindle speed ns = 18,000 RPM and an orbital rotational speed no = 300 RPM. In the left picture of Fig. 12 an axial feed rate
4.3 Influence of Tool and Bore Diameter Ratio on the Cap Formation in Aluminum 2024 T351
Fig. 12 Axial feed rate influence, left: vf,a = 60 mm/min, right: vf,a = 180 mm/min
In Fig. 13 the influence of the ratio between the bore diameter and the tool diameter on the cap formation is pictured. In the left picture of Fig. 13 a tool with a diameter of d = 3.5 mm and in the right picture a tool with d = 4 mm for a bore diameter of D = 5.1 mm is used. The cap formation benefits with increasing the ratio between bore and tool diameter. The higher borehole/tool ratio leads to the production of a small sickle cap. This means in the total ratio (G) between peripheral cut and front cut the part of the peripheral cut increases. Otherwise, boreholes drilled with a smaller tool are characterised by a lower accuracy. One reason for this is the higher tool deflection of the 3.5 mm tool due to a lower bending strength in comparison to the 4 mm tool. In
36
E. Brinksmeier and S. Fangmann
Primer (15–25 μm)
Cladding (0.2 mm)
Drill direction
Anodic layer (2–5 μm)
Aluminum 2024 T 351 Fig. 14 Multilayer workpiece material Fig. 13 Diameter ratio borehole/tool, left: 1.45; right: 1.275
5.1 Primed Clad Aluminum aluminum 2024 ratios between the bore and the tool diameter D/d ≥ 1.45 show a very small sickle cap which could not be machined with the front cutting edges of the tool. The material rest in the moment of the tool exit is removed by the peripheral cutting edges in the orbital movement, so that the small cap can be removed easily by vacuum. In machining with ratios greater than 1.45 also the use of lubrication shows no problems regarding the removement of caps and chips.
5 Cap and Burr Formation in Primed Clad Aluminum 2024 T351 Clad aluminum sheets show the worst results regarding adherent caps and burr formation. Furthermore, the surface finish (i.e., primer, paint) influences the formation and detachment of caps. Therefore in the second step of the investigation a primed clad aluminum 2024 T351 material was selected to investigate the effects of five different tool geometries and thirty process parameter sets on cap and burr formation. Five levels of spindle speed: 18,000, 21,000, 24,000, 27,000 and 30,000 RPM and three level of orbital speed: 199, 299, 399 RPM with two axial feed, 0.0033 and 0.0066 mm using up milling were selected for each type of tool, resulting in 30 cutting conditions. Furthermore the parameter sets with an orbital speed of 199 RPM were tested in down milling and with minimum quantity lubrication.
Fig. 15 Tool geometries
The workpiece material is a multilayer surface system (Fig. 14). The different characteristics of the individual layers are explained afterwords. The basic material consists of an aluminum alloy 2024 T351, which was thermal treated and stretched. The annealing condition is naturally aged. The cladding layer is a pure aluminum layer, applied as corrosion protection on the basic material. The anodic layer consists of an aluminum oxide layer, which is applied by a chromic acid bath on the surface of the material. The primer is based on an epoxy resin. This primer offers excellent stability apart from its anticorrosion characteristics against chemicals, solvents as well as fuels.
5.2 Tool Geometries Tool 1, 2, 3 and 4 are cemented carbide double cutters without coating and with different geometries in the range of the front cutting edge. The different variances of the tool geometries are compared. Tool 5 is a cemented carbide double cutter with a diamond coating (CVD process). All tools have a diameter of 3 mm (Fig. 15). The second step of the orbital drilling experiments showed the influence of some discrete tool geometry elements on the cap and burr formation. These results are basics for a better process understanding and for further target orientated constructions of orbital drilling tools.
Burr and Cap Formation by Orbital Drilling of Aluminum
37
5.3 Forces in Orbital Drilling Primed Clad Aluminum 2024 T351
Fn no
During the investigation the forces were measured with a Kistler 3-component dynamometer (Fx , Fy and Fz ). For the discussion of the experimental results these forces are calculated into rotating forces: tangential force Ft , normal force Fn and resulting force Fres (Fig. 16). Figure 17 pictures the axial feed forces Fz versus the cutting speed by using of two axial feeds fa and three orbital speeds no for tool 1, 2, 3 and 4. All experiments were carried out three times. The highest axial feed forces for all parameters were obtained with tool 1. In orbital drilling with the higher axial feed fa = 0.0066 mm there is a decrease of the axial force with increasing cutting speeds. Comparing the two axial feeds, the highest forces are obtained with the higher axial feed. The lowest axial forces are observed with tools 3 and 4. In Fig. 18 the resulting feed forces Fres versus the cutting speed under use of two axial feeds fa and three orbital speeds no for tools 1, 2, 3 and 4 are shown. Even for the result-
no = 199 min–1 120 fa = 0.0033 mm
Fre s e
vft
Ft ns
Fy Fx Fz Fig. 16 Forces in orbital drilling
ing force tool 1 shows the highest values for all parameters. Working with the high axial feed fa = 0.0066 mm there is a decrease of the axial force with increasing cutting speeds. Tool 4 shows the lowest resulting feed forces in the investigation.
no = 299 min–1
fa = 0.0066 mm fa = 0.0033 mm
no = 399 min–1
fa = 0.0066 mm fa = 0.0033 mm
fa = 0.0066 mm
tool 2
100
axial feed force Fz [N]
tool 1 80
tool 3
60
40 tool 4
20
286
258
230
202
173
286
258
230
202
173
285
257
229
201
172
285
257
229
201
172
284
256
228
200
171
284
256
228
200
171
0
cutting speed vc [m/min] Fz tool 1
Fz tool 2
Fz tool 3
Fz tool 4
Process: Orbital drilling; Method: upmilling; Material: primed clad aluminum 2024 (6 mm);Tool-∅ d: 3 mm; Bore-∅ D: 4.2 mm; Substrate: H10F; Coating: no; Number of cutting edges: 2; Spindle speed ns: 18000– 30000 RPM; Orbital rotational speed no: 199-399 RPM; Feed rate vfa: 60–200 mm/min; Cutting speed vc: 170–286 m/min; Tangential feed ft: 0.013–0.041 mm; Depth setting per helix course ap: 0.15–1 mm; No lubrication
Fig. 17 Axial feed force Fz over the cutting speed with two axial feeds fa and three orbital speeds no
38
E. Brinksmeier and S. Fangmann
120
no = 199 min–1
no = 299 min–1
no = 399 min–1
fa = 0.0033 mm fa = 0.0066 mm
fa = 0.0033 mm fa = 0.0066 mm
fa = 0.0033 mm fa = 0.0066 mm tool 1
100
tool 3 tool 4
resulting feed force Fres [N]
tool 2
80
60
40
20
cutting speed vc [m/min] Fz tool 1
Fz tool 2
Fz tool 3
Fig. 18 Resulting feed force Fres over the cutting speed with two axial feeds fa and three orbital speeds no
Fig. 19 Cap weight over the cutting speed with two axial feeds fa and three orbital speeds no
Fz tool 4
286
258
230
202
173
286
258
202
230
173
285
257
201
229
172
285
257
229
201
172
284
256
228
200
171
284
256
228
200
171
0
Burr and Cap Formation by Orbital Drilling of Aluminum
Figure 19 illustrates the averaged cap weight of three caps over the cutting speed under use of two axial feeds fa and three orbital speeds no for tools 1, 2, 3 and 4. The least caps with tool 4 are produced with the cutting speed vc = 256 m/min, the orbital speed no = 199 RPM and the low axial feed fa = 0.0033 mm.
5.4 Geometries of the Orbital Drilling Caps and Burrs Figure 20 shows four pictures of the last sequence in cap and burr formation with tool 1. The cutting parameters in the next four figures are no = 199 RPM, fa = 0.0033 mm and a cutting speed vc = 171 m/min in up milling without minimum quantity lubrication (MQL). Tool 1 produces heavy caps (Fig. 20) and burr, due to the full radius of the front cutting edge. One disadvantage of using this geometry is the occurrence of cutting speed vc = 0 m/min on top of the tool radius resulting in
39
high axial forces. Regarding the cap and burr formation this geometry is the worst one in the investigation. Figure 21 illustrates four pictures of the last moments in cap and burr formation with tool 2. The exit area surface curves first elastic, then plastic due to the high axial feed force. The cap is pressed in axial feed direction and develops a circulating bending load. The bending moment results from the axial feed force. The material, in this case the cladding layer (pure aluminum), begins to flow if the bending stress is larger than the yield stress. The yield stress depends on the prevalent temperature. The curved material cannot absorb the axial feed force under the cap because only the primer is still attached to the material. In the range of the highest points of the front cutting edges a crack is initiated. A small cap emerges. The burr still remaining at the workpiece is transformed now completely by the pheripheral cutting edges into the final burr. Figure 22 shows four pictures of the last moments in cap and burr formation with tool 3. The process of the cap formation is the same as in Fig. 21.
1 mm
Fig. 20 Cap formation in primed clad aluminum with tool 1
1 mm
Fig. 21 Cap formation in primed clad aluminum with tool 2
1 mm
Fig. 22 Cap formation in primed clad aluminum with tool 3
40
E. Brinksmeier and S. Fangmann
1 mm
Fig. 23 Cap formation in primed clad aluminum with tool 4
Figure 23 shows four pictures of the last moments in cap and burr formation with tool 4. This tool produces the least cap and burr. Due to this reason a closer look on the cap formation is taken. Figure 24 illustrates the cap formation with tool 4 in 20 detailed screens. The cap and burr formation is explained in 5 steps: 1. The tool is still in a sufficient distance from the exit surface. The exit material is not deformed either plastic or elastic. 2. On the workpiece surface the exit area surface curves first elastic, then plastic due to the high axial feed force.
1
3. A cap is formed, which is only adhered at the edge within a thin range. The cap is only affected by the feed force at the border edge. 4. The remaining material cannot absorb the axial feed force under the cap. In the range of the front cutting edges there is no more chip formation, but shaping. At the border of the cap in the range of the transition between workpiece and cap the cap is pressed upward in a plastic deformation procedure, and separated by the peripheral cut from the workpiece. 5. The remaining burr is transformed now completely by the pheripheral cutting edges into the final burr.
2
1 mm 3
4
5
cap 1 mm
Fig. 24 Cap formation in primed clad aluminum with tool 4 (vc = 171 m/min, no = 199 RPM, fa = 0.033 mm, up milling, no MQL)
Burr and Cap Formation by Orbital Drilling of Aluminum
41
1 mm
Fig. 25 Cap formation in primed clad aluminum with tool 4 (vc = 171 m/min, no = 199 RPM, fa = 0.033 mm, down milling, no MQL)
Figure 25 pictures the cap formation with tool 4 in down milling. This process strategy generally produces lager caps as up milling. Figure 26 shows the cap formation with tool 4 in up milling under use of minimum quantity lubrication. This process strategy generally produces bigger caps as dry up milling. This might be due to a micro lubricating film which occurs between material and front cutting edge. This benefits a gliding of the cladding material over the front cut-
ting edges. The micro lubricating film prevents the necessary adhesive forces between the cap material and the front cutting edges [15]. An increased shaping process occurs. The following diagram (Fig. 27) illustrates the burr formation versus the cutting speed under use of two axial feeds fa and three orbital speeds no for tool 4. Orbital drilling with a higher axial feed and an increasing cutting speed improves the burr formation.
1 mm
Fig. 26 Cap formation in primed clad aluminum with tool 4 (vc = 171 m/min, no = 199 RPM, fa = 0.033 mm, up milling, with MQL)
42
E. Brinksmeier and S. Fangmann no = 299 min−1
no = 199 min−1
280
no = 399 min−1
fa = 0.0033 mm
fa = 0.0066 mm fa = 0.0033 mm
fa = 0.0066 mm
fa = 0.0033 mm
fa = 0.0066 mm
171
171
172
173
173
240
burr height [µm]
200
160
120
80
40
0 228
284
228
284
172
229
285
229
285
230
286
cutting speed vc [m/min]
230
286
tool 4
Process: Orbital drilling; Method: upmilling; Material: primed clad aluminum 2024 (6 mm); Tool-∅ d: 3 mm; Bore-∅ D: 4.2 mm; Substrate: H10F; Coating: no; Number of cutting edges: 2; Spindle speed ns: 18000–30000 RPM; Orbital rotational speed no: 199–399 RPM; Feed rate vfa: 60–200 mm/min; Cutting speed vc: 170–286 m/min; Tangential feed ft: 0.013–0.041 mm; Depth setting per helix course ap: 0.15–1 mm; No lubrication
Fig. 27 Burr formation in primed clad aluminum with tool 4 over the cutting speed with two axial feeds fa and three orbital speeds no no = 199 min−1 up milling
down milling
up milling with MQL
100
fa = 0.0033 mm
fa = 0.0066 mm
fa = 0.0033 mm
fa = 0.0066 mm
fa = 0.0033 mm
fa = 0.0066 mm
171
171
171
171 228
171 228
burr height [µm]
80
60
40
20
0
171
228
284
228
284
228
284
228
284
cutting speed vc [m/min]
284
284
tool 4
Process: Orbital drilling; Material: primed clad aluminum 2024 (6 mm)> Tool-∅ d: 3 mm; Bore-∅ D: 4.2 mm; Substrate: H10F; Coating: no; Number of cutting edges: 2; Spindle speed ns: 18000–30000 RPM; Orbital rotational speed no: 199 RPM; Feed rate vfa: 60–200 mm/min; Cutting speed vc: 170–286 m/min; Tangential feed ft: 0.013–0.041 mm; Depth setting per helix course ap: 0.15–1 mm
Fig. 28 Burr formation in primed clad aluminum with tool 4 over the cutting speed with two axial feeds fa using different process strategies
Burr and Cap Formation by Orbital Drilling of Aluminum
43
look into the borehole pin, produced with the free grinded front cutting edge
cap
Fig. 29 Cap with pin, tool 5
Figure 28 shows the burr formation versus the cutting speed by using two axial feeds fa with tool 4 for different orbital drilling strategies. The lowest burr is measured in up milling with minimum quantity lubrication. Figure 29 pictures the disadvantage of tool 5. This geometry tends to produce high burr formation due to the free grinded front cutting edges. In this range a small pin of exit material produces a high axial pressure. The cladding material is highly shaped, while in the range of the cap edge the material cracks. The curved exit material is not separated by the peripheral cut from the workpiece. It cracks because of
no = 199 min−1 600
the high plastic deformation. The remaining burr is transformed now completely by the peripheral cutting edges into the final burr. Figure 30 shows the burr formation in orbital drilling with tool 5. High cutting depth settings per orbital path (low orbital speed) with increasing cutting speed tend to high burr formation. In Fig. 31 the burr formation versus the cutting speed using two axial feeds fa with tool 5 for different orbital drilling strategies is shown. Even here the lowest burr is measured in up milling with minimum quantity lubrication.
no = 299 min−1
no = 399 min−1
fa = 0.0033 mm
fa = 0.0066 mm fa = 0.0033 mm
fa = 0.0066 mm
fa = 0.0033 mm
fa = 0.0066 mm
171
171
172
173
173
550 500
burr height [µm]
450 400 350 300 250 200 150 100 50 0 228
284
228
284
172
229
285
229
285
cutting speed vc [m/min]
230
286
230
286
tool 5
Process: Orbital drilling; Method: upmilling; Material: primed clad aluminum 2024 (6 mm); Tool-∅ d: 3 mm; Bore-∅ D: 4.2 mm; Substrate: H10F; Coating: Diamond; Number of cutting edges: 2; Spindle speed ns: 18000-30000 RPM; Orbital rotational speed no: 199–399 RPM; Feed rate vfa: 60–200 mm/min; Cutting speed vc: 170–286 m/min; Tangential feed ft: 0.013–0.041 mm; Depth setting per helix course ap: 0.15–1 mm; No lubrication
Fig. 30 Burr formation in primed clad aluminum with tool 5 over the cutting speed with two axial feeds fa and three orbital speeds no
44
E. Brinksmeier and S. Fangmann no = 199 min−1 up milling
down milling
up milling with MQL
600 fa = 0.0033 mm
fa = 0.0066 mm fa = 0.0033 mm
fa = 0.0066 mm
fa = 0.0033 mm
fa = 0.0066 mm
550 500
burr height [µm]
450 400 350 300 250 200 150 100 50 0 171 228 284 171 228 284 171 228 284 171 228 284 171 228 284 171 228 284
cutting speed vc [m/min]
tool 5
Process: Orbital drilling; Material: primed clad aluminum 2024 (6 mm)> Tool-∅ d: 3 mm; Bore-∅ D: 4.2 mm; Substrate: H10F; Coating: Diamond; Number of cutting edges: 2; Spindle speed ns: 18000–30000 RPM; Orbital rotational speed no: 199 RPM; Feedrate vfa: 60–200 mm/min; Cutting speed vc: 170–286 /min; Tangential feed ft: 0.013–0.041 mm; Depth setting per helix course ap: 0.15–1 mm
Fig. 31 Burr formation in primed clad aluminum with tool 5 over the cutting speed with two axial feeds fa and three orbital speeds no
6 Conclusion and Outlook Extensive orbital drilling experiments on aluminum materials using different tool geometries and process strategies were carried out to investigate the influence on cap and burr formation. The following conclusions can be drawn: The first part of the investigation showed three forms of cap formation in aluminum 2024 T351. The sickle cap is the best cap formation for the manufacturing of closed structures. Sickle caps are produced with high axial feed rates vf,a = 150 mm/min and high depths of cut ap = 0.75 mm. For bore diameters greater than 5 mm a ratio between the bore diameter D and the tool diameter d of D/d ≥ 1.45 produces the best results regarding the cap formation. In the bore diameter range ≥ 5 mm the use of minimum quantity lubrication is possible without declined chip transportation by vacuum if the ratio is D/d ≥ 1.45. Summarising the results of the second part of the investigation the geometry of tool 4 (sharp front cutting edge without radius and a face grinding angle about 10◦ ) shows the best results regarding the cutting forces, the cap and the burr formation in primed clad aluminum 2024 T351. Down
Fig. 32 Tool 4, lowest cap
milling and the use of MQL generally produces lager caps as up und dry milling. In this experiment the lowest caps, easily to transport by vacuum, are produced with the cutting speed vc = 256 m/min, the orbital speed no = 199 RPM and the low axial feed fa = 0.0033 mm (Fig. 32). Acknowledgments Results presented in this paper were achieved with projects funded by the Airbus Deutschland GmbH. The authors gratefully thank the Airbus Deutschland GmbH.
Burr and Cap Formation by Orbital Drilling of Aluminum
References 1. Brinksmeier, E., Fangmann, S., Walter, A., 2007, High SpeedMachining of Multilayer Composite Materials by Orbital Drilling. Sixth International Conference of High Speed Machining, 21–22 March, San Sebastian, Spain:193–197. 2. Dix, M., Leopold, J., Neugebauer, R. 2007, Modelling, Simulations and Experimental Verification of Size Effects in Burr formation, 2nd ICNFT, International Conference on new Forming Technology, 20–21 September, Bias Verlag, Bremen, Germany: 471–480. 3. Heisel, U., Schaal, M., 2008, Burr formation in intersecting holes, Production Engineering (WGP), 2/1:55–62. 4. Stoll, A., Leopold, J., Neugebauer, R., 2005, Modellierung von Größeneinflüssen bei der Gratbildung, Prozessskalierung, Strahltechnik Band 27, BIAS Verlag, 241–252. 5. Basavarajappa, S., Chandramohan, G., Paulo Davim, J., Prabu, M., Mukund, K., Ashwin, M., Prasanna Kumar, M., 2008, Drilling of hybrid aluminum matrix composites, International Journal of Advanced Manufacturing Technology, 35:1244–1250. 6. Sharif, S., Rahim, E. A., 2007, Performance of coated- and uncoated-carbide tools when drilling titanium alloy, Ti-6Al4V, Journal of Materials Processing Technology 185:72–76. 7. Ramulu, M., Branson, T., Kim, D., 2001, A studiy on the drilling of composite and titanium stacks, Composite Structures, 54:67–77.
45 8. Dornfeld, D. A., Kim, J. S., Dechow, H., Hewson, J., Chen, L.J., 1999, Drilling Burr Formation in Titanium Alloy, Ti-6Al-4V, Annals of the CIRP, 48(1):73–76. 9. Brinksmeier, E., Janssen, R., 2002, Drilling of Multi-Layer Composite Materials Consisting of Carbon Fiber Reinforced Plastics (CFRP), Titanium and Aluminum Alloys, Annals of the CIRP, 51(1):87–90. 10. Brinksmeier, E., Krogmeier, F., Walter, A., 2005, Bohrungsbearbeitung von mehrschichtigen Compound-Werkstoffen, Spanende Fertigung, 4. Ausgabe, Vulkan Verlag, Essen:43–50. 11. Tönshoff, H.-K., Friemuth, T., Groppe, M., 2001, High Efficient Circular Milling – A Solution for an Economical Machining of Bore Holes in Composite Materials (CFK, Aluminum), Proc. Third International Conference on Metal Cutting and High Speed Machining, Metz, France, 27–28 June. 12. Weinert, K., Hammer, N., 2004, Zirkularfräsen von Bohrungen im Leichtbau, WB Werkstatt und Betrieb, 137(10):54–58. 13. Brinksmeier, E.; Fangmann, S., 2007, Orbital Drilling of High Tolerance Boreholes, International Conference on Applied Production Technology (APT`07), BIAS-Verlag:75–84. 14. Brinksmeier, E., Fangmann, S, Meyer, I., 2008, Orbital drilling kinematics, Production Engineering (WGP), 2/3:277–283. 15. Luik, M., 2006, Gratbildung und Gratminimierung bei asymmetrisch mit Hartmetall-Wendeschneidplatten bestückten Bohrwerkzeugen, Institut für Werkzeugmaschinen der Universität Stuttgart.
Cutting Force Model for Analysis of Burr Formation in Drilling Process T. Matsumura and J. Leopold
Abstract Burr formation in the drilling process is discussed with the cutting force and the chip flow direction. An analytical model is presented to predict the cutting force based on the minimum cutting energy. Because the burr occurs on the backside of the machined plate in drilling, burr formation is associated with the axial component of the cutting force at the exit of the drill from the workpiece. The chip should also be controlled to flow toward the radial direction of the drilling tool. The lip geometry of the twist drill is discussed to reduce burr formation with the curved edges in the cutting simulation. Keywords Drilling · Simulation · Cutting force · Chip flow direction · Burr formation
1 Introduction Most of manufacturing industries perform a huge number of drilling operations in machine shops. The drilling technology has been studied to improve the cutting performance with optimizing the cutting parameters and the tool geometry. However, burrs are sometimes formed when the drill exits the workpiece and the exit burrs have to be removed in the deburring process. The control of burr formation, therefore, has been strongly required to reduce the post process of the drilling operation. Many researches have been made on burr formation so far. The earlier works associated the burr shape with the cutting parameters experimentally. Some mathematical models based on the experiments were pre-
T. Matsumura () Department of Mechanical Engineering, Tokyo Denki University, 2-2 Kanda Nishiki-cho, Chiyoda-ku, Tokyo 101-8457, Japan e-mail:
[email protected] url: www.skynet.m.dendai.ac.jp J. Leopold Fraunhofer Institute for Machine Tools and Forming Technology, IWU Chemnitz, Reichenhainer Str. 88, 09126 Chemnitz, Germany
sented to determine the cutting conditions [1]. Burr control charts were also proposed to optimize the cutting parameters with considering burr formation [2, 3]. A statistical analysis was presented to estimate the burr height in the drilling process [4]. Burr formation is discussed with the microstructure of the material in micro-scale drilling [5]. Some of researches tried to associate the drill geometries with burr formations. A special tool, which removed the chisel edge and increased the wedge angle and the web thickness, was designed to suppress burr formation and evaluated in cutting of some materials [6]. Step drills were tried to reduce the exit burrs [7]. Because burr formation originally depends on the stress and strain in the material to be cut, deformation of the material at the end of the cut has been analyzed in the finite element method. The FE analysis is effective to understand the material behavior in the process [8–10]. However, it has not been applied to the design of the drill geometry yet because it takes a long time for the simulations. Other analytical models, therefore, is required to optimize the geometry in the simulation. From the point of view of the cutting force, the axial component at the end of the cut promotes burr formation in drilling of the plates. When the cutting chip also flows upward, a large burr is left at the edge of the hole. Therefore, the axial component of the cutting force and the chip flow direction should be controlled to reduce burr formation. This study focuses on the cutting force and the chip flow at the exit of the drill from the workpiece. An analytical model based on the minimum cutting energy is presented to predict the cutting process. Many force models in the drilling process have also been presented as well as those of the milling one so far [11, 12]. A model based on the minimum cutting energy was proposed to predict the cutting force with the chip flow direction in turning [13] and was applied to the drilling process [14]. However, the transient cutting processes at the penetration and the exit processes of the drill have not been analyzed yet in the model. In order to reduce burr formation at the backside of the machined plate, the presented model simulates the axial component of the cutting force with the chip flow direction in the transient cutting process when the drilling tool exits the workpiece. Then, the lip geometry of
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_5, © Springer-Verlag Berlin Heidelberg 2010
47
48
T. Matsumura and J. Leopold
the twist drill is discussed to reduce burr formation with the cutting force and the chip flow. Several curved lip shapes have recently been applied to the machine shops in the automobile industry. The paper shows the effect of the curved lip shape on the cutting process at the exit of the drill in the simulation.
Wedge angle
Relief angle
Helix angle Web Inclination Diameter
2 Force Model in Drilling Process
Web size
2.1 Definition of Tool Geometry The general geometry of twist drills is defined as shown in Fig. 1. The twist drills consist of the chisel and the lip. The paper discusses the cutting process of a drill with a X type thinning, which inclines at an angle of 0◦ in the axial direction. Because burr formation strongly depends on the cutting process with the outer side of the lips, the lip geometry is discussed with not only the straight edge but also the curved one. Table 1 shows the tool geometry with the edge shape viewed from the bottom of the tool, where the lines show the cutting edges. Figure 2 shows the change of the axial and the radial rake angles with the rotational radius of the cutter, which can be derived from the shape of the grinding stone in manufacturing of the drills.
Axial inclination angle
Web center thickness Fig. 1 Parameters of drill geometry
Table 1 Tool geometry Diameter Web inclination Web size Wedge angle Helix angle Relief angle Web center thickness Thinning Axial inclination angle at thinning
10 mm 135◦ 1.52 mm 118◦ 30◦ 10◦ 0.2 mm X type 0◦
5
4
4
4
3
3
3
2
2
2
1
1
1
0 −1
0 −1
Y mm
5
Y mm
Y mm
Lip shape 5
0 −1
−2
−2
−2
−3
−3
−3
−4
−4
−4
−5 −5 −4 −3 −2 −1 0 1 2 3 4 5 X mm
−5 −5 −4 −3 −2 −1 0 1 X mm
−5 −5 −4 −3 −2 −1 0 1 X mm
Straight
2
3
4
5
Clockwise curved roundness: 4mm
2
3
4
5
Counterclockwise curved roundness: 4mm
49
40
40
20
20
Rake angle deg
Rake angle deg
Cutting Force Model for Analysis of Burr Formation in Drilling Process
0 −20 Axial Radial
−40 −60
–20 –40 –60
0
1
2
3
4
5
Axial Radial
0
0
1
Radius mm (a) Lip shape, straight
2 3 4 Radius mm (b) Lip shape, clockwise
5
40 Rake angle deg
20 0 –20 –40 Axial Radial
–60 –80
0
1
2 3 Radius mm
4
5
(c) Lip shape, counterclockwise
Fig. 2 Axial and rake angles of drills
2.2 Chip Formation in the Drilling Process
2.3 Analysis of Cutting Force
The chip formations on a chisel and a lip were observed in the cutting experiments. Figure 3 shows a picture of the chips at the interruption of cutting. The picture proves that the cutting chip is formed on the chisel as well as the lip. However, the chip formation on the chisel is different from that of the lip. The chip flows on the chisel and that of the lips should be considered independently to predict the cutting processes.
An analysis model is presented to predict the cutting force with the chip flow direction in the drilling process as shown in Fig. 4. The chip flow is interpreted as a piling up of the orthogonal cuttings in the plane containing the cutting velocity V and the chip flow velocity Vc . The cutting edges are divided into small segments to make the orthogonalcutting models using the following data:
Rotation axis
Chip formation on a chisel edge Chip
Chip flow angle ηc Chip flow velocity Vc P Cutting velocity V
Orthogonal cutting
Chip formation on a cutting lip Fig. 3 Chip formation in the drilling process
Plane containing V and Vc Fig. 4 Chip flow model
50
T. Matsumura and J. Leopold
⎫ φ = f (V,t1 ,α) ⎪ ⎪ ⎬ τs = g (V,t1 ,α) ⎪ ⎪ ⎭ β = h (V,t1 ,α)
(1)
where φ, τs and β are the shear angle, the shear stress on the shear plane and the friction angle. V, t1 and α are the cutting velocity, the undeformed chip thickness and the rake angle. Equation (1) can be obtained in the orthogonal cutting tests. Because the cutting velocity V is the sum of the rotational speed and the feed rate f, V is given by:
where φ ∗ e is be given by Eq. (1) at α ∗ e , V∗ and t∗ 1 . Then the shear stress on the shear plane τs and the friction angle β can also be given by Eq. (1). The shear energy in a segmented area dUs is: dUs = τs ls dLs
(RP ω)2 + f 2
φ=
δs + φe∗
(3)
αe α*e δs
δs Vc
Q
A
V* V
δs
(5)
where dFt is given in the orthogonal cutting model: dFt = τs t1∗
cos
φe∗
sin β dLf + β − αe∗ sin φe∗
(6)
where dLf is the width of the tool chip contact area in a segmented area. The chip flow velocity at the center of the cutting area removing material is: sin φe∗ + δs V Vc = cos φe∗ + δs − αe
(7)
The chip flow velocities in the other segmented areas on the cutting edge are calculated geometrically to keep the angular velocity of the chip curl constant without the plastic deformation in the chip. The cutting energy U, then, is given by the integration over the height [hmin , hmax ] of the cutting area as follows: U=
h max
(dUs + dUf )dh
(8)
h min
The chip flow angle ηc can be determined to minimize U in the iterative calculation. The cutting force,then, is predicted in the model at the minimum cutting energy. Figure 6 shows an orthogonal cutting plane with cutting force components loaded on edge point P. X -Y -Z is the coordinate system rotating with cutting edge at angular velocity ω, where the direction of cutter radius is defined as X -axis. The tangential cutting force in a segmented area dFH is:
dFH = dUs + dUf V
t*1 t1
dUf = dFt Vc
(2)
where ω and RP are the angular velocity and the radius at a position P on an edge. Because the cutting direction is inclined at an angle of tan–1 (f/Rp ω), the axial rake angle during cutting is regarded as α A = α A + tan–1 (f/Rp ω), where α A is the nominal angle given by the tool geometry. When a chip flow angle ηc on the rake face is assumed in Fig. 4, the orthogonal cutting model can be made by Eq. (1) with calculating the effective rake angle α e . The workpiece surface is inclined with respect to the cutting direction when the edges enter and exit workpiece, where the undeformed chip thickness increases and decreases with the cutter feed. Therefore, the orthogonal model in such the transient cutting process is made in the coordinates system inclined at the angle δ s as shown in Fig. 5. The rake angle of the cutting edge is regarded as α ∗ e = α e –δ s in the model. The cutting edge removes the material in the direction of the surface inclination at a cutting velocity of V∗ = V{cosδ s –sinδ s tan(α e –δ s )} in an undeformed chip thickness of t∗ 1 = t1 sinφ ∗ e /sinφ. The nominal shear angle φ is:
(4)
where ls and dLs are the length and the width of the shear plane on a segmented area. The friction energy dUf is given by the friction force dFt and the chip flow velocity Vc :
V=
cos αe V cos φe∗ + δs − αe
(9)
φ*e The normal force on the rake face dFn is given by:
P Fig. 5 Orthogonal cutting model in transient cutting process
dFn =
dFH − dFt sin αe cos αb cos αR
(10)
Cutting Force Model for Analysis of Burr Formation in Drilling Process Table 2 Cutting conditions
Rake face
Z' αR Y'
(dFx')1 A H I α e ηc η0 η'c (dFz')1 αb α'A
O'
(dFx')2 αR
P φe
Aluminum alloy (ADC12) Coated HSS 640 rpm 110 mm/min 19 mm
Cutting edge (dFz')2
αb
Q
dFH
Material cut Tool material Spindle speed Feed rate Width of plate
dFt
X' dFn Measured thrust Measured torque Calculated thrust Calcualed torque
600
Orthogonal cutting plane
500
Fig. 6 Components of cutting force
where dFt is the friction force given by Eq. (6). α R is the radial rake angle of the edge viewed from the inclined direction at an angle of tan–1 (f/Rp ω); and α b is the inclination angle of the rake face with respect to Z-axis direction. The radial component dFT and the axial one dFV are given by: dFT = −dFt cos αe sin ηc + dFn cos αb sin αR dFV = dFt cos αe cos ηc − dFn sin αb
Thrust N Torque Ncm
Rotation axis n
51
Thrust
400
C
B
300
D 200 100
A
Torque
E
0 O –100
0
5
10
15
Time s
(11)
ηc is the projected angle of the chip flow direction onto the vertical plane. dFH , dFT , and dFV is converted into the components of X-, Y-, and Z-axis. The chip flow models are made on the chisel and the lip independently to calculate the cutting forces with the chip flow directions. The thrust is the sum of Z component of the cutting forces loaded on all of the cutting edges. The torque can be given by the integration over the radius [Rmin , Rmax ] of the cutting area as follows:
Fig. 7 Cutting force in drilling process
Figure 7 shows the change of the thrust and the torque with the cutting time. The process consists of the following steps according to the cutting position:
3.1 Case Study
(1) The chisel penetrates into the material. The cutting area on the chisel increases with the cutter feed (Process O-A). (2) The chisel and the lips remove the material. The cutting areas on the lips increases with the cutter feed (Process A-B). (3) The chisel and the lips remove the material in the steady cutting process, where the cutting areas do not change (Process B-C). (4) The chisel exits the material. The cutting area on the chisel decreases with the cutter feed (Process C-D). (5) The lips exit the material. The cutting areas on the lips decreases with the cutter feed (Process D-E).
The simulation based on the presented model was performed to predict the cutting force in the drilling process in the conditions of Table 2. The following data was used in the simulation: ⎫ φ = 0.073124V − 47656.1t1 + 1.5328α − 2.0940 ⎪ ⎬ (13) τs = 0.008939V − 32602.7t1 + 0.3000α + 20.0459 ⎪ ⎭ β = −0.040498V + 9636.43t1 + 0.6328α − 0.2913
The thrust and the torque change with the depth of the cutting position according to the above steps in the simulation. The cutting force increases rapidly at the chisel penetration; and increases gradually during the lips penetration. The cutting force does not change when the chisel and the lips remove the material simultaneously. The cutting force drops suddenly just after the exit of the chisel from the material; and then decreases gradually with the cutter feed.
T=
R max
r · dFH dr
(12)
R min
3 Force Analysis for Burr Formation
52
T. Matsumura and J. Leopold
The simulation based the presented model can be verified in Fig. 7. The simulated results agree with the measured forces not only in the steady process but also the transient one.
3.2 Effect of Tool Geometry on Burr Formation The cutting simulations were tried to discuss the effect of the lip shape on the cutting force and the chip flow. The curved lip drills shown in Table 1 were examined in the simulation. The cutting conditions are the same as those in Table 2. Figure 8 shows the thrust and the torque of the drill having the clockwise curved lips with the cutting time. Figure 9 shows those of the drill having the counterclockwise curved lips. The cutting force of the clockwise curved lips is almost
same as that of the straight lips in Fig. 7. However, the cutting force of the counterclockwise curved lips is slightly larger than those of the straight and the clockwise lips due to large negative radial rake angles around the connection of the thinning and the lip, where the rotational radius is 1.06 mm in Fig. 2(c). Figure 10 compares the axial components of the cutting forces loaded on a cutting edge around the end of the cut, where the drill exits the workpiece. Because the examined drills have two edges, the total axial components are twice as large as the results in Fig. 10. The axial component of the straight lips is largest in the examined tools. The curved lips, therefore, is expected to reduce burr formation with the axial component. Figure 11 shows the chip flow angles, which is the inclination angle with respect to the axial direction as shown in Fig. 4. The chip flow angles of the curved lips are larger than the angle of the straight lips due to the
100
600
Thrust N Torque Ncm
Axial component N
Thrust
500
Torque
400
Thrust
300 200 100
Torque
Straight Clockwise Counterclockwise
80 60 40 20
0 –100
0
5
Time s
10
0 11
15
11.6
60 Chip flow angle deg
Thrust Torque
500 400
Thrust
300 200 Torque
100 0 0
5
11.8
12
Fig. 10 Change of axial component loading on a lip. Tool geometry and cutting conditions are shown in Tables 1 and 2
600
Thrust N Torque Ncm
11.4
Time s
Fig. 8 Cutting force of clockwise curved drill. Tool geometry and cutting conditions are shown in Tables 1 and 2
–100
11.2
10
15
Time s Fig. 9 Cutting force of counterclockwise curved drill. Tool geometry and cutting conditions are shown in Tables 1 and 2
Straight Clockwise Counterclockwise
50 40 30 20 10 0 11
11.2
11.4
11.6
11.8
12
Time s Fig. 11 Change of chip flow angle. Tool geometry and cutting conditions are shown in Tables 1 and 2
Cutting Force Model for Analysis of Burr Formation in Drilling Process
change of the radial rake angle as shown in Fig. 2. The large chip flow angle means that the cutting chip flows toward the radial direction of the tool. The curved lip is effective for the reduction of burr formation with changing the chip flow direction.
4 Conclusion An analytical force model was presented to predict the cutting force and the chip flow direction. Because burr formation occurs on the backside of the machined plate, this study focused on the cutting process around the exit of the drill from the workpiece. The presented model interprets the chip flow in drilling as a piling up of the orthogonal cuttings in the plane containing the cutting velocities and the chip flow velocities. The orthogonal cutting data, which associates the shear angle, the shear stress on the shear plane and the friction angle with the cutting velocity, the undeformed chip thickness and the rake angle of the tool, are employed to make the piled cutting models with calculating the cutting energy consumed in the models. Because the cutting energy of a chip flow model can be calculated with assuming the chip flow angle, the chip flow direction is determined to minimize the cutting energy. The force model, therefore, predicts not only the cutting force but also the chip flow direction. This study associates the axial component of the cutting force and the chip flow direction with burr formation. The small axial components with the chip flow toward the radial direction of the tool can be expected to reduce burr formation at the end of the cut. The simulations were performed in the drilling processes with changing the lip shape. Little change was observed in the maximum cutting forces of the curved lip drills. The axial component of the cutting force at the end of the cut can be reduced by the curved lips. The chip flow angle increases in cutting with the drill having the curved lips. Although this paper shows the effect of the curved lips shape on burr formation in the simulation, the presented model can be utilized in design of the drill geometry with considering burr formation.
53
References 1. Pande, S. S., ReleKar, H. P., 1986, Investigations on Reducing Burr Formation in Drilling, International Journal of Machine Tool Design Researches, 26, 3 339–348. 2. Min, S., Kim, J., Dornfeld, D. A., 2001, Development of a Drilling Burr Control Chart for Low Alloy Steel, AISI 4118, Journal of Material Processing Technology, 113, 4–9. 3. Kim, J., Min, S., Dornfeld, D. A., 2001, Optimization and Control of Drilling Burr Formation of AISI 304L and AISI 4118 Based on Drilling Burr Control Charts, International Journal of Machine Tools & Manufacture, 41, 923–936. 4. Lauderbaugh, L. K., 2008, Analysis of the Effects of Process Parameters on Exit Burrs in Drilling Using a Combined Simulation and Experimental Approach, Journal of Material Procesing Technology, doi:10.1016/j.jmatprotec.2008.04.062. 5. Sugawara, A., Inagaki, K., 1982, Effect of Workpiece Structure on Burr Formation in Micro-Drilling, Precision Engineering, 4, 1, 9–14. 6. Ko, S. L., Lee, J. K., 2001, Analysis of Burr Formation in Drilling with a New-Concept Drill, Journal of Material Processing Technology, 113, 392–398. 7. Ko, S. L., Changa, J. E., Yang, G. E., 2003, Burr minimizing Scheme in Drilling, Journal of Material Processing Technology, 140, 237–242. 8. Marusich, T. D., Usui, S., Ma, J., Stephenson, D. A., Shih, A., 2007, Finite Element Modeling of Drilling Process with Solid and Indexable Tooling in Metal and Stack-ups, 10th CIRP International Workshop on Modeling of Machining Operations, 51–57. 9. Regel, J., Stoll, A., Leopold, J., 2007, Numerical Analysis of Crack Propagation during the Burr Formation Process of Metal, 10th CIRP International Workshop on Modeling of Machining Operations, 125–132. 10. Makarov, V. F., Chigodaev, N. E., Tokarev, D. I., 2007, 10th CIRP International Workshop on Modeling of Machining Operations, 97–102. 11. Armarego, E. J., Cheng, C. Y., 1972, Drilling with Rake Face and Conventional Twist Drills. 1: Theoretical Investigation, International Journal of Machine Tool Design Researches, 12, 17. 12. Elhachimi, M., Torbaty, S., Joyot, P., 1999, Mechanical Modelling of High Speed Drilling. 1: Predicting Torque and Thrust, International Journal of Machine Tool Design Researches, 39, 553–568. 13. Usui, E, Hirota, A., Masuko, M., 1978, Analytical Prediction of Three Dimensional Cutting Process – Part1 Basic Cutting Model and Energy Approach, ASME Journal of Engineering for Industry, 100, 222–228. 14. Hirota, A., Tanaka, M., Kasahara, K., 1981, Analytical Preditcion of Chip Formation and Cutting Forces in Drilling Operation (1st Report), Journal of Japan Society for Precision Engineering. 47, 8, 987–992.
Burr Formation in Microstructuring Processes B. Denkena, L. de Leon, and J. Kästner
Abstract Due to high form accuracy and flexibility concerning the structure geometry, cutting is predestined to produce structures in the micron range. At the Institute of Production Engineering and Machine Tools (IFW), a fly-cutting process is used for manufacturing microstructures. These structures are used to improve the tribological performance of thermo mechanically highly stressed surfaces. The structures thus produced are able to hold some lubricant and can also be used to increase the hydrodynamic pressure built up between the friction partners already at low sliding speeds. When machining the small structures, material of the compression zone in front of the cutting edge is squeezed in the direction of the free lateral surface areas and causes burrs. For an unrestricted functionality of the structured surface and for avoiding down-stream finishing processes, these burrs have to be minimized. In this paper, the basic relationship between the cutting edge geometry and the burr formation in machining ductile materials with depths of cut below 50 μm will be investigated. In particular, the influences of the material, the tool cutting edge radius and the tool cutting edge angle on the cutting forces and burr formation will be described. Based on these correlations, a special cutting edge microgeometry will be developed which prevents material flow in the direction of the lateral free surface areas. Keywords Microstructuring · Fly-cutting · Burr formation
1 Introduction Microstructured surfaces offer properties that are of high interest for tribological systems which are subject to high mechanical and thermal stress. A well directed structuring
of highly loaded surface areas can reduce the effects of mixed friction. Consequently, friction losses and wear are minimized [1, 2]. Microstructured surfaces are able to hold lubricant, which can improve the dry running properties considerably. Furthermore, they can act as so-called micropressure chambers. Due to their closed structure, micropressure chamber systems prevent lateral lubricant leakage. The hydrodynamic effect of microstructures is illustrated in Fig. 1. When the friction partners are moved in relation to each other, the lubricant is compressed in the narrowing lubrication gap and causes a hydrodynamic floating of the conforming friction partner. This effect is supported by microstructures already at low sliding speeds. The bigger the pressure which is built up, the more rarely do mixed friction effects appear. Cylinder running surfaces are a typical application for microstructures. At the piston assembly group, especially the piston/piston-ring connection, the areas of dead centres are affected by mixed friction due to insufficient lubricant supply and low sliding speeds. Deterministic structures generated by a laser-based material treatment are already representing a promising approach for reducing friction losses and wear without increasing emissions caused by lubricants [2]. Due to high form accuracy, low surface roughness and very high flexibility regarding the structure geometry, cutting processes are ideally suited for the production of structures in the micron range [3]. The focussed structure dimensions have depths in the range of 10–30 μm. Within these dimensions, material properties like grain boundaries, inclusions and pores as well as hardness deviations have a significant impact on the cutting forces and the geometry of the generated structure [4–6].
B. Denkena, L. de Leon, J. Kästner () Institute of Production Engineering and Machine Tools, Leibniz Universität Hannover, An der Universität 2, 30823 Garbsen, Germany e-mail:
[email protected]
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_6, © Springer-Verlag Berlin Heidelberg 2010
55
56
B. Denkena et al.
Fig. 1 Pressure due to microstructures
2 Burr Formation in Microstructuring Processes 2.1 Experimental Setup of the Fly-Cutting Process In order to investigate the basic interactions between the material, the process parameters and the resultant structure, fly-cutting experiments with a Blohm-Profimat 307 profile grinding machine were conducted on plane surfaces. This machine meets the high demands on the positioning accuracy (≤ 0.5 μm) and repeatability (≤ 1 μm) in the production of microstructures and is thus suited for the investigations. The kinematics of the fly-cutting process are the following: By an axial movement of the machine spindle, the rotating
Fig. 2 Experimental setup of the fly-cutting process
tool, mounted on the peripheral surface of an aluminum disk, is moved along the workpiece on a helix path, whereas the workpiece is clamped in a slightly inclined position. Thus, several parallel structures are produced within one movement along the workpiece with a constant increase in the depth of cut. Due to the incline of the workpiece, depths of 0–50 μm can be produced within one movement along the workpiece. The axial feed is higher than the maximum structure width to be expected, which means adjoining structures cannot impair each other. The experimental setup and the measuring section for the measurement of the cutting forces are shown in Fig. 2. To guarantee an exact adjustment of the specimen, a levelling device which enables a vernier adjustment within two independent directions was developed. Thus the surface of the specimen can be positioned accurately. In order to avoid
Burr Formation in Microstructuring Processes
vibrations due to the discontinuous engagement of the single cutting edges, the platform itself is prestressed by an inner spring system. In preexaminations, no influence of potential vibration on the development of the cutting force was observed. During the test series, the cutting forces, the passive forces and the feed force (for asymmetrical tool profiles) are measured with a piezoelectric three-component force measurement platform of the type KISTLER 9253-A2. Due to the small dimensions of 91 mm × 80 mm × 25 mm, the resultant high natural frequencies of f0 (x) = 3.8 kHz, f0 (y) = 3.5 kHz and f0 (z) = 4.5 kHz and a responsiveness < 2 mN, the system is well suited for the highly dynamic measurement of small forces. The lowest natural frequency of the measuring setup is 340 times higher than the maximum operating frequency used in the experiments, e.g. 10 Hz at vc = 480 m/min. Therefore no superposition of these frequencies is to be expected. The force curve follows a symmetrical parabolic arc during the period of engagement of the single cutting edges along one structure. The maximum forces at the maximum depth of cut within one structure will be used for the evaluations. The evaluation of the forces comprises the range of the depth of cut of 0–50 μm. The maximum structural depth of one band is determined with a confocal white light microscope of the type Nanofocus μ-Surf, which scans the layers of the surface by moving along an optical axis. An objective lens with a fiftyfold magnification is used for these measurements, which leads to a lateral resolution of 0.3 μm and a vertical resolution of 0.0015 μm. Thus the forces can be plotted against the depth of cut. By means of the measured depth and the detailed tool profile, the cross-sections of undeformed chip are calculated. Consequently the specific forces can be plotted. Within the frame of this work, the specific forces are the main basis for comparison because, depending on the tool profile, they are influenced by both the width of cut, which varies with the depth of cut, and the machining conditions. Furthermore, scaling effects can be identified.
57
ished surface of steel S235JR. The irregular run of the burr along the edge of the structure is clearly visible. These burrs, formed by plastic material flow in the direction of the lateral free surfaces, have to be removed by additional finishing procedures. Therefore, burr formation has to be avoided or minimized by applying adapted cutting edge geometries and process parameters. For the quantification of the burr dimensions, the relative chipping volume is calculated according to [7]. This value is used to determine the efficiency of machining processes and describes the ratio between the cross section of the removed material and the bulging cross section. The structures produced by machining processes within the frame of this work have no constant burrs, so the volumes are used instead of the cross sectional areas (Fig. 3). Therefore, the structures are measured with a confocal white light microscope of the type Nanofocus μ-Surf. The set of measured data is evaluated with the software Nanoexplorer Premium 4.1. With this software, the volumes above and beneath a definable area can be calculated and the relative chipping volume can be determined. SEM micrographs and topographic images are used for describing the burr shape and the burr distribution.
3 Influence of the Process Input Parameters on the Burr Formation The machining experiments show that in microstructuring, besides the workpiece material, the geometric dimensions in the orthogonal plane and the profile plane of the tool also have a significant influence on the burr formation. In the following, these characteristics will be examined regarding their influence on the burr formation and the mechanisms involved.
2.2 Characterization of the Burr Formation
3.1 Influence of the Material and the Cutting Edge Radius on the Burr Formation
The SEM-micrographs in Fig. 3 show microstructures with a maximum depth of cut of ap = 30 μmproduced on a pol-
Figure 4 shows the basically different structural conditions of the investigated materials and contrasts the produced surface
Fig. 3 Quantification of the burr dimensions by means of the relative chipping volume
58
B. Denkena et al.
Fig. 4 Influence of the structural conditions of employed materials on the cut surface topography
topographies with each other. The materials used are a hypereutectic Aluminum-Silicon alloy (AlSi17Cu4Mg), globular cast iron (GJS400) and an iron molybdenum layer produced by plasma spraying (Fe50Mo). These materials are very frequently used for tribological systems subject to high thermomechanical stress, like cylinder liners. Due to its fine grained structure the steel S235JR is used as reference material. The notchy surface of the primary silicon crystals on the AlSi17Cu4Mg alloy, based on the brittle material separation, differs significantly from the cut surface topography of the aluminum matrix. Furthermore, silicon crystals are ripped out of the matrix during the separation process or dislocated within the matrix. In the cut surface topography of the globular cast iron, there are open graphite pores which partly bunch out of the cut surface. In the cross-section polish, it is clearly visible that the subsurface has been plastically deformed by the separation process. In the transition sections between the graphite pores, ferrite edges and the pearlitic basic matrix, the edge layer is interrupted due to the different mechanical phase properties of these sections and is partly peeled off the substrate material. In spite of the inherently porous structure of the material, the cut surface of the thermally sprayed FeMo layer suggests that ductile material separation has taken place. Outbursts of single lamellae and of complete groups of lamellae mainly occur at the transition sections to the free lateral surface areas, which is due to insufficient reinforcement within these sections. The polished cross-sections and the cut surface topography of S235JR show that the subsurface is also plastically deformed during the separation process. Besides the reproduction of the chipping of the tool cutting edge, no structural influence can be observed. The topographic images show that AlSi17Cu4Mg and GJS400
are much more sensitive to burr formation than S235JR and Fe50Mo. This can also be derived from the relative chipping volume, which tends to decline with an increase in the cutting edge radius and from the resultant increase in the burr dimensions (Fig. 5). It can be observed that the burr dimensions increase much more with AlSi17Cu4Mg (3.1%/μmrβ ) and GJS400 (2.61%/μmrβ ) compared with S235JR (0.4%/μmrβ ) and Fe50Mo (0.3%/μmrβ ). It has been proven that materials which already tend to an increased burr formation, if ideally sharp cutting edges are applied, will tend to much more increased burr dimensions if large cutting edge radii are applied. With large cutting edge radii, the thermally sprayed layer slightly tends to burr formation, too, but there are considerably less lamellae outbursts in the areas of the structure flanks. It is assumed that the higher deflection forces in the direction of the lateral free surface areas, which are due to a big cutting edge radius, work against the entrainment of lamellae. For all applied materials, the cutting forces and the passive forces tend to increase with the cutting edge radius. The highest difference in the force level can be observed between 5 and 45 μm. Figure 6 shows the development of the cutting forces for the hypereutectic AlSi alloy and the Fe50Mo layer. The influence of the cutting edge is very low in the machining of the AlSi alloy, even at low depths of cut. Only the cutting edge radius of 45 μm leads to slightly increased cutting forces and considerably increased passive forces (≤ 80%) at depths of cut between 10 and 20 μm. With the FeMo layer, the specific forces increase more considerably if the cutting edge radius is increased. The cutting force and its change due to the cutting edge radius is similar to that
Burr Formation in Microstructuring Processes
59
Fig. 5 Burr formation depending on material and cutting edge radius rβ
of S235JR and GJS400. However the most significant influence of the cutting edge radius on the specific passive forces is in the machining of the Fe50Mo layer. With a cutting edge radius of 45 μm, these forces are increased by almost 300% compared to a cutting edge radius of 5 μm. It is assumed that this extreme increase is due to the very low ductility of the material, which prevents the relief of stress due to plastic deformation. In analogy to [8], the increase in the passive forces with higher cutting edge radii observed for different materials correlates with the increase in the burr dimensions.
3.2 Influence of the Tool Cutting Edge Angle on the Burr Formation Preexaminations already showed that there is a relationship between the tool cutting edge angle and the burr dimensions at the structure flanks. In order to investigate this relationship and to determine the limit ranges, the tool cutting edge angle is varied between 5◦ and 95◦ and the relative chipping
volume is calculated in order to classify the burr dimensions (Fig. 7). The inclination of the tool leads to asymmetrical structure profiles and an increased tool cutting edge angle at the one and a decreased angle at the opposite structure flank. A proof of the relationship between the tool cutting edge angle and the burr dimensions is given by the three-dimensional topographic images. The burr dimensions increase progressively with the tool cutting edge angle. There is an exponential decrease in the relative chipping volume for κ = 5◦ –90◦ . For tool cutting edge angles above κ = 90◦ , there is a decrease in the burr dimensions. However the mechanisms of the burr formation in the undercut area will not be investigated within the frame of this work. With an increase in inclination of the tool, a slight increase in the burr dimensions of the whole structure (left and right structure flank) has been observed in the experiments. This means that the increase in the burr volume at one structure flank is higher than the decrease in the volume at the opposite structure flank. This development is reversed for tool cutting edge angles above κ = 90◦ . Additionally, the decrease in the burr dimensions is clearly visible
60
Fig. 6 Specific cutting forces depending on material and cutting edge radius
Fig. 7 Influence of the tool cutting edge angle on the burr dimensions
B. Denkena et al.
Burr Formation in Microstructuring Processes
61
Fig. 8 Geometry and cutting performance of the adapted cutting edge
in the topographic images. From a certain limit of tool cutting edge angle, no burrs are visible. If a sharp cutting edge with a cutting edge radius of 5 μm is used, this limit is at κ = 10 –20◦ when machining steel S235JR. This is also confirmed by the fact that the relative chipping volume converges the value one asymptotically, where no burr occurs. The fact that the burr dimension decreases with a decrease in the tool cutting edge angle is due to the inhibition of material flow in the direction of the lateral free surfaces (which is itself due to a higher deflection). Furthermore, the passive force causes thrust forces which are pointing orthogonally from the cutting edge to the workpiece material and thus determine the direction and development of material deformation at the edge of the structure (cutting edge exit in the
Fig. 9 Deburring of microstructured surfaces
profile layer). With an increase in the tool cutting edge angle, the vector of this thrust force is shifted towards the lateral free surfaces and thus leads to an increase in the burr dimensions [9, 10]. In order to use the positive effect of small tool cutting edge angles and to avoid the resultant increase in the structure volume at the same time, a tool profile with a gradually decreasing cutting edge angle is developed. The necessary concave chamfer at the cutting edge corner is shown in Fig. 8. In this case, the minimum radius which can be produced by grinding is rε,2 = 50 μm. The cutting performance of the adapted tool geometry confirms the deburring mode of operation. However total deburring occurs when depth of cut equals the depth of the
62
concave chamfer (ap,limit ). When depth of cut goes below ap,limit burr dimensions increase again due to the increase of the effective tool cutting edge angle. Consequently the presented cutting edge geometry can be used for deburring only in the middle of the structure. When ap exceeds ap,limit burr formation also increases.
4 Deburring Process By means of adapted tool geometries, the burr dimensions can be decreased considerably, but they can only be completely avoided in rare cases. But for ideal tribological characteristics of microstructured surfaces and for avoiding damages during the running-in, the structures have to be free of burrs. In the experiments conducted within the frame of this work, the burrs have been removed mechanically with extremely fine abrasives and by polishing. Figure 9 shows one surface with symmetrical structures and one surface with asymmetrical structures both before and after the deburring procedure. The SEM micrographs show that the burrs are completely removed. Furthermore, there are no remains of burr in the structures or in the areas of the structure edges which could have had a negative influence on the tribological characteristics of the surface and reinforce surface wear in case of delamination.
5 Summary The investigations presented in this paper treated the burr formation in microstructuring by means of a flycutting process and the essential determining factors on this burr formation. The lateral burrs in the area of the structure edges are due to material flow towards the lateral free surfaces. For the production of microstructures with minimized burrs, small cutting edge radii and small tool cutting edge angles have proven to be successful. They reduce the volume of deforming material in the compressive zone in front of the cutting edge and inhibit the material flow in the direction of the lateral free surfaces. By means of an adapted cutting edge geometry which holds both of these properties, burr free structures
B. Denkena et al.
were produced. Furthermore, it has been shown that adequate mechanical deburring processes have no influence on the structure geometry and the surface quality. Future investigations will focus the causes and mechanisms of burr formation more in detail. By a targeted investigation of chip formation and work piece microstructures as well as highly dynamic force measurement, the knowledge on the mechanisms of burr formation will be gained. Furthermore, the influence of the surface topography and the subsurface will be included in the investigations. Acknowledgments The investigations presented in this paper are supported by the German Research Foundation (DFG) within the research unit 576 “Microstructuring of thermomechanically highly stressed surfaces”.
References 1. Lindner, H.; Bermann, H.W.; Brandenstein, C; Lang, A.; Queitsch, R; Reichsten, S.; Stengel, E.: Precision Processing of Cast Iron Cylinder Surfaces of Combustion Engines with UV-Photons. University of Bayreuth, Bayreuth, 2003 2. Golloch, R.; Merker, G.P.; Kessen, U.; Brinkmann, S.: Benefits of Laser-Structured Cylinder Liners for Internal Combustion Engines. 14th International Colloquium Tribology, Technische Akademie Esslingen, Ostfildern, 2004 3. Fischer, S: Fertigungssysteme zur spanenden Herstellung von Mikrostrukturen. PHd-Thesis, RWTH Aachen, 2000 4. Bissaco, G.; Hansen, H.N.; De Chiffre, L.: Size Effects on Surface Generation in Micro Milling of Hardened Tool Steel. Annals of the CIRP Vol. 55, No. 1, 2006, Technical University of Denmark, pp. 593–596 5. Quoteiro, J.C.; Dias, A.M.; Jawahir, I.S.: On the Effects of Residual Stresses Inducted by Coated and Uncoated Cutting Tools with Finite Edge Radii in Turning Operations. Annals of the CIRP Vol. 55, No. 1, University of Coimbra Portugal, University of Kentucky, 2006, pp. 111–116 6. Hüntrup, V.: Untersuchungen zur Mikrostrukturierbarkeit von Stählen durch das Fertigungsverfahren Fräsen. PHd-Thesis, Universität Karlsruhe, 2000 7. zum Gahr, K.H.; Mewes, D.: Werkstoffabtrag beim Furchungsverschleiß. Metall 37, 1983, pp. 1212–1217 8. Riemer, O.: Trennmechanismen und Oberflächenfeingestalt bei der Mikrozerspanung kristalliner und amorpher Werkstoffe. PHdThesis, Universität Bremen, 2001 9. Schäfer, F.: Entrgaten in der Fertigungstechnik – Grundlagen des Entgratens, Werkstatttechnik-Zeitschrift für industrielle Fertigung, Ausgabe 63, 1973, pp. 692–696 10. Schäfer, F.: Entgraten – Theorie, Verfahren, Anlagen. Krauskopfverlag, 1975
Analytical Modeling and Experimental Investigation of Burr Formation in Grinding H. Sudermann, I.G. Reichenbach and J.C. Aurich
Abstract Increasing industrial requirements on the precision of edge geometry lead to the investigation of burr formation, particularly in finishing operations such as grinding. In the presented paper the burr formation process at the workpiece edges during surface grinding is analyzed. Based upon a geometrical examination of the macroscopic contact situation between grinding wheel and workpiece, a qualitative analytical process model of burr formation is developed. Following experimental investigations verify this qualitative model and enable quantitative results. Several so called “grinding stops” are realized by specifically interrupting the grinding process. Thereby, the burrs at the edges of the temporary contact area are detected. To analyze the burr formation, metallographic sections of the workpiece edges are prepared and characteristic burr parameters are measured. Based on experimental results, a thermal impact on burr formation in grinding can be detected. Keywords Grinding · Burr formation modeling · Experiment · Thermal impact
·
Analytical
1 Introduction Reliable and cost efficient manufacturing of workpieces with defined geometrical and technological properties is a main objective of industrial production. Due to plastic material deformation, cutting operations always lead to the creation of micro- or macroscopic burrs. If the size and shape of burrs affects the functionality of a workpiece, a subsequent deburring process is required. This additional production step
H. Sudermann (), I.G. Reichenbach, J.C. Aurich Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany e-mail:
[email protected] url: www.fbk-kl.de
extends the process chain and increases manufacturing time and cost. Since burr generation in machining of ductile materials cannot be avoided totally, it is necessary to create burrs of predefined size and shape. Thereby, either tolerable or easy removable burrs can be generated. The first step towards this objective is the understanding of the burr formation mechanism. In the past decades, research of burr formation focused on cutting processes with defined tool geometry, which are commonly used in industrial manufacturing. For these cutting processes, the chip formation and the burr formation at the workpiece edges are directly linked. However, the burr formation in cutting processes with geometrically undefined cutting edges, such as grinding, is insufficiently investigated. The main reason therefore is a predominantly smaller burr size, which only interferes with increasing requirements on the precision of edge geometry. Furthermore, the relation between microscopic chip formation in grinding, due to multiple single grits, and macroscopic burr formation all along the workpiece edges is more complex compared to cutting processes with defined tool geometry.
2 Burr Formation in Grinding For the classification of grinding burrs usually the terms entrance burr, side burr and exit burr are applied according to the cutting direction (Fig. 1a). Independent from the feed direction, the entrance burr always occurs at the workpiece entrance edge, where the cutting direction of the single grits point into the workpiece. The exit burr similarly occurs at the exit edge, where the cutting direction of the grinding grits point out of the workpiece material. In pendulum grinding (Fig. 1b) the exit burr formation at the exit edge in an up-cut grinding pass is therefore alternated by an exit burr formation in a down-cut grinding pass. This also applies to the entrance and side burr formation. Previous research on burr formation in metal grinding was done by Gillespie [1], Kawamura and Yamakawa [2],
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_7, © Springer-Verlag Berlin Heidelberg 2010
63
64
H. Sudermann et al.
Fig. 1 Classification of grinding burrs
Yamakawa et al. [3], Hofman and Kvasnicka [4], Barth et al. [5], Warnecke et al. [6] and Aurich et al. [7–9]. Gillespie [1] analyzed the burr formation in pendulum grinding without coolant supply. He detected different macroscopic burrs at the workpiece edges and explained them according to the engagement of a single grain as follows: • At the side edges of a single grain a poisson burr is forming, caused by lateral forces. A superposition of numerous, randomly distributed single grain engagements is continuously pushing the side burr formation. • A rollover burr formation occurs at the exit edge. By the end of a single grain engagement the chip formation stops and the workpiece material is plastically bent around the exit edge. • An entrance burr can occur at the entrance edge, due to a plastically material deformation in opposite moving direction of the single grain. Kawamura and Yamakawa [2] carried out experimental work on a reciprocating table type surface grinder without coolant supply. They obtained the following results: • Among the analyzed entrance, side and exit burrs, the exit burr is by far the largest in size. Therefore, the subsequent investigation concentrated on the formation mechanism of the exit burr. • At the end of a grinding pass, a plastic deformation zone appears at the exit edge. By increasing the exit edge angle, the expansion of this plastic zone can be decreased. At the same time the exit burr size is reduced. • The exit burr formation is explained by a material deformation. A small area of the theoretically to be removed workpiece material is bent around the exit edge. Consecutive grinding passes cause an accumulation of the exit burr. Parallel to the experimental investigations, Yamakawa et al. [3] investigated the exit burr formation in grinding by a two-dimensional Finite Element Model. Since the material removal during grinding was not considered, the simulation results showed no real burr formation. However, the
simulation results confirmed an increasing plastic flow area in the vicinity of the exit edge. Further on, a tendency for heat concentration at the workpiece end could be observed as the exit edge angle becomes more acute. Hofman and Kvasnicka [4] also investigated the burr formation at the exit edge on a reciprocating table type surface grinder without coolant supply. Due to similar cutting conditions, they obtained similar results as Kawamura and Yamakawa [2]. Moreover, a higher microhardness in the burr root area compared to the core material was detected. Barth et al. [5] investigated the burr formation in surface grinding (up-cut) with coolant supply. Besides a conventional grinding wheel, a superabrasive grinding wheel was used for the experimental research. Contrary to previous studies, no entrance burr formation occurred. The observed side burr formation showed a strongly curled geometry, but was not analyzed in detail. Main focus of this research again was the exit burr formation, which is influenced by the exit edge geometry, the grinding wheel and the cutting conditions. According to the authors, burr formation in general can be explained by plastic deformation. Due to insufficient resistance against acting forces, the workpiece edge deforms. In addition to the described experimental analysis of Barth et al. [5], Warnecke et al. [6] developed a two-dimensional Finite Element Model to simulate the thermal and mechanical loads on the exit edge. Besides the identification of the heat concentration at the exit edge, the impact of the exit edge angle on the mechanical deformability of the exit edge was observed in the simulation. The exit burr formation in surface grinding (up-cut) was also investigated by Aurich et al. [7–9]. Experimental studies in wet grinding [7], applying conventional and superabrasive grinding wheels with different grain sizes and materials, showed a direct correlation between grinding forces, temperatures and burr height. Furthermore, an increasing heat concentration and burr size for more acute exit edge angles was detected. Later on, a three-dimensional Finite Element Model was developed to simulate the exit burr formation [8]. Due to the physical modeling of a rotating grinding wheel
Analytical Modeling and Experimental Investigation of Burr Formation in Grinding
and a moving workpiece, as well as the consideration of the material removal, the burr formation at the exit edge could be simulated qualitatively. Based on the results of the experimental investigation and the Finite Element Simulation, a descriptive model of exit burr formation in grinding was developed [9]. Similar to the burr formation in processes with defined cutting edges, the exit burr formation in up-cut grinding was subdivided into the five steps: continuous grinding, pre-initiation, burr initiation, burr development and final burr formation. The approach to the analysis of burr formation in grinding shows parallels to the investigation of burr formation in machining processes with geometrically defined cutting edges. To obtain a basic understanding of burr formation, a surface grinding process with simple kinematics is applied. In principle, the same kind of burrs as in orthogonal cutting are found at the workpiece edges (entrance burr, side burr, exit burr). According to previous research results, the largest burr occurs at the exit edge. Therefore, the analysis of the exit burr formation mechanism and the determination of the exit burr geometry for varying input parameters were the main focus of previous research. A comprehensive investigation of the side burr formation and the occasionally occurring entrance burr formation was not done before.
65
3 Analytical Modeling of Burr Formation in Up-Cut Grinding To illustrate the burr formation at different workpiece edges, a geometric examination of the contact situation between grinding wheel and workpiece is done. For surface grinding of a so called geometrically long workpiece, whose length exceeds the geometrical wheel-workpiece contact length (lg ) at the full depth of cut (ae ), the three sequences cut in, steady state and cut out can be considered [10]. Figure 2 shows an idealized illustration of these sequences for a cuboidal workpiece • without consideration of the cutting direction, • neglecting elastic and plastic workpiece deformation owing to acting forces and/or varying temperatures, • without consideration of the dislocation between grinding wheel and workpiece as a consequence of the acting forces and • assuming an ideal geometric penetration of the workpiece by the grinding wheel, in combination with a complete material removal. The cut in starts with the first contact between grinding wheel and workpiece at the upper edge of the workpiece
Fig. 2 Sequences of a surface grinding pass for a geometrically long workpiece
66
contour (sequence Ia). During cut in, the wheel-workpiece contact length increases from zero to its geometrical steady state value lg as the depth of wheel engagement with the workpiece in feed direction vf increases to the specified wheel depth of cut ae (sequence Ic/IIa). Steady state grinding (sequence II) is characterized by a constant geometric contact length and area moving in feed direction and leaving a ground workpiece surface. Cut out (sequence III) occurs during disengagement of the grinding wheel at the end of the grinding pass. Thereby, the wheel-workpiece contact length decreases from its steady state value lg back to zero. A close examination of the contact geometry between grinding wheel and workpiece shows the relevant sequences for the burr formation at different workpiece edges. Figure 3 illustrates the analytical model of burr formation in up-cut grinding of a geometrically long workpiece. The symbols for simple designation of the burrs in Fig. 3 (M-B, M-S and M-F) go back to the machining burr classification system of Nakayama and Arai [11]. Their combined classification system considers the directly concerned cutting edge (M – major cutting edge; C – Corner or minor cutting edge) as well as the mode and direction of burr formation (backward flow – B; sideward flow – S; forward flow – F; leaning to feed direction – L). According to the geometrical contact situation between grinding wheel and workpiece, the burr formation has to occur at the edges of the temporary contact area. Therefore, the formation of an entrance burr (M-B) starts immediately
Fig. 3 Analytical model of burr formation in up-cut grinding
H. Sudermann et al.
after the first contact between grinding wheel and workpiece (sequence Ia) and ends with the beginning of the steady state sequence (Ic/IIa). Simultaneously to the beginning of the entrance burr formation also the side burr formation starts. Since side burrs are generated at the side edges of the temporary contact area, the side burr formation is an ongoing process during the complete grinding pass. The exit burr formation starts with the beginning of the cut out sequence (IIIa) and ends with the last contact between grinding wheel and workpiece (IIIc).
4 Experimental Investigation of Burr Formation in Up-Cut Grinding A study of the burr formation at the edges of the temporary contact area is not possible with conventional grinding experiments. Therefore, so called “grinding stops” are realized for the verification of the analytical model of burr formation. For the preparation of each grinding stop two up-cut grinding passes are carried out. At first a conventional grinding pass with a depth of cut ae = 0.1 mm was realized to achieve a coplanar workpiece surface. Afterwards, the second grinding pass with a depth of cut ae = 1.0 mm was specifically interrupted by using the consecutive control commands: stop of the feed motion and operating the grinding spindle vertically in +Y-direction. The supply of grinding oil and the
Analytical Modeling and Experimental Investigation of Burr Formation in Grinding
67
Fig. 4 Grinding stops for up-cut grinding
spindle rotation is stopped when reaching a distance of 50 mm between grinding wheel and workpiece. Grinding stops are carried out every 5 mm distance in feed direction (Fig. 4). As a consequence of the chosen cutting parameters and the workpiece geometry, ten grinding stops resulted. Furthermore, one workpiece is completely ground. The optical analysis of the adhering burrs enables the verification of the analytical model of burr formation in up-cut grinding. In detail the following results can be obtained: • The exit burr formation starts during the disengagement of the grinding wheel (grinding stop at position 35 in Fig. 4) and continues until the end of the grinding pass (completely ground workpiece). • In addition to the completely developed side burrs (M-S) in the area of the completely ground side edges of the workpieces, a burr formation at the temporary side edges of the contact area can be observed. • An entrance bur formation is not detectable for any of the workpieces.
4.1 Exit Burr Formation To investigate the exit burr formation, the grinding stops during the cut out sequence are analyzed. In Fig. 5a the side views of the relevant workpieces and the metallographic sections of the individual exit edges are illustrated. The metallographic sections show the continuous growth of the exit burr. Thereby, the small exit burr at position 30 is not a result of the grinding stop. It is in fact the remaining exit burr of the first grinding pass with a depth of cut ae = 0.1 mm. Due to the accumulation of the exit burr, this small exit burr from the first grinding pass is detectable at the top of all subsequent exit burrs. A diagram of the burr thickness profiles is displayed in Fig. 5b. It exemplifies the continuous burr growth at the burr root. The differences in burr length among the grinding stop positions bl result from a continuously deformation of the theoretically to be removed workpiece material at the exit edge (Fig. 5c). Starting with the disengagement of
68
H. Sudermann et al.
Fig. 5 Exit burr formation in up-cut grinding
the grinding wheel (theoretical grinding stop position 32 in Fig. 5c), the burr length is increasing proportional to the depth of cut ae . According to the workpiece exit angle of 60◦ , the proportionality factor is 1/(sin 60◦ ) or about 1.15.
4.2 Side Burr Formation For the investigation of the side burr formation a grinding stop from the steady state grinding process is analyzed (position 25 in Fig. 4). Therefore, the temporary contact area along the geometrical contact length lg = 20 mm is cut in regular distances of five millimeter. Figure 6a illustrates the cutting positions in the temporary contact area (0, 5, 10, 15 and 20 mm) and the corresponding metallographic sections of the left and right workpiece side burr. The metallographic sections show the growth of the side burrs along the temporary
contact area. Compared to the exit burr geometry, the side burrs are curled much stronger. The microscopic side burrs for the first cutting position (0 mm) are again the result of the first grinding pass with a depth of cut ae = 0.1 mm. Due to the side burr accumulation, these microscopic burrs are also detectable at the top of all subsequent side burrs. Because of the rectangular side edge angle of 90◦ , the differences in side burr length bl increase simultaneously with the depth of cut ae (Fig. 6b, c). Thus, the side burr formation is caused by a continuously deformation of the theoretically to be removed workpiece material at both sides of the temporary contact area. The metallographic sections of the cutting positions 5–20 in Fig. 6a show a thermally induced change of the microstructure near the surface of the temporary contact zone. Since the temperature distribution in the temporary contact zone and the heat flux into the workpiece surface
Analytical Modeling and Experimental Investigation of Burr Formation in Grinding
69
Fig. 6 Side burr formation at the side edges of the temporary contact area
during a grinding pass are changing, a thermal impact on the side burr formation must be detectable. Therefore, the side burr formation along the completely ground workpiece is analyzed.
4.3 Thermal Impact on Side Burr Formation During a Grinding Pass The temperature during initial wheel-workpiece engagement (cut in) in up-cut grinding rises rapidly. Main reason therefore is the insufficient lubrication of the temporary contact zone, due to the absence of a lubricating oil wedge in the gap between grinding wheel and the already ground workpiece surface. If the workpiece is sufficiently long, a quasi-steady
state temperature distribution is reached during the steady state grinding process. During the final wheel-workpiece disengagement (cut out), as workpiece material is suddenly unavailable to dissipate heat by conduction, the temperature may exceed the quasi-steady value [10]. To detect the varying thermal impact on the surface of the completely ground workpiece (post-process), a visual examination of the black workpiece discoloration at the upper edge of the workpiece side view in Fig. 7a is done. This black discoloration, which in literature is also described as temper colors [12], is an oxide layer. Higher temperatures promote the oxidation process near the workpiece surface and increase the broadness of the oxide layer. Therefore, the visual examination of the oxide layer broadness allows a reverse conclusion on the workpiece temperatures during the complete grinding pass.
70
H. Sudermann et al.
Fig. 7 Thermal impact on side burr formation
The oxide layer formation on the side view of the completely ground workpiece in Fig. 7a confirms the prior explained temperature profile of an up-cut grinding process. After a small area without oxide layer formation a rapid enlargement up to the maximum broadness can be observed, due to a considerable temperature increase. Afterwards, a decreasing temperature is leading to a decreasing broadness of the oxide layer. Right at the end of the workpiece a small temperature increase is again broadening the oxide layer. The absence of an area with constant oxide layer broadness is an indication of the transient temperatures during the grinding pass. Due to the short workpiece length, a quasi-steady state temperature distribution during the steady state grinding sequence was not reached.
For the analysis of the thermal impact on side burr formation, the completely ground workpiece was cut in 5 mm distances. The metallographic sections of the side burrs are shown in Fig. 7a. In accordance with the bigger broadness of the oxide layer, a thermal surface layer impact on the metallographic sections can be detected for the cutting positions 5, 10 and 15 mm. Exemplary microhardness measurements in the surface layer confirm a strongly increased hardness of about 63 HRc, due to a martensitic hardening. Compared to the side burrs of the first three cutting positions, the burrs on the following cutting positions appear increasingly thinner. This optical impression is confirmed by the measured burr thickness profiles, illustrated in Fig. 7b. Thus, the higher temperatures in the beginning of the grinding pass support the
Analytical Modeling and Experimental Investigation of Burr Formation in Grinding
material deformation and therefore the side burr formation. In Fig. 7c the measured values of burr length, burr height and burr thickness at root are illustrated. Additionally, the measured values of the so called “minimum burr thickness near the root”, which affect a deburring process, are displayed. Particularly, the burr thickness at root and also the “minimum burr thickness near the root” are strongly affected by the higher temperatures for the first three cutting positions. The burr length and burr height on the other hand remain thermally unaffected.
5 Conclusion and Outlook In the presented work, an analytical model of burr formation in up-cut grinding was developed and experimentally verified. Moreover, parameters with influence on the geometry of grinding burrs were identified. The side and exit burr formation in up-cut grinding are the result of a plastic deformation process at the edges of the temporary contact zone. Due to the insufficient resistance against acting forces, a small area of the theoretically to be removed workpiece material at the side edges and the exit edge is continuously deformed. These experimental results verify the proposed analytical model of burr formation. While the side burr formation is an ongoing process during the complete grinding pass, the exit burr is only formed during the disengagement of the grinding wheel. Therefore, the side burr formation is affected by a changing thermal impact during the grinding pass. Higher temperatures, which occur at the beginning of the up-cut grinding process, cause a bigger burr thickness. The burr length however remains thermally unaffected. While the exit burr formation occurs by a successively deformation of the uncut workpiece material in cutting direction, the side burr formation requires a stronger material deformation in orthogonal direction to the grinding forces. Therefore, the side burrs are curled stronger. The observed side burr formation shows a good correlation to the side burr formation in orthogonal cutting, as detected by Nakayama and Arai [11]. For the up-cut grinding process as well as for an orthogonal cutting process with a negative rake angle a side flow of workpiece material causes the sideward burr formation. Therefore, the side burr formation accumulates for consecutive cuts. During this experimental research, an entrance burr formation did not occur. The main reason therefore is probably
71
the high thermal convection at the entrance edge, which is caused by a big volume flow of grinding oil directly supplied at the entrance edge. Due to the cold entrance edge, a backward flow of workpiece material is suppressed. For that reason, the entrance burr formation in dry conventional grinding [1, 2] as well as the absence of entrance burr formation in wet grinding [5, 6] can be explained. The presented results enhance the understanding of burr formation in grinding. According to the analysis of burr formation in up-cut grinding, the burr formation in down-cut grinding has to be investigated in future. Based on these combined results, burr control strategies for grinding processes have to be developed.
References 1. Gillespie, L.K., 1975, The Burrs Produced by Grinding, Report BDX-613-1572, Kansas City, MO: Bendix Corporation (available from NTIS). 2. Kawamura, S., Yamakawa, J., 1989, Formation and Growing up Process of Grinding Burrs, Bulletin of the Japan Society of Precision Engineering, 23(3): 194–199. 3. Yamakawa, J., Kawamura, S., Okuyama, S., 1989, Effect of Plastic Zone near Workpiece Ends on Grinding Burrs, Bulletin of the Japan Society of Precision Engineering, 23(3): 224–229. 4. Hofmann, P., Kvasnicka, I., 1999, A Study of Grinding Burrs Formation, Abrasives Magazine, October/November: 29–33. 5. Barth, C., Dollmeier, R., Warnecke, G., 2001, Burr Formation in Grinding of Hardened Steel with Conventional and Superabrasive Wheels, Transactions of NAMRI/SME, 29(1): 273–278. 6. Warnecke, G., Dollmeier, R., Barth, C., 2002, Gratbildung beim Schleifen von gehärtetem Stahl mit konventionellen und hochharten Schleifscheiben, Industrie-Diamanten-Rundschau, 36(3): 202–207. 7. Aurich, J.C., Sudermann, H., Braun, O., 2004, Experimental Investigation of Burr Formation in Surface Grinding of Tool Steel, Proceedings of the 7th International Conference on Deburring and Surface Finishing, Berkeley, CA, pp. 29–37. 8. Aurich, J.C., Sudermann, H., Bil, H., 2006, 3D Finite Element Modeling of Burr Formation in Grinding, Proceedings 9th CIRP International Workshop on Modeling of Machining Operations, Bled, Slovenia. 9. Aurich, J.C., Sudermann, H., Bil, H., 2005, Characterisation of Burr Formation in Grinding and Prospects for Modelling, Annals of the CIRP, 54(1): 313–316. 10. Malkin, S., Guo, C., 2007, Thermal Analysis of Grinding, Annals of the CIRP, 56(2): 760–782. 11. Nakayama, K., Arai, M., 1987, Burr Formation in Metal Cutting, Annals of the CIRP, 36(1): 33–36. 12. Rowe, W.B., Black, C.E., Mills, B., Qi, H.S., Morgan, M.N., 1995, Experimental Investigation of Heat Transfer in Grinding, Annals of the CIRP, 44(1): 329–332.
Developing a Process Model for Abrasive Flow Machining E. Uhlmann, V. Mihotovic, H. Szulczynski, and M. Kretzschmar
Abstract Abrasive flow machining (AFM) is a unique machining method used to achieve high surface quality on inner, difficult-to-access contours and on outside edges. Using AFM, it is possible to deburr complex shaped intersecting holes and to realize pre-defined edge rounding on any brittle or hard material. Moreover it is easy to integrate into an automated manufacturing environment. Reproducibility of results saves various time and cost consuming manual operations in industrial applications. However for an implementation in new applications costly and time-consuming preliminary investigations are required that have to be carried out by trained staff. Therefore the aim of recent research activities is to identify the fundamentals of the process, the functional correlations between setting parameters and work results. This paper presents an approach using a numerical simulation to develop a process model. Furthermore a model is introduced that describes the fundamental principles of the deburring process using AFM. Keywords Abrasive flow machining · Deburring · Surface finishing · Numerical simulation
1 Introduction Abrasive flow machining is an innovative finishing and polishing operation with a gentle material removal mechanism. In contrast to other machining methods for deburring and polishing, it is possible to machine difficult-to-access cavities, inner contours and undercuts in a reproducible
E. Uhlmann, V. Mihotovic (), M. Kretzschmar Institute for Machine Tools and Factory Management (IWF), Technische Universität Berlin, Office PTZ 1, Pascalstr. 8-9, 10587 Berlin, Germany e-mail:
[email protected] url: www.iwf.tu-berlin.de H. Szulczynski Robert Bosch GmbH, Diesel Systems – Manufacturing Department 7, Engineering Nozzle BaP/MOE7, 96045 Bamberg, Germany
manner. Typical components that could be machined by AFM are extrusion and compression molding dies as well as crimping and stamping tools. Use of AFM on these tools showed that within 2 minutes processing time an improvement of the surface roughness from Ra = 2 μm to Ra = 0.2 μm could be achieved. The medium applied during AFM is a fluid consisting of a polymer which carries silicon carbide or super abrasive grains. With a specified pressure and temperature, this fluid flows in alternating directions along the contours of the work piece resulting in an abrasive effect. Figure 1 depicts the process principle during machining. Unlike in conventional grinding process the path curves of the active grains in AFM depend on the existing contact conditions, the geometrical constraints of every workpiece and the actual state of the abrasive media. By using modern simulation techniques it is possible to enhance the knowledge about the fundamental principles of the flow process and to anticipate the complex behavior of the visco-elastic grinding medium. For this purpose three topics will be presented: • Deburring mechanisms using AFM, • a numerical model for the flow process and • an approach to develop a process model for AFM.
2 Experimental Setup The experiments were carried out on a Delta Towers MF100 type abrasive flow machine by Micro Technica Technologies, Kornwestheim, Germany. The selected workpieces are displayed in Table 1. The investigations of deburring behavior in AFM were conducted on geometry 1. The chosen geometry allows an easy access for observing the effects obtained by applying AFM. The workpieces were prepared by milling the offset without subsequent deburring. The components were made of steal C45. In addition a second cylindrical geometry with an inner diameter of 32 mm was used to observe the flow velocity
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_8, © Springer-Verlag Berlin Heidelberg 2010
73
74
E. Uhlmann et al.
Fig. 1 Process principle during abrasive flow machining
of the visco-elastic grinding medium. For this purpose the workpiece was made of an acrylic polymer. The intention of this experimental approach was to provide data for evaluation of a numerical simulation. A flow profile as well as a numerical model have been developed using the CFD-application Fluent 6.3 of the company ANSYS, Southpointe, USA. A grinding medium with the specification MF10-36S-200 has been used. The first term of the nomenclature indicates the viscosity of the polymeric carrier material, which is a specification given by the manufacturer. The second term specifies the grain size of the used abrasive; in this case 36S for silicon carbide with a grain size diameter of 525 μm according to FEPA F36 or to DIN 69101. The last term stands for a weight ratio between carrier medium and abrasive grains of 1–2.
3 Deburring Mechanism Using AFM Table 1 shows the geometry 1. The prepared burr displayed various characteristics. In some cases burrs appeared with a loose connection to the base material. Loosely attached burrs could be assessed qualitatively by SEM-images. Burrs firmly attached to base material were assessed quantitatively by a mechanic scanner of the company Mitutoyo, type Digimatic, Japan, displayed in Fig. 2. The burr height is distributed statistically along the workpiece edge. The burr was classified before and after AFM according to the model of Schäfer [1].
Table 1 Workpiece geometries for technological investigations Geometry 1
Geometry 2
Figure 3 summarizes the results of the investigations as well as the underlying mechanisms according to Szulczynski [2]. The diagram shows the chronological sequence of the deburring and edge rounding process. In addition the dependency of the deburring behavior on the first impulse is illustrated. The left column shows the burr behavior, if the first flow impulse is directed downwards. For a vertical abrasive flow upwards the behavior is given in the right column. The SEM-images were made after 0, 1/2, 11/2, 41/2, 9 operating cycles. During the first impulse a steady flow occurs without return flux. The arrows depict the flow direction during the first impulse. The described mechanisms are based on intermediate results observed throughout the process. Experiments showed that mostly a mixture of both processes occurs.
Developing a Process Model for Abrasive Flow Machining
75
Fig. 2 Burr formation before abrasive flow machining
The primary mechanism characterizes the first phase of the machining immediately at the beginning of the first piston lift. The burr offers resistance to the abrasive flow. The grinding medium bends the burr in flow direction at the thinnest point. In the left column the burr is pressed towards the surface with the downward flow. When the impulse is in an upwards direction, the burr likewise moves with the impulse
to decrease the resistance to the abrasive flow. The thinnest point begins to crack due to the high mechanical stress. The secondary mechanism describes the processes at the end of the first piston lift, directly after the reversal of the flow direction. In the left column the burr is bent to the surface. In this case grinding medium can no longer reach the underside of the burr. Subsequently the flow gets deflected at the burr root, thus causing an increased material removal rate at this location. The burr in the right column, already under high mechanical stress, is bent along the flow direction. The plastic deformation ultimately leads to the burr breaking off. The burr gets carried away by the flow and pulverized within the grinding medium. The residual fracture surface is ragged, again causing an increased material removal rate. During the tertiary mechanism a continuous edge rounding takes place. In the left column the burr breaks off at the burr root. This fracture surface gets polished by abrasive flow. As the machining time increases, the rounding of the edge is practically completed. The flow resistance and thus the material removal rate decreases. Similarly, a continuous edge rounding of the burr root takes place in the right column as well. In both cases the edge rounding process is completed after 9 operating cycles. In spite of the different processes and the chronological sequence of events the resulting edge rounding is the same. Furthermore, workpieces show traces of grinding marks in vertical direction caused by the abrasive grains.
4 Developing a Numerical Model
Fig. 3 Deburring mechanisms during AFM according to Szulczynski [2]
The optimization of the manufacturing process can be accelerated significantly using modern computer and simulation technologies [3]. In order to determine optimal production parameters, nowadays cost and time intensive experimental research is necessary. By improving simulation technologies, it will be possible to limit experimental setup to a minimum, improve process comprehension and reduce development costs. Below an approach for modeling flow processes is presented.
76
E. Uhlmann et al.
The precise mechanisms and the kinematical principles of AFM are still not extensively explored, due to the fact that the medium is non-Newtonian and the behavior can not be directly observed [4]. The non-Newtonian fluid properties have an extraordinary impact on the material removal rate. Therefore relevant rheological parameters must be included in a numerical model of the grinding medium. In the field of fluid mechanics extensive investigations were conducted for a cylindrical geometry (Fig. 4). The workpiece was made of an acrylic polymer. Under variation of temperature and working pressure the abrasive grain movements along the surface were observed by using a high-speed-camera. The mean velocity of a cutting grain could be determined by analyzing the distance covered by particles as observed in image sequences. The numerical model was validated by the experimental results. The rheological properties of a grinding medium were measured with a rotational rheometer of the company HAAKE GmbH, type RT 20, Germany. The measurements were used to determine the relation between shear rate γ˙ , shear stress τ and viscosity η, presented in Eq. (1). η=
τ γ˙
(1)
The visco-elastic behavior of the grinding medium could be described by the power of law according to Ostwald and De Waele [5]. η = K · |γ˙ |n−1
(2)
The model of Ostwald and De Waele is a simple approach for non-linear flow behavior. Two material parameters are required for this approach: • Dimensionless viscosity exponent n and • the consistency parameter K.
Fig. 5 Regression of the viscosity curve according to Ostwald/De Waele
The flow curves resulting from the rheometrical measurements are then analyzed mathematically using a software application. The parameters of the chosen regression are obtained as a result. Figure 5 shows an example of the regression analysis. Although the deviation between the measured values and the regression values is relatively high at a low shear rate of γ˙ = 0.2s−1 , the full flow profile evolves at significantly higher shear rates. Hence the regression meets the requirements for modeling non-Newtonian fluids like an abrasive medium.
4.1 Comparison of the Mean Cutting Velocity of Abrasive Grains
Fig. 4 Velocity of a cutting grain along the observed surface
The comparison between Ostwald/ De Waele and the experimental results are displayed in Fig. 6 exemplarily for the temperature ϑ = 30◦ C. The differences in velocity profiles, especially in the wall vicinity, are very slight. The comparison also shows that the correlation between the displayed curves is high. The deviation between the
Developing a Process Model for Abrasive Flow Machining
Fig. 6 Mean cutting velocity at different working pressure
77
model to evaluate the numerical simulation of the abrasive flow behavior. However, this analytical approach is not valid for rectangular cross-sections. In order to be able to allow the modeling of arbitrary, not just simple-shaped geometries, complex flow equations must be solved. As described earlier, the shear thinning properties of the abrasive fluid can be determined using rheometrical devices. The underlying process model is given through the flow-chart in Fig. 7. Firstly, the complex characteristics of the grinding medium are investigated on circular cross-sections. Subsequently, the numerical model is applied in a complex case to create a connection between the local flow ratio and machining results, such as the surface finish. In order to apply the process model on any arbitrary geometry, relevant correlations are stored in a technological database. In this matter a CFD-simulation, which takes most common constraints like working pressure, temperature and visco-elastic behavior of the grinding medium in account is capable of determining local velocities on complex shaped geometries. Combined with the technological data that are stored in a database, it is possible to anticipate work results. The functional capability of the process model remains to be proven by a selection of complex-shaped work pieces.
numerical results and the measured results is between 6 and 15% of the absolute values.
6 Conclusion and Outlook 5 Approach for a Process Model The velocity of the abrasive grains in the contact zone can be calculated for circular cross-sections with an analytical
Fig. 7 Flow-chart showing the development of a process model
Three different approaches have been provided. Mechanisms for deburring metallic materials have been investigated extensively. The results are summarized by Szulczyn-
78
ski in a diagram about the deburring process using AFM. A second approach shows the development of a numerical model for the flow behavior of visco-elastic grinding media. In this matter enhancement of knowledge about the fundamental flow principles in AFM can be achieved by using numerical simulation techniques. For the simulation of a non-Newtonian fluid rheological parameters for the grinding medium have to be integrated in the numerical model. The quality of the numerical results could be evaluated by measured results. The deviation between the observed values was low. The last approach shows an overview of a process model that is capable of anticipating work results like surface quality and edge rounding on complex-shaped geometries by using a numerical simulation. So far, results of extensive technological investigations have been made available in a database. However the development of the functional capability is still in progress. Acknowledgements The technological investigations regarding the Abrasive Flow Machining process on metallic materials were carried out by Dr. Szulczynski during his scientific activities at the Institute for Machine Tools and Factory Management. All authors acknowledge the financial support provided by the German Research Foundation.
E. Uhlmann et al.
References 1. Schäfer, F., 1997, Entgraten – Theorie, Verfahren, Anlagen, Krauskopf-Verlag, Mainz, Germany 2. Szulczynski, H., 2007, Verfahrensgrundlagen und Technologie des Hubschleifens mit viskosen Medien, Dissertation, TU Berlin, Germany 3. Klocke, F.; Beck, T., 2002, Examples of FEM application in manufacturing technology, Journal of Materials Processing Technology, 120 (1–3), 450–457 4. Uhlmann, E.; Mihotovic, V., 2006, Precise Finishing of Ceramic Materials with Abrasive Flow Machining. CIRP – 56th General Assembly, STC “G”, Kobe, Japan. 5. Carreau, P.J.; De Klee, D.C.R.; Chhabra, R.P., 1997, Rheology of Polymeric Systems, Carl Hanser Verlag, Munich, Germany
Modeling and Simulation of Burr Formation: State-of-the-Art and Future Trends J. Leopold and R. Wohlgemuth
Abstract Modeling and Simulation of Burr Formation is one of the main topics in contemporary basic scientific investigations. Using state-of-the-art numerical methods, the process of burr-development can be investigated with 2Dsimulations and increasingly for 3D-applications. The aim of the numerical analysis (e.g. FE simulation) of burr formation is burr minimization. Until the present time, numerical simulations of burr formation, especially for 3D-applications – have been very time consuming. Hence, in contrast to fundamental research investigations into the material removal process during burr formation, more practical methods of predictive modeling are need for industrial purpose. In view of Virtual Manufacturing, the response time must be dramatically faster than usual in numerically calculations. The most relevant methods currently applied to burr formation modeling and simulation are accurately reported and discussed in this paper. Keywords Modeling · Simulation · Burr formation · Predictive modeling · Numerical methods
1 Introduction Today, the focus is on how to maximize utilization of machining centres in order to expedite returns on these higher technology machine tools, which are also more expensive. Due to the rapid growth and competitiveness of the automobile industry, reduction in total machining time has become a basic necessity. In precision manufacturing, e.g. the automobile industry, burr minimization and deburring is key challenge. J. Leopold () Am Bahrehang 95, 09114 Chemnitz, Germany e-mail:
[email protected] url: www.tbz-pariv.de R. Wohlgemuth TBZ-PARIV GmbH, Bernsdorfer Strasse 210-212, 09126 Chemnitz, Germany
It is necessary to push the understanding of the burr to something much greater than just a nuisance to be removed or ignored. In [1] the investigation was focused on three areas of study (a) development of experimental databases on burr formation, (b) development of analysis tools (including FEM) for understanding of burr formation, especially in relation to tool geometry, and (c) development of software tools linked to CAD for process planning (e.g. tool path and process parameters), as well as demonstrating how the need of clear standards for edge quality can be beneficial to precision manufacturing. Edge finishing and deburring operations are often neglected in the design and process planning of precision parts. These finishing operations are often the source of many dimensional discrepancies and they usually occur in the final stages of manufacturing, after the completion of a series of processes which add value to a precision part. When the deburring of a precision part is not considered until the final stages of manufacturing, the potential loss due to any failure in the selection, planning or execution of the edge-finishing process is great. The cost of deburring these components may contribute as much as 30% to the cost of finished parts [2]. The selection of capable deburring and finishing processes for precision components is highly dependent upon knowledge of burr proper-ties. Burr size, shape and location, as well as the allowable surface finish, are the primary factors in the selection of a deburring process. Burr properties are influenced by part design and process planning decisions. To classify whether a particular burr property is influenced primarily by the design stage or the manufacturing stage requires burr formation data, burr formation models and burr formation mechanism identification. The results of such studies and data on burr formation form the basis of analytical and numerical tools, incorporated in software with CAD data input, for engineers at several levels from design to process planning to shop floor trouble-shooting, for effective burr prevention or minimization. For virtual manufacturing or virtual enterprises, theoretical investigations of modeling and simulation of burr formation and burr minimization are more highly utilized. There are two main trends – the first deals with mechanical- and material sciences based
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_9, © Springer-Verlag Berlin Heidelberg 2010
79
80
J. Leopold and R. Wohlgemuth
Fig. 1 Manufacturing methods and burr sizes: macro- and micro- scale modeling
investigations of burr-formation, and the second is focused on fast response to burr minimization during the manufacturing process. Due to the complexity of the burr formation process, typically two-dimensional investigations are more frequently being replaced by three-dimensional investigations. This paper will cover the state-of-the-art in burr modeling and simulation, from typically macroscopic processes down to the small scale details of the burr formation process at the material micro scale (Fig. 1).
Fig. 2 Burr types [3]
2 Predictive Models 2.1 “Simplified” Mechanical Models From the mechanical point of view, burrs can be divided into sub-groups (Fig. 2). From the industrial point of view, petal burrs (Fig. 3) are the most dangerous material overhang. Using the well-
Modeling and Simulation of Burr Formation: State-of-the-Art and Future Trends
81
Fig. 3 Petal burr at the drilling tool exit area (F – Force; l – overhang length)
known continuum bending theory, the stability of such types of burrs including the type of material used can be calculated [4]. bf =
4
2 3 · kc1.1 · fz2−2mc · (rβ )2 · (Kvc ) · (Kγ )2 Rm · ET · KT · Kϕ
(1)
where rβ is the cutting edge radius, fz is feed per teeth; Rm , ET and Kc1.1 are material constants and other parameters are determined from machining experiments (Kvc , Kγ , KT , Kφ).
2.2 Slip-Line Models Slip-line models have been well established in machining simulations since 1945. The most important steps forward in order to use this simulation technique for burr
Fig. 4 Slip-line and first slip-line model [5]
formation are given in the next few models (Figs. 4 and 5). Experiments have shown that plastic strain lines and breakout curves are similar to arcs of a circle. If the material is rigid-plastic, it can be accepted – that the direction of these lines is in agreement with the direction of flow lines. The arc of flow line AB in Fig. 6 determines the lower limit of plastic deformation in the first moment of burr formation. According to the boundary conditions for the stresses, the initial arc of flow line AB crosses the (cutting) face and the exit surface BC at the angle π2 + γ and π4 respectively. In this equation γ the angle of plastic friction, determined from the equation: γ =
π 1 τc − arccos 4 2 τβ
(2)
82
J. Leopold and R. Wohlgemuth
Fig. 5 Extended slip-line model and application for chip-curl determination [5]
If the arc of flow line AB is changed on the standard initial shear plane, that initial negative angle of shear can be obtained from geometric ratio as: β0 =
1 π − γ + γ − , 2 4
(4)
The length of shear plane AB can be calculated by the equation: AB = R 2(1 − cos ).
(5)
If during the burr formation process breakout doesn’t occur, the maximum pivoting angle of the burr from the tool exit surface can be determined from the proportion: max =
π + γ + γ 4
(6)
The dimensions of the formed burr are calculated according to the following formulae: Fig. 6 Slip-line model of burr formation for the tool exit in orthogonal cutting [5]
bf = AB
sin β0 cos max
b0 = C(bf + ap cosmax ) sin max where τc is the shear stress in the plastic strain area on the tool face and τβ is the limit shear stress of hardened chip material. In the general case the angle between the exit surface BC and the perpendicular to the machining surface is the exit angel . From the condition of balance of element ABCDEM, the radius R of initial flow line AB may be obtained as: R=
N cos γ + F sin γ (3) τβ [(1 + 2)(1 − 2 cos ) − sin + 2 cos ]
where N and F are the axial force and friction force on the tool face, respectively, = (π/4) − γ − γ is angle of sector AOB and γ is the rake angle.
(7) (8)
where ap is the chip thickness; C is a constant (coefficient) describing the protrusion of the exit surface as a result of deformation.
3 Numerical Methods 3.1 General Overview In the last 50 years methods of numerical mechanics and numerical mathematics have been introduced in many fields of manufacturing. The following Fig. 7 summarizes the most important methods.
Modeling and Simulation of Burr Formation: State-of-the-Art and Future Trends
83
Fig. 7 Numerical methods – applied in chip- and burr-formation
In the following, some results of FE-calculations applied to burr formation will be described.
3.2 Influence of Friction Limit Friction modelling in metal cutting has been recognized as one of the most important and challenging tasks facing researchers engaged in modelling of machining operations. The process of orthogonal metal cutting of aluminium alloys used in the automobile industry is studied with the finite element method under plane strain conditions. A series of finite element simulations have been performed to model the friction along the tool–chip interface and the interior of the cutting tool. Friction coefficients ranging from 0.01 to 1.5 have been considered in the simulations. The results of these simulations are consistent with experimental observations in the literature. The distribution seemed to be a bifurcate when changing the friction coefficient from 0.01 to 0.2. Furthermore, the maximum temperature, contact length, burr-geometry, hydrostatic pressure and cutting force are found to depend strongly on the coefficient of friction.
Fig. 8 Influence of the friction coefficient on the chip-tool contact length
Fig. 9 Influence of the friction on the chip-geometry
There are three interesting results. First the chip-tool contact length is influenced by the friction coefficient; at the beginning of the chip formation process the contact length increases linearly, but at very high friction there is some nonlinearity in the behavior caused by stick-slip motion of the chip at the rake face of the tool (Fig. 8). In comparison with previous investigations, the chip geometry is strongly influenced by friction (Fig. 9). On the other hand, the tool/chip friction conditions influence the burr formation (Fig. 10).
Fig. 10 Influence of friction on the burr geometry
84
J. Leopold and R. Wohlgemuth
3.3 New Continuum-Mechanical Parameters Applied to Burr-Formation 3.3.1 The Principle of the Hydrostatic Bowl In [5, 6] it was noticed that at transition from a steady-state cutting process to burr formation a region of high negative hydrostatic pressure exists. That region is called a “hydrostatic bowl” (Fig. 11). Hydrostatic pressure is the negative value of mean stress σ m . At the same time it is the first invariant of the stress tensor. Hence high negative hydrostatic pressure means large mean stress p = −σm
(9)
According to previous investigations the hydrostatic stress ratio is said to be the most significant influencing variable for the occurrence of cracks in forming processes. That means the hydrostatic pressure improves the formability of a material whereas hydrostatic tension (p < 0) affects a lower forming work. In addition to this the influence of the hydrostatic stress ratio on the formability increases with increasing temperature. Until now it is unknown exactly what effect the hydrostatic bowl has on burr formation. Therefore, the location and the dimensions of the bowl have to be monitored more closely. Indications of the location of crack initiation and of the path of the crack may be given by the hydrostatic bowl. First of all, it is reasonable to analyze the development of the hydrostatic bowl using failure criteria. It is essential to choose a reasonable mean stress value as threshold for the hydrostatic pressure, if it is chosen too small, the area is very large and vice versa. As expected, the results show large stress peaks at the crack tip, which means the crack pushes the hydrostatic bowl
Fig. 12 Shape of the hydrostatic bowl during crack formation
forward. This is clarified in Fig. 12 for the example of the Oyane failure criterion at a cutting speed of 400 m/min. The threshold value is about 800 MPa and the tool distance to the workpiece edge is 0.4, 0.3 and 0.1 mm. The left picture shows the state before crack initiation, which is equal for all failure criteria. With ongoing crack formation the shape and dimensions of the hydrostatic bowl vary with different failure criteria.
3.4 Material Influence on Burr Formation The workpiece material has an important influence on burr formation. Ductility of the material is the most important material property. Beyer [4] clarifies in his formula for the burr affinity N that the elongation without reduction of the cross-sectional area, εgl-, and the tensile strength, Rm-, are the most important factors. Another formula for burr affinity N from Link [7] includes the strength Rp0,2 , the yield strength A and tensile stretch Z, but all include parameters dependent on the material ductility. NBeyer = εgl · Rm
(10)
NLink = (Rm − Rp0,2 )(A + Z)
(11)
With numerical calculations, the influence of heat-treated material on burr sizes can be investigated.
3.5 Concept of Ductile Damage According to Lemaitre and Sievert
Pressure threshold Smallest pressure value
Fig. 11 Hydrostatic bowl
Sievert et al. introduce a model for ductile damage under high-speed-cutting (HSC) load conditions [8]. They analyze the machining behavior of nickel-base alloy Inconel 718 via experimental and numerical tests. The verification of the simulations has been done by comparing the numerical results with experimental data.
Modeling and Simulation of Burr Formation: State-of-the-Art and Future Trends
The basic principle of these constitutive equations is formed by the Johnson-Cook material model, which has strain- and strain-rate-dependent hardening as well as thermal softening. To this basic model Sievert adds a term, that contains the degree D of ductile damage:
σV = A + Bε
n
ε˙ 1 + C ln ε˙ 0
1−
T − TR TM − TR
m (12)
for 0 ≤ D ≤ 1 This form of description goes back to Lemaitre [9], who assumes in his concept of effective stress, that with increasing damage the load capacity of each single element decreases. Hence he introduces an effective equivalent stress for each element, which varies, depending on damage: σeff,v =
σV 1−D
(13)
Thus the constitutive equation for the deformation behavior of a damaged material has the same form as that of an undamaged material. The equation solved for equivalent stress is: σV = f (ε,˙ε ,T) 1−D
(14)
The degree of damage D has a direct influence on the material behavior. It differs clearly from the failure criteria, described in Sect. 2, where element failure occurs abruptly. To emphasize this difference, it is identified with D instead of the terminology C used for the failure hypothesis. To increase the degree of damage, a simple, direct nonlinear approach was chosen with dependence on a standardized interior time s, the lifetime consumption: D = sk
ζ1 , if σm ≥ 0 ζ2 , if σm < 0
4 Conclusions and Outlook Theoretical investigations into the mechanism of burr formation are the foundation of efforts to avoid or minimize burrs. Predictive models are suitable for a fast response time and can be applied in virtual manufacturing environments. On the other hand, to understand the mechanism of burr formation and burr minimization, numerical models are widely used. There are some trends for future developments. First, models must be focused on real three dimensional objects. Burr minimization in crossing holes is a problem in the automotive industry as well as in the manufacturing of microparts. Next, although homogeneous material behavior is typically appropriate for macroscopically manufacturing, in burr formation the workpiece dimensions are sometimes reduced to the size of the material structure. Hence new material properties must be included in modeling and simulation and new algorithms and computational principles must be developed to reduce the calculation time dramatically. The advantages and disadvantages of other simulation methods like Boundary Element Method, Finite Difference Method, Grid Free Methods and Molecular Dynamic Simulations will be discussed in the presentation. Acknowledgements The authors thank all colleagues and students (G. Schmidt, K. Hoyer, A. Stoll, M. Dix, J. Regel, A. Poppitz) for a long time research co-operations.
References (16)
with the intensity ζ of the influence of the multiaxiality on the ductile damage, which can be ζ1 for medium compressive or ζ2 for medium tensile stress: ζ (σm ) =
whose dimension is described by the parameter a. It leads to a decrease of cutting force with increasing cutting speed [10].
(15)
The lifetime consumption is calculated as
σm σeff,v ε˙ a s˙ = exp ζ (σm ) ε˙ 1+ σV Wc ε˙ 1
85
(17)
The critical value Wc of the plastic shear work per unit volume specifies when the material is completely damaged at a shear stress state of σ m = 0. The exponent k of the non-linearity describes when the achieved plastic shear work affects into a rigidity fall. Above the critical strain rate ε˙ there is an additional velocity dependence of ductile damage
1. Dornfeld, D., Min, S., Kim, J., Hewson, J., Hsing Chu, Ch., Tyler, P., Field, P. and Askari, A., 1999–2000, Burr Prevention and Minimization doe the Aerospace Industry, Research Reports 1999– 2000, CODEF 2. Gillespie, L. K., 1979, Deburring Precision Miniature Parts. Prec. Engg, 1 (4), pp. 189–198, 3. Dornfeld, D., 2002, Burr Prevention and Minimization Activities in CODEF, Consortium on Deburring and Edge Finishing, Presentation 4. Beyer, H. M., 1991, “Gratentstehung – ein umformtechnischer Ansatz”, Zeitschriftenaussatz: wt Werkstatttechnik online Volume 91 (2001) Magazine 12, S. 765–772 5. Leopold, J., Schmidt, G., Hoyer, K., Freitag, A., 2005, Modeling and Simulation of Burr Formation – State-of-the-Art and Future Trends, 8th CIRP International Workshop on Modeling and Machining Operations, Germany, Chemnitz, Proceedings, pp. 73–83 6. Regel, J., Stoll, A., Leopold, J., 2007, Numerical Analysis of Crack Propagation during the Burr Formation Process of Metals, 10th
86 CIRP-International Workshop on Modeling of Machining Operations, Reggio Calabria, Italy 7. Link, R., 1992, Gratbildung und Strategien zur Gratreduzierung bei der Zerspanung mit bestimmter Schneide, Dissertation, RWTH Aachen 8. Sievert, R., Noack, H.-D., Hamann, A., Löwe, P., Singh, K. N., Künecke, G., Clos, R., Schreppel, U., Veit, P., Uhlmann, E., Zettier, R., 2003, Simulation der Spansegmentierung beim Hochgeschwindigkeitszerspanen unter Berücksichtigung duktiler Schädigung, Technische Mechanik, Band 23, Heft 2–4, pp. 216–233
J. Leopold and R. Wohlgemuth 9. Lemaitre, J., 1996, A Course on Damage Mechanics, Springer Verlag, Berlin 10. Sievert, R., Hamann, A., Noack, H.-D., Löwe, P., Künecke, G., Kiyak, Y., Steppa, D., Haftaoglu, C., 2005, Werkstoffmechanik einer Nickelbasislegierung beim Hochgeschwindigkeitsspanen – Werkstoffverhalten und Modellierung, Abschlussbericht zum Forschungsvorhaben im Rahmen des DFGSchwerpunktprogramms 1057, Bundesanstalt für Materialforschung und -prüfung, Berlin
Burr and Chip Formation Mechanisms
Interfacial Burr Formation in Drilling of Stacked Aerospace Materials S.N. Melkote, T.R. Newton, C. Hellstern, J.B. Morehouse, and S. Turner
Abstract Interfacial burr formation during through-hole drilling in stacked aluminum sheets is a common problem in aircraft assembly operations. Burrs formed at the interface of the sheets are removed through non-value added destacking and deburring operations that increase the overall assembly time and costs. This paper presents experimental work aimed at understanding the impact of drilling parameters including drill geometry, cutting conditions, clamping configuration and drill wear on interfacial burr formation. Specific conclusions regarding the influence of these parameters on burr sizes and drilling forces are presented. Keywords Drilling · Burr formation · Clamping · Aircraft assembly · Stacked sheet metal · Step drill
1 Introduction Drilling burrs are commonly formed when through-holes are created in sheet metal. In the drilling of stacks of sheet metal (“sheets”) to be fastened together, burrs form at the interface between the sheets, as illustrated in Fig. 1, requiring the layers to be de-stacked and de-burred before final assembly is completed and fasteners are installed. Since drilling of stacked metallic sheets is a frequent operation in aircraft structure assembly, minimizing or eliminating these burrs
S.N. Melkote (), T.R. Newton, C. Hellstern George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, 801 Ferst Drive, Atlanta, GA 30332, USA e-mail:
[email protected] url: http://pmrc.marc.gatech.edu J.B. Morehouse Manufacturing Research Center, Georgia Institute of Technology, 813 Ferst Drive, Atlanta, GA 30332, USA S. Turner Lockheed Martin Aeronautics Corporation, 86 South Cobb Drive, Marietta, GA 30063, USA
Fig. 1 Interfacial burr formation in drilling of stacked sheet metal layers
will prevent the need for non-value-added and costly destacking and deburring operations. Many experimental investigations have been performed to understand and minimize burr formation in drilling of a single layer of material [1–3] and in other processes [4]. Analytical models have been developed to predict exit burr sizes based on given drilling parameters [5–8], and burr formation mechanisms have also been analyzed with the use of finite element modeling [9–11]. The effect of both solid and bushing-type backup materials on hole exit burr has also been studied [9, 12–13]. However, relatively little has been reported on the topic of burr formation at the interface of two or more stacked sheets in hole drilling operations. The inter-layer gap and resulting burrs for drilling of stacked sheets of SS304L have been simulated using finite element methods [14–15], although no experimental validation was conducted, and the effects of drill geometry, feed and speed were neglected. Burr formation in drilling of graphite/bismaleimide-titanium alloy stacks [16] has been studied, although interfacial burr was not considered. Work on burr formation in drilling of composite-titanium stacks [17] has been reported, but it is unclear how applicable these findings would be to the case of other metal sheets. Limited findings relating to interfacial burr heights in drilling of titanium-titanium and titanium-aluminum stacks under varying clamping forces have been reported [18], but no analysis of varied drill geometries, feeds and speeds was done.
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_10, © Springer-Verlag Berlin Heidelberg 2010
89
90
S.N. Melkote et al.
While all of the aforementioned studies deal with burr formation, it is clear that reported knowledge of experimental work for the case of interfacial burr formation in stacked aluminum sheets under varying drilling and clamping conditions, in addition to knowledge of the impact of drill wear on interfacial burr formation, is lacking. This paper presents a summary of recent experimental work by the authors on interfacial burr formation in drilling of stacked sheet metal that addresses the aforementioned deficiencies. The first set of experiments was designed to determine the impact of several drilling parameters, including the helix angle, point angle, coating type, point type, feed, speed, and clamping distance on the resulting interfacial burr sizes produced in stacked aircraft-grade aluminum sheets, in the absence of drill wear. These experiments (referred to as “Drilling Parameter Experiments” throughout this paper) were conducted with unused drills in order to achieve this goal. The second set of experiments (referred to as “Drill Wear Experiments” throughout this paper) was performed to determine the effect of drill wear on interfacial burr formation for three different drill types and to determine which drill performed the best in terms of minimizing interfacial burrs over the useful life of the drill.
2 Experimental Work 2.1 Drilling Parameter Experiments Owing to the small quantity of work done in the area of interfacial burr formation in drilling of stacked aluminum sheets, factors that have been shown to be significant for burr formation in general were selected for study. The factors and their levels can be seen in Table 1. Helix angle, point angle, coating type and point type comprise the various elements of drill geometry examined. Helix and point angles have been shown to have a significant effect on exit burr thickness in drilling of aluminum [1]. Drill coatings have also demonstrated an influence on burr heights [19–20]. Drill point type is known to have an impact on burr heights in aluminum [21]. For these experiments, the drill Table 1 Experimental factors and levels Factor Level 0 Helix angle Point angle Coating Point type Feed (/rev) Speed (RPM) Clamp type Clamp distance Frame material
30◦ 118◦ Black oxide Standard 0.10 mm (0.004 ) 3000 2 Hand clamps 15 mm (0.59 ) 2024-T3
Dimensions: D 1 = 4.91 mm D 2 = .4.41 mm L = 1.52 mm Ө 1 = 118/ 135° Ө 2 = 40° Fig. 2 Step drill geometry used [22]
point type factor consisted of three levels: standard-point, split-point and step drill. The step drill, displayed in Fig. 2, has been shown to minimize burr heights [22]. Note that in the current experiments, the step drill could have either a standard-point or a split-point. All drills were made of high speed steel (HSS) and were #10 in size (4.91 mm diameter). In order to minimize the effects of drill wear, each run of the experiment was conducted with a new drill. The influence of drill wear is analyzed in the second set of experiments reported later in the paper. The feeds and speeds selected were based on reported burr formation research in drilling of aluminum [19–20]. The method of clamping stacked sheets is often dictated by the particular situation, and can have an effect on burr formation [18]. One of the two types of clamps used in these experiments was a squeeze-action hand clamp (De-Sta-Co model 424), shown in Fig. 3a. The other clamp, shown in Fig. 3b, is a plier-operated spring-loaded hole clamp. This type of clamp is frequently used in sheet-metal assembly. It serves to align pre-existing holes in sheets and to provide moderate clamping force. The sheet materials used in the experiments were 2024-T3 and 7075-T6 aluminum alloy, both very common in aircraft structures. The top sheet, referred to here as the “skin”, was always 2024-T3, while the bottom sheet, Level 1
Level 2
38◦ 135◦ TiN Split 0.17 mm (0.0065 ) 4500 1 Hand clamp 35 mm (1.38 ) 7075-T6
– – – Step 0.23 mm (0.009 ) 6000 1 Hand clamp, 1 Hole clamp 55 mm (2.17 ) –
Interfacial Burr Formation in Drilling of Stacked Aerospace Materials
91
Fig. 3 (a) Squeeze action hand clamp (De-Sta-Co 424); (b) Springloaded hole clamp and plier
referred to as the “frame”, was either material. Both the skin and frame were 1.59 mm in thickness. Aluminum 2024-T3 was always used for the skin since it is more ductile than 7075-T6 and will presumably be the “worst-case” skin material in terms of burr formation. A picture of the actual experimental setup is shown in Fig. 4. It can be seen that in this particular example a hand clamp has been placed to the left of the drilled hole and a hole clamp has been placed to the right, and in line with the hand clamp and drilled hole. The lower sheet, or frame, was a section of 90◦ angle screwed to a fixture which was mounted to a Kistler piezoelectric machining force dynamometer. All holes were drilled dry in a Fadal VMC-15 vertical machining center. Both the skin and the frame were 1.59 mm in thickness. The three clamping arrangements studied in the present work are illustrated in Fig. 5. Note that a hand clamp is always placed to the left of the drilled hole, and the “distance from clamp” factor is always measured relative to the
Fig. 5 Sheet clamping configurations used [24]
center of this clamp. Hence, in the third case shown, the hole is drilled 15 mm from the left hand clamp, but is also 55 mm from the hole clamp located on the right. Given the large number of factors to be analyzed, a 36-run restricted orthogonal array design of experiment with two replicates was used. The tests were conducted in random order. The experimental design was restricted in the sense that the drill geometry related factors were not completely balanced. This is due to several reasons, including: (1) the desire to utilize drill geometries that are readily available, (2) the step drills were created by custom-grinding the step geometry into standard twist drills, (3) it was determined that certain combinations of the geometry factors are not feasible since they cannot be realized physically, and (4) in two cases, a coating of black oxide was not available with the given geometry, so an uncoated HSS drill with a bright finish was substituted. It should be noted that black oxide coating is a thin conversion layer of magnetite (Fe3 04 ), formed by a process in which oxidizing salts react with the iron in the High Speed Steel at elevated temperatures [23]. The magnetite layer (i.e. the black oxide layer) acts as an integral protective surface for the drill, thereby increasing wear resistance as well as inhibiting corrosion. It should be emphasized that the black oxide is not an “applied” coating. Rather, it is a coating formed by the “conversion” of the surface of the HSS base metal according to the process described above.
2.2 Drill Wear Experiments
Fig. 4 Photograph of sample drilling setup in the Fadal VMC-15 [24]
A second set of experiments was performed to determine the effect of drill wear on interfacial burr formation. The wear
92
S.N. Melkote et al.
Table 2 Summary of drills used in drill wear experimental (all HSS and 4.91 mm diameter) Helix angle Point angle Coating Point type Drill 1 Drill 2 Drill 3
20◦ 20◦ 20◦
118◦ 118◦ 135◦
Uncoated tip Black oxide Black oxide
Split/step Split/no step Split/no step
experiments were limited to the three drill types listed in Table 2. The choice of the drill types was based on results of the drilling parameter experiments discussed above (and presented later). The same experimental setup that was used in the drilling parameter experiments was also utilized in the drill wear experiments although a Kistler 9272 drilling thrust and torque dynamometer was substituted for the Kistler 9257B three-component force dynamometer. In addition, the frame and skin material used for all drill wear experiments were both 2024-T3, each with a thickness of 1.59 mm. Two hand clamps were used, with the hole being drilled at distances of 25 mm from each clamp. The spindle speed and feed used in the tests were 4500 rpm and 101.6 μm/rev, respectively. Each of the three drills listed in Table 2 were used to drill 1500 holes and two replications of the experiment were performed for each drill geometry. All tests were performed dry.
3 Measurements 3.1 Drilling Parameter Experiments Burrs are small, uneven and easily deformable, making their measurement inherently difficult. For simplicity and speed, two primary methods for interfacial burr height measurement were used for the drilling parameter experiments. The first method involved measuring the pre- and post-drilling separations of the two sheets with a micrometer and using their difference as a measure of the burr height. Details of this measurement method can be found elsewhere [24]. The second burr measurement method was to use an optical comparator to determine the maximum burr height. By placing a precision gage block near the burr and bringing both into focus, the distance from the highest point of the burr to the top of the gage block could be accurately recorded. Knowing the thickness of the gage block, the maximum burr height can be found. Utilizing this method, both the skin exit burr height and the frame entry burr height could be measured individually. In order to reduce variability, measurements of the maximum burr height were repeated four times for each hole drilled by rotating the workpiece by 30–50◦ and recording the maximum burr height in each of the four views. The highest three burr measurements were then averaged to obtain a single measure of the burr height. The skin
entry and frame exit burrs were not examined. Though both measurement methods were used for each drilling test, the optical comparator measurements were used as the primary source of burr height data for further analysis. In addition, a video recording of each test was also made that enabled analysis of the instantaneous separation of the stacked sheets.
3.2 Drill Wear Experiments A similar set of measurements to those performed in the drilling parameter experiments were also performed in the drill wear experiments. For each hole drilled, the instantaneous thrust force and torque at the moment the tip of the drill broke through the first sheet (i.e. the “breakthrough force” and “breakthrough torque”) were measured using the Kistler 9272 drilling dynamometer. The sizes of the skin exit burr and frame entry burr were measured using the optical comparator method once every 50 holes, for a total of 31 burr measurements in each 1500-hole experiment. In addition, the widths (or thickness) of the skin exit burrs were measured using a toolmaker’s microscope. Measurements of the drill flank wear and outer corner wear were also taken once every 100 holes using a toolmaker’s microscope following the method described by Kanai, et al. [25] and illustrated in Fig. 6, which provides an illustration of the end view of a split point drill. The flank wear was calculated by averaging a total of 18 measurements made along the two cutting lips (note that flank wear and outer corner wear measurements are illustrated along one cutting lip in Fig. 6). The outer corner wear was calculated by averaging the outermost wear measurement on each of the two cutting lips on the drill.
Fig. 6 End view of a split point drill depicting outer corner and flank wear measurements performed
Interfacial Burr Formation in Drilling of Stacked Aerospace Materials
4 Results 4.1 Drilling Parameter Experiments Still frames from the videos of an example case are shown in Fig. 7 where only a single hand clamp was located to the left. Notice the initial separation of the two sheets, which was closed by the thrust force of the drill. Once the hole in the skin was completely drilled, the skin sheet sprung up and allowed chips to become entrapped in the interface. Lastly, a cap burr is evident at the frame exit. In a similar case where the hole was drilled near a hole clamp, interfacial separation when the drill breaks through the skin is not evident. Figure 8 shows the main effects of various factors on the skin exit burr and frame entry burr heights. Notice that the skin exit burr height has an average value of 89 μm and the frame entry burr height has an average value of 76 μm. Similar trends for both entry and exit burrs are apparent. The skin exit burr was the larger of the two burrs in nearly every case. Significant trends were determined by an analysis of variance (ANOVA). Effects were considered significant for an alpha value of 0.1, though most effects had considerably smaller p-values. The significant main effects for the skin exit burr measurements are shown in Fig. 9. Note that the 118◦ point angle outperformed the 135◦ point angle. The maximum average burr height increased with increasing feed, which is consistent with data reported in the literature on drilling in single layer of material [1–2, 5].
93
Figure 10 shows the interaction effect of clamp type and clamping distance. Notice that both cases of two supports exhibited a similar trend. In both, the burr was smallest near a hand clamp or hole clamp, and largest in the center. In the arrangement of a single clamp, the burr height increases with the distance of the hole away from the clamp. These trends can be easily explained by the fact that the stacked sheets act as either simply supported or cantilevered beams that deflect elastically under the drilling thrust force. The deflection is maximum in the center, thus causing greater separation between the sheets. In the case of a single clamp (cantilevered beam arrangement), the deflection (and separation) is greatest further away from the clamped end. Consequently, the burr height increases with clamp distance. Performing the same analysis on the frame entry burr height measurements yielded the effects found in Fig. 11. Point angle, clamp type and clamp distance were again found to be significant. In place of feed, the point type is now found to be a significant factor. It is seen that the step drill outperformed either non-step drill. Observing the interaction of split- and standard-points with a step (see Fig. 12), it is clear that the presence of the step has a far greater effect than the non-step standard- or split-point. The interaction effects of clamp type and clamp distance for the frame entry burr were very similar to those for the skin exit burr. The drilling thrust force data was examined, and the instantaneous force when the drill point broke through the skin was determined and used for analysis. The statistically significant main effects are shown in Fig. 13. Two components of drill geometry which impacted thrust force are helix angle and coating type. Point type, feed and speed had a significant effect on thrust force, as well as on interfacial burr formation which indicates a possible correlation. Interfacial burr heights followed the same trends as thrust force for the effects of feed and speed. Step drills reduced thrust force as well as burr height. Conversely, standard-points demonstrated a lower burr height than split-points, however, they required a much greater thrust force, which is consistent with results reported in the literature [26].
4.2 Drill Wear Experiments
Fig. 7 Images from a drilling test; single hand clamp, 3000 RPM, 0.5 m/min feed, 2024-T3 frame
The graph in Fig. 14 shows the average skin exit burr height as a function of the number of holes drilled for each of the three drills studied. As can be seen, Drill 1 (Step drill) produced the smallest mean exit burr height throughout its lifetime, while Drills 2 and 3 produced larger burrs. It is also interesting to note that the increase in skin exit burr height with hole number (or drill wear) is negligibly small for Drill 1 and very mild for Drills 2 and 3.
94
S.N. Melkote et al. Helix Angle
Point Angle
Point Type
Coating
Feed
Clamp Type
Speed
Clamp Frame Distance Material
Burr Height (µm)
110
95
80
65
N d. lit p Ti St Sp Ste
5 e 13 xid O ck
8
11
5 9 00 00 00 30 45 60
4
6.
a
Skin Exit Burr
75
70
1
Bl
m m m 4 m m m 202 5 1 35 55
H 2 an H d, an 1 d H ol e
38
30
50
Frame Entry Burr
Fig. 8 Main effects plot for optical comparator readings of skin exit burr and frame entry burr heights [24] Fig. 9 Significant main effects for skin exit burr height measurements
Point Angle (deg.)
Clamp Type
Feed (mm/rev)
Distance from Clamp
Burr Height (µm)
115
102
89
76
23
ps
m
0.
17 0.
5 13
0.
8 11
10
63 la
2
C
1
p m la C
p, 15 m la co C le 1 C 1
Burr Height (µm)
55
Clamp Type 2 Hand 1 Hand 1 Hand, 1 Hole
140
115
90
65
Fig. 10 Interaction of clamp type and clamp distance on the skin exit burr height
35
15
35 Clamp Distance (mm)
55
Interfacial Burr Formation in Drilling of Stacked Aerospace Materials Point Angle (deg.)
95
Point Type
Clamp Type
Clamp Distance (mm)
Burr Height (µm)
95
85
75
65
55 15
1 H 1 and Ho , le
Ha n
2
St
d
ep
St
1
lit
Sp
d
d
ar
d an
an
5
13
H
8
11
35
55
Fig. 11 Significant main effects for frame entry burr height measurements [24]
Fig. 12 Interaction of standardor split-point and a step for frame entry burr
Point Type Standard Split
Burr Height (µm)
90 80
70
60 50 No Step
Step Step
Coating
Helix Angle (deg.)
Point Type
Feed (0.001"/rev)
Speed (RPM)
340
Thrust Force (N)
300
260
220
180
Fig. 13 Significant main effects on thrust force
de
38
20
k
ac
Bl
xi
O
N
Ti
St
d
ar
d an
lit
Sp
ep
St
4
5
6.
9
00
30
00
45
00
60
96
S.N. Melkote et al.
Fig. 14 Skin exit burr height as a function of hole number
180 Skin Exit Burr Heights (µm)
160 140 120 Drill 1 100 Drill 2 80 Drill 3 60 40 20
00 14
00 12
00 10
0 80
0 60
0 40
0 20
1
0
Number of Holes Drilled
140 Averaged Skin Exit Burr Widths (µm)
Fig. 15 Skin exit burr width as a function of hole number
120 100 Drill 1
80
Drill 2 60
Drill 3
40 20
00 14
00 12
00 10
0 80
0 60
0 40
0 20
1
0
Number of Holes Drilled
The average frame entry burr heights, when plotted as a function of the number of holes drilled, did not reveal a clear trend due to the large variability in the data and hence are not shown here. However, on average, Drill 1 still produced the smallest burrs in this case as well. The widths of the skin exit burrs as a function of the number of holes drilled can be seen in Fig. 15. The graph clearly shows that Drill 1 produced thinner burrs than the other two drills, while Drill 3 produced slightly thinner burrs than Drill 2. The effect of drill wear (with hole number) on the skin exit burr widths is negligible for Drill 1, while its effect on the burr widths produced by Drills 2 and 3 is slightly more pronounced. The averaged flank wear for each drill as a function of the number of holes drilled is shown in Fig. 16. From the graph it is evident that the wear for Drill 1 was smaller than the wear for Drills 2 and 3 over 1500 holes. It is possible that this is partially due to the fact that Drill 1 has a step ground into it, making the area over which measurements were made
slightly smaller. However, it is clear that Drill 1 wears at a slower rate, thus implying longer useful life than Drills 2 or 3. The averaged outer corner wear (not shown here) shows similar trends. The graph of the breakthrough thrust force (measured at the instant the drill breaks through the skin) as a function of hole number is shown in Fig. 17. Drill 1 generated a higher thrust force than the other drills. This is attributed to the fact that the black oxide coating was ground off of the tip of Drill 1 when the step was created. An additional factor that contributes to the higher thrust force is the increase in total cutting edge length in contact with the workpiece because of the presence of the step. Based on the drill wear results presented above it is evident that the step drill (Drill 1) outperforms the other two drills in every measurement taken except for thrust force, and therefore is the most suitable drill geometry for minimizing interfacial burr formation in stacked aluminum sheets.
Interfacial Burr Formation in Drilling of Stacked Aerospace Materials Fig. 16 Averaged flank wear for each drill as a function of hole number
97
180
Averaged Flank Wear (µm)
160 140 120 Drill 1 100 Drill 2 80 Drill 3 60 40 20
00 14
00 12
00 10
0 80
0 60
0 40
0 20
0
0
Number of Holes Drilled
250 Breakthrough Thrust Force (N)
Fig. 17 Breakthrough thrust force for each drill as a function of hole number
200 Drill 1
150
Drill 2 100
Drill 3
50
00 14
00 12
00 10
0 80
0 60
0 40
0 20
0
0
Number of Holes Drilled
5 Conclusions It was found from the drilling parameter experiments that interfacial skin exit burrs and frame entry burrs demonstrated a similar dependence upon all the examined parameters. Averaged across all tests, frame entry burrs were approximately 15% smaller than skin exit burrs. It was found that the most significant factors in interfacial burr formation were drill point angle, clamp type and clamping distance. A combination of a 118◦ point angle with a hole location near either a hand clamp or hole clamp reduced the interfacial burr heights. Additionally, the analysis revealed that smaller feeds produced smaller skin exit burrs, while the use of a step drill reduced the frame entry burr. It is believed that these trends will hold true for both types of interfacial burrs. The most significant finding in the drill wear experiments was that a step drill with a 118◦ point angle (Drill 1) pro-
duced the smallest interfacial burrs, including skin exit burr heights, frame entry burr heights, and skin exit burr widths over the useful life of the drill. This agreed with results found in the drilling parameter experiments. The drill wear experiments also showed that for Drill 1 the burr heights and widths over the complete range of 1500 drilled holes remained nearly constant rather than increasing with drill wear. Furthermore, measurements of drill wear showed that Drill 1 had similar or even lower wear rates than the other two drills. Therefore, the results indicate that Drill 1 (step drill) yields the best results as far as interfacial burr formation is concerned and is the preferred drill for minimizing the same.
Acknowledgements The authors would like to thank Lockheed Martin and the Georgia Research Alliance for sponsoring this project.
98
References 1. Hasegawa, Y., Zaima, S., Yuki, A., 1975, Burr in Drilling Aluminum and a Prevention of It, SME Technical Paper, MR75480. 2. Safronas, A., Taraman, K., 1976, Model Development for Exit Burr Thickness as a Function of Drill Geometry and Feed, SME Technical Paper, MR76-253. 3. Pande, S.S., Relekar, H.P., 1985, Investigations on Reducing Burr Formation in Drilling, International Journal of Machine Tool Design and Research, 26(3):339–348. 4. Aurich, J.C., Sudermann, H., Bil, H., 2005, Characterization of Burr Formation in Grinding and Prospects for Modeling, Annals of CIRP, 54(1):313–316. 5. Safronas, A., Spurgeon, W.M., Taraman, K., 1975, Reduction of Burr Formation in Drilling, SME Technical Paper, MR75-376. 6. Kim, J., Dornfeld, D.A., 2002, Development of an Analytical Model for Drilling Burr Formation in Ductile Materials, Transactions of the ASME, Journal of Manufacturing Science & Engineering, 124: 192–198. 7. Heisel, U., Luik, M., Eisseler, R., Schaal, M., 2005, Prediction of Parameters for the Burr Dimension in Short-Hole Drilling, Annals of CIRP, 54(1): 79–82. 8. Toropov, A.A., Ko, S.L., 2006, A New Burr Formation Model for Orthogonal Cutting of Ductile Metals, Annals of CIRP, 55(1): 55–58. 9. Park, I.W., Dornfeld, D.A., 2000, A Study of Burr Formation Processes Using the Finite Element Method: Part II—The Influences of Exit Angle, Rake Angle, and Backup Material on Burr Formation Processes, Journal of Engineering Materials and Technology, 122(2): 229–237. 10. Min, S., Dornfeld, D.A., Kim, J., Shyu, B., 2001, Finite Element Modeling of Burr Formation in Metal Cutting, Machining Science and Technology, 5(3): 307–322. 11. Saunders, L.K.L., 2003, A Finite Element Model of Exit Burrs for Drilling of Metals, Finite Elements in Analysis and Design, 40: 139–158. 12. Gillespie, L.K., Albright, R.K., Neal, B.J., 1976, Burrs Produced by Drilling, Bendix Corporation, Topical Report BDX-6131248. 13. Guo, Y., Dornfeld, D.A., 1998, Finite Element Analysis of Drilling Burr Minimization with a Backup Material, Transactions of NAMRI/SME, 26: 207–212.
S.N. Melkote et al. 14. Choi, J., Min, S., Dornfeld, D.A., Mahboob, A., Tzong, T., 2003, Modeling of Inter-Layer Gap Formation in Drilling of MultiLayered Material, Proceedings of the 6th CIRP Workshop on Modeling of Machining, 19–20 May, McMaster University, Hamilton, Ontario. 15. Choi, J., Min, S., Dornfeld, D.A., 2004, Finite Element Modeling of Burr Formation in Drilling of a Multi-Layered Material, Proceedings of 7th International Conference on DeBurring and Surface Finishing, 7–9 June, University of California, Berkely, CA. 16. Kim, D., Ramulu, M., 2004, Drilling Process Optimization for Graphite/Bismaleimide-Titanium Alloy Stacks, Composite Structures, 63: 101–114. 17. Ramalu, M., Branson, T., Kim, D., 2001, A Study on the Drilling of Composite and Titanium Stacks, Composite Structures, 54: 67–77. 18. Sisco, T., 2003, Achieving “One Up Assembly” by Reduction of Interface Burr Height in Aluminum, Graphite, and Advanced Titanum/Graphite Hybrid (TiGr) Material, SAE Technical Paper Series, 2003-01-2896. 19. Huang, M.-F., Lin, T.-R., 2004, Application of Grey-Taguchi Method to Optimize Drilling of Aluminum Alloy 6061 with Multiple Performance Characteristics, Materials Science and Technology, 20: 528–532. 20. Rivero A., Aramendi, G., Herranz, S., López de Lacalle, L.N., 2006, An Experimental Investigation of the Effect of Coatings and Cutting Parameters on the Dry Drilling Performance of Aluminum Alloys, International Journal of Advanced Manufacturing Technology, 28: 1–11. 21. Bahr, B., Mankad, T., 1999, Study of Drill Tool Geometry in HighSpeed Drilling of Aluminum Sheet Metal, SAE Technical Paper Series, 1999-01-2295. 22. Ko, S. L., Chang, J.E., 2003, Development of Drill Geometry for Burr Minimization in Drilling, Annals of CIRP, 52(1): 45–48. 23. http://www.clevelandblackoxide.com/what.aspx; accessed on 6 February, 2009. 24. Newton, T.R., Morehouse, J., Melkote, S.N., Turner, S., 2007, An Experimental Study of Interfacial Burr Formation in Drilling of Stacked Aluminum Sheets, Transactions of NAMRI/SME, 36: 437–444. 25. Kanai, M., Iwata, K., Fujii, S., Kanda, Y., 1978, Statistical Characteristics of Drill Wear and Drill Life for the Standardized Performance Tests, Annals of CIRP, 27: 61–66. 26. Kinman, M.D., 1963, Precision Drilling with Standard Twist Drills, Machinery, 102(2633): 1014–1018.
Burr Formation in Drilling Intersecting Holes L. Leitz, V. Franke, and J.C. Aurich
Abstract The increasing power density in engine manufacturing as well as the complexity of parts in automotive production demand for an entire control of burr formation especially in regard of intersecting holes. This paper presents an approach to control burr formation concerning reproducible generation of burrs depending on intersecting geometry and process parameters. Therefore two process models describing burr formation and burr cap formation and results of experimental investigations of different workpiece materials (AISI 4140H (42CrMo4), 226D AlSi9Cu3)) were combined. Keywords Drilling · Burrs · Intersecting holes
1 Introduction Drilling is of particular importance within the field of machining. As an effect of this process particularly burrs are generated besides chips. On the one hand handling and mounting burr affected components causes a considerable risk of injury. On the other hand especially burrs getting separated during operating time lead to the risk of possible functional surface damage or flow disturbances in ductworks (e.g. at oilway holes of cylinder heads), which can result in total loss of the functional units. Therefore, machining processes are mostly followed by cost-intensive cleaning and deburring processes. Furthermore, burr affected parts require a high inspection effort to assure component quality. Especially components whose reliability is affected by burrs need a 100% checking resulting in high time and cost efforts. In this context burr formation at intersecting holes is
L. Leitz (), V. Franke, J.C. Aurich Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany e-mail:
[email protected] url: www.fbk-kl.de
proved to be exceptionally critical because of a high detection and removal effort for burrs located inside of drills. A study carried out in the German automotive and machine tool industries is focusing on costs associated with burr minimization, deburring and part cleaning. To evaluate the economic impact of expanses caused by burrs the participants of the survey were asked to name the manufacturing share related to burrs for a specific workpiece. The expenses are caused by an increase of about 15% in manpower and cycle time. In addition, 2% share in the reject rate and 4% share in machine breakdown times were reported. Summarizing the presented distribution without regarding additional efforts the share accounts up to 9% of the total manufacturing cost [1]. The presented investigation within this paper aims at an enhanced understanding of burr formation mechanisms of intersecting holes during drilling. Based on typical industrial applications geometry relations of intersecting holes and influences of drilling parameters are analyzed. These results enable the control of burr formation in terms of reproducible production of burrs. Furthermore, a considerably increase of process reliability of deburring processes (e.g. abrasive flow machining or vibratory grinding) and remarkably reductions of process times for deburring and quality inspections can be achieved.
2 State of the Art Intersecting holes (base hole (first hole) and cross hole (second hole)) are common in industrial production processes, e.g., as coolant ducts in cylinder heads, crank shafts or gear shafts, as ducts in hydraulic elements or in the area of fuel injection in common rails or injections valves. Park [2] investigated the influence of the exit surface angle and identified that burr formation decreases with increasing exit surface angles. The pivoting point that initiates plastic bending leading to large burr formation appears very close to the machined surface when the exit surface angle is 30◦ . As the exit surface angle decreases, the pivoting point moves
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_11, © Springer-Verlag Berlin Heidelberg 2010
99
100
L. Leitz et al. Tangential line to exit surface
Exit surface
Exit surface
Tangent to the exit surface “Burr Angle”, α 1
analytical process model is developed to visualize the process kinematics when drilling intersecting holes. On the other hand experimental investigations with varying hole diameter, geometry relations and process parameters are conducted. The combination of both results, considering material properties as well, enables the prediction of burr formation and burr shape.
Drill path “Burr Angle”, α2 (a) Exit surface is normal to a drill path
(b) Cylindrical exit surface
Fig. 1 Burr formation when drilling intersecting holes [4]
farther away from the machined workpiece and causes larger burrs. A burr formation model related to drilling of intersecting holes was developed in [3]. An interaction angle that defines the interaction between the cutting edge and the exit surface was proposed under assuming constant exit surface geometry. It includes dynamic motion of the cutting edge induced by feed and cutting speed. If the interaction angle is positive, the cutting edge exits from the workpiece and vice versa. The model can predict the likely burr formation area that can be represented as the positive interaction angle. The area increases as feed increases, speed decreases, and the exit surface angle decreases. An effective exit surface angle was proposed in order to incorporate the modification of the exit surface geometry during drilling. Due to plastic deformation at the end of a cutting process, the exit surface geometry changes. Depending on the angular position of the exit surface, the effective exit surface angle changes. A small negative exit surface angle leads to early initiation of the bending mechanism and results in a large burr. Hence, thinner parts of a workpiece may have a larger burr. The interaction angle determines exiting and entering of the cutting edge, and predicts the likely burr formation area. The effective exit surface angle defines the size of burr and shifts the likely burr formation area calculated through the interaction angle in the rotational direction of the drill [3] (Fig. 1). In [5] it is shown that results regarding the drill exit from perpendicular and angled surfaces can be applied to concave exit surfaces. Hence the geometry of cutting edge plays an essential part in burr formation. However, burr formation is decisively influenced by the radius of the exit surface. This is shown for burr formation on concave surfaces without offset.
3 Matter of Investigation The following presented investigations on control of burr formation are structured into two sections. On the one hand an
4 Process Model The presented process models enable the explicit description of kinematics of the cutting edge relative to the exit surface. The correlation between tool exit and burr formation are analyzed. This approach serves to improve the comprehension of the drilling process when producing intersecting holes. In a first step sixteen selected geometry relations were investigated to analyze tool exit and tool entry conditions by using simulation. These results enable to develop an analytical process model. Figure 2 shows the varied geometry parameters as well as a visualization of production processes of intersecting holes by simulation. Additionally high speed cinematography was applied to record the cutting edge exit. Combining these results with visualizations from above as well as further calculations enable a graphical description of the intersection line of base and cross hole and the identification of cutting edge exit depending on feed. Figure 3 shows an eddy line at an inclination angle of 90◦ , 0% eccentricity, and base hole dB = 16 mm and cross hole dQ = 8 mm. The specific tool exit phases, reveal the exit conditions of the cutting edge when drilling a cross hole. Apparently no continuous tool exit occurs contrary to planar exit. In fact the special geometry relations in intersecting holes lead to a permanent alteration of tool entry into and tool exit out of the workpiece material. Therefore, continuous burr formation at the exit surfaces of the cross hole is not possible. Burr values vary depending on exit conditions. The developed analytical process model was applied to selected geometry relations. These reveal a major influence of inclination angle and eccentricity on tool exit and entry conditions and thus on burr formation when drilling intersecting holes. In addition to the analysis of tool exit conditions the remaining material in front of the cutting edge at the position of tool exit has been investigated. The remaining material volume is split into four sections which are related to each other. Four possible cases to define the cutting edge exit position resulted. In the first case the tool exits with the drill point (α = 90◦ , e = 0%). In the second case the cutting edge and in the third case the major cutting edge exit first. The tool exit positions depend on inclination angle α and the ratio
Burr Formation in Drilling Intersecting Holes
101
Fig. 2 Variation of geometry relation and visualization applying simulation technique
0°
0° 360°
360°
0°
360°
270° 180°
0° 90°
02
1 1Material 234-
4 3
5No material 60°
45°
90°
135°
180°
225°
270°
315°
360°
Angle position
Fig. 3 Visualization of cutting edge exit and identification of critical sections
of base hole to cross hole diameter. The tool never exits the material entirely in the fourth case, due to a large eccentricity. Because of drill cap formation depends significantly on the tool exit position this approach helps to determine the tendency of drill cap formation for different geometry relations.
5 Experimental Investigations For a better understanding of burr formation in intersecting holes metallographic cross sections illustrating the different phases of burr formation are shown in Fig. 4 which
102
L. Leitz et al. Sectional View:
Diameter:
Material:
4140H
Cutting speed.:
100 m/min
dB
16 mm
Feed:
0,25 mm/rev
dQ
8 mm
Coolant concept:
Internal coolant
Eccentricity:
0%
Coolant:
Emulsion
Concentration:
9%
Intersection angle:
90°
B1 A1
B1 A2
A1
A2 B2
B2
A1-A2 Phase 1
A1-A2 Phase 2
A1-A2 Phase 3
A1-A2 Phase 4
B1-B2 Phase 1
B1-B2 Phase 2
B1-B2 Phase 3
B1-B2 Phase 4
Fig. 4 Phases of burr formation in intersecting holes
were enabled by feed stops. It points out how burrs and burr caps are formed in intersecting holes. Furthermore, burrs are forming during different points of time along the intersection line. In section A1-A2 burr formation is finished in phase 3 whereas in section B1-B2 at the same time burr formation is not yet finished. Thus, the geometry of intersecting holes leads to an alteration of tool exit and entry after phase 3 and to a variation of burr formation mechanisms. This causes an irregular burr height distribution and complicating following deburring processes difficult.
Fig. 5 Drill cap formation in varying geometry relations
Within the experimental investigations on burr control eight out of sixteen different geometry relations have been regarded in the kinematical process model examined. These intersecting geometries varied in hole diameter, inclination angle, and eccentricity. Furthermore, the influence of drilling parameters, cutting speed, and feed, as well as, different workpiece material (AISI 4140H, 226D) were analyzed. To measure the resulting burr heights a laser triangulation sensor (LTS) was applied [6], whereas an available system was enhanced by two linear axes and a rotational axis to enable
Burr Formation in Drilling Intersecting Holes
103
In Fig. 5 drill caps at different geometry relations are visible in AISI 4140H at a cutting speed of vc = 100 m/min and feed of f = 0.25 mm/rev. A detailed analysis of the test series reveals a large impact of intersecting geometry on burr formation. For illustration, two examples with different intersecting geometry are given below. In both examples the varying tool exit conditions cause an irregular burr height distribution (Fig. 4). In the first example using an intersecting angle of α = 90◦ maximum burr height occurs between 0◦ (B1) and 100◦ (A1), as well as between 180◦ (B2) and 280◦ (A2) caused by material deformation in these areas and the cutting edge exit conditions as shown in the analytical process model. This leads
burr height measurement at intersecting holes. The application of these axes ensures a continuous perpendicular guide way of the laser beam was. Measurements were verified with measurements at cross sections by preparing two cross sections (90◦ rotated) for each test series. A preliminary visual inspection of the workpieces was carried out to detect drill caps. Thereby, the intersecting geometry has a major effect on drill cap formation. Independent of the cutting parameters drill caps appeared particularly at inclination angles of α = 90◦ for AISI 4140H. Machining of the aluminum alloy 226D revealed no drill caps, as the aluminum alloy tends to break out due to its high silicon content.
Example1: Material:
AISI 4140H
Cutting speed.:
100 m/min
Feed:
0,15 mm/rev
Diameter: dB
Geometry relation: 16 mm
Eccentricity:
0%
Coolant:
Emulsion
Concentration:
9%
Intersection angle:
90° A1
Burr height [µm]
Winding off of burr height measurement 400 B1 350 300 250 200 150 100 50 0 0
B2
A1
90°
4 mm
dQ
Coolant concept: Internal coolant
0%
A2
B1
B1
A2
A1
A2 B2
B1
B2
B1 60
120 180 240 Angle position [°] hole 1
hole 2
300
360
A2
A1 B2
hole 3
Example2: Material:
AISI 4140H
Cutting speed.:
100 m/min
Feed:
0,15 mm/rev
Diameter: dB
Coolant concept: Internal coolant Coolant:
Emulsion
Concentration:
9%
Geometry relation: 12 mm
dQ Eccentricity:
0%
6 mm
Intersection angle:
90° A1
Burr height [µm]
Winding off of burr height measurement 400 B1 350 300 250 200 150 100 50 0 0
60°
0%
A2 B1
A1
A2
B2
B1 A1
A2 B2
B1 60
120
180
240
300
Angle position [°] hole 1
Fig. 6 Burr formation in intersecting holes
hole 2
360
A1
A2 B2
hole 3
B2
B1
104
L. Leitz et al.
intersecting geometry. The kinematical process model was enhanced by areas with increased burr formation confirming the results of burr measurement. In the first of three examples only tool exit and entry conditions are responsible for burr formation. The kinematical process model shows that areas between 0◦ and 10◦ as well as between 180◦ and 280◦ can be identified as areas with increased burr formation. Whereas for the following examples exit/entry conditions along with a lack of supporting material result in higher burrs (Fig. 7). Figure 8 demonstrates the results of combining the kinematical process model and experimental investigations for 226D. For better comparison the same geometries were chosen. For the aluminum alloy areas with breakouts can be defined due to described tool exit and tool entry correlation of cutting edges and the lack of supporting material. Combining analytical process model for tool exit and tool entry conditions and the calculation of remaining material in front of the cutting edge at the position of tool exit with experimental investigations allows drawing conclusions in which area burrs or breakouts are formed in intersecting holes. Furthermore, a probability of drill cap formation can be given subject to intersecting geometry.
to an increased burr formation. In the second example with an intersecting angle of α = 60◦ , maximum burr height was measured in the section between 180◦ (B2) and 320◦ (A2). For position B2 the same burr formation mechanisms as shown in the first example occur. However, at position A2 the lack of supporting material is responsible for burr formation (Fig. 6). These results show that burr formation and drill cap formation depend on the geometry relations of intersecting holes chosen.
6 Relationship of Kinematical Process Model and Experimental Investigation Combining the presented analytical kinematical process models and the experimental results enables to indicate critical regions at intersecting holes and to predict areas of maximum burr height. The process model concerning tool exit and entry conditions provides explanations for areas of maximum burr height whereas additionally regions of little burr formation can be described. This approach enables to optimize subsequent deburring operations for their application. Further it is revealed that drill cap minimization can be provided in changing inclination angle or eccentricity of the drill axis. In these cases a more favorable tool exit point with regard of drill cap formation can be realized. Figure 7 shows the relationship between kinematical process model and experimental investigation for AISI 4140H and three examples of
7 Summary Focus of the presented work is the control of burr formation in terms of reproducible generation of burrs in intersecting holes depending on intersecting geometry and process
Kinematical process model with increased burr formation areas
Intersecting Geometry
Experimental Investigation
16 mm 8 mm 90°
material
90°
1 2
0%
3
180°
4
0°
No material
5
Burr height [μm]
0
dB dQ
400 350 300 250 200 150 100 50 0 0
60
120
180
240
300
360
Angle position [°] 6 0∞
45 ∞
90 ∞
135 ∞
180 ∞
225 ∞
270 ∞
315 ∞
360 ∞
270°
Winkellage
0
16 mm 8 mm
1
material
90°
2
2500
3 4
90°
50%
5 6
180°
No material
7
0°
8
Burr height [μm]
dB dQ
2000 1500 1000 500 0
9
0
50
0∞
45∞
90∞
135 ∞
180 ∞
225∞
270 ∞
315 ∞
360 ∞
0
60°
250
300
350
400
material 4
0%
200
90°
2
6
180°
8
0°
10
Burr height [μm]
16 mm 8 mm
150
270°
Winkellage
dB dQ
100
Angle position [°]
10
350 300 250 200 150 100 50 0 0
No material 12 0∞
45∞
90∞
135 ∞
180 ∞
225∞
270 ∞
315 ∞
360 ∞
60
120
180
240
Angle position [°]
270°
Fig. 7 Relationship between kinematical process model and experimental investigation for AISI 4140H
300
360
Burr Formation in Drilling Intersecting Holes
105
Fig. 8 Relationship between kinematical process model and experimental investigation for 226D
parameters. Two kinematical process models describing tool exit and entry conditions, as well as the calculation of remaining material are developed to enhance the understanding of burr formation in intersecting holes. Further experimental tests are conducted to reveal the influence of intersection geometry, cutting speed, feed, and workpiece material on burr formation. The combination of process models and experimental results enables well funded declarations on burr position and shape depending on intersection geometry. Hence, areas with increased burr formation when drilling intersecting holes can be identified. Moreover, the experimental examination of two workpiece material of tool exit allows conclusions on drill cap formation. In future steps the influence of tool wear on burr formation in intersection holes will be analyzed. Furthermore, a guideline, presenting a detailed description of the approach on control of burr formation at intersecting holes, will be prepared. Acknowledgments The presented investigations are sponsored by the Research Association for Machine Tools and Manufacturing Technology (FWF) which is an independent part of the VDW responsible for industrial research.
References 1. Aurich, J.C., 2006, SpanSauber, Untersuchung zur Beherrschung der Sauberkeit von zerspanend hergestellten Bauteilen, University of Kaiserslautern 2. Park, I., 1996, Modeling of Burr Formation Process in Metal Cutting, PhD. Thesis, University of California at Berkeley, Berkeley, CA 3. Min, S., Dornfeld, D.A., Nakao, Y., 2003, Influence of Exit Surface Angle on Drilling Burr Formation, Transactions of the ASME, Journal of Manufacturing Science and Engineering 125(4): 637–644 4. Min, S., 2001, Modeling of Drilling Burr Formation and Development of Expert System, Ph.D. Thesis, University of California, Berkeley 5. Heisel, U., Schaal, M., 2008, Burr formation in intersecting holes, Production Engineering, Research and Development (Online) 2(1):55–62 6. Aurich, J.C., Wiese, M., Tries, T., Gsänger D, 2004, Gratbildung und Gratmessung beim Bohren, WB Werkstatt und Betrieb, 137(12):44–47
Chip Breakage Prediction by a Web-based Expert System F. Klocke, D. Lung, and C. Essig
Abstract Chip control is essential for the manufacturing with geometrical defined cutting edge to avoid interruptions of the process, damages on the machine or injuries of the machine operator. Hereby the prediction of the chip shape is highly relevant and scope of the present research work. High speed filming of the flowing chips, analyses of the contact zones between cutting tool and chip; as well as TM cutting simulations using the software DEFORM 3D are conducted to achieve an understanding of fundamental relations and the effectiveness of different chip groove geometries. A newly developed web-based expert system searches for similar datasets in a database with captured applications and is furthermore able to predict the chip shape for a planned application. Therefore the chip breakage borders for many tool-material-combinations are determined empirically and captured in the database. The software judges by these borders, if for a certain parameter set point, either chip breakage or unfeasible chip shapes occurs. Interand extrapolations for so far unknown datasets are also based on the borders. These operations are operating with correction values and the data output is combined with a declaration of the output accuracy. Beside the search for similar applications and the prediction of chip breakage, the expert system is also capable to evaluate the effectiveness of chip groove geometries of cutting tools. Keywords Chip control · Chip breakage prediction · Process reliability
ity by the customers and a growing cost pressure due to an increasing competition. By the increase of automation in manufacturing processes the demands on process reliability also increase. For cutting processes this demands chip control, to avoid interruptions of the process, damages on the machine or injuries of the machine operator. At the same time the implementing costs for new manufacturing processes are high in automated production due to the higher integration of manufacturing cells. Therefore it should be ensured to achieve good chip shapes without the execution of timeconsuming preliminary testing. Hereby the prediction of chip shape is highly relevant.
2 Approach The research deals with the prediction of chip breakage for cylindrical turning operations with continuous cut. A database with captured applications is the core of the developed software. On this base, an expert systems searches for similar application and furthermore predicts the chip shape for planned applications. High speed filming of the flowing chips, analyses of the contact zones between cutting tool and chip as well as cutting simulations using the software TM DEFORM 3D are conducted to achieve an understanding of fundamental relations and the effectiveness of different chip breaker geometries.
1 Introduction Essential drivers for the increase of automation in manufacturing are the growing demand on manufacturing qual-
F. Klocke (), D. Lung, C. Essig WZL, Laboratory for Machine Tools and Production Engineering, Aachen University, 52056 Aachen, Germany e-mail:
[email protected] url: www.wzl.rwth-aachen.de
2.1 Investigations of Chip Flow and Chip Breakage Mechanisms The analysis of the contact zones in comparison to the chip flow recorded by a high speed filming device have shown which elements of a chip breaker geometry of a tool are guiding the chip effectively. Furthermore the chip guiding mechanisms are identified. Therefore the depth of cut and the feed
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_12, © Springer-Verlag Berlin Heidelberg 2010
107
108
F. Klocke et al.
Fig. 1 Examples for the analysis of the chip flow and the contact zones
rate respectively are decreased until the borders of controlled chip breakage are exceeded. In Fig. 1 snap shots of the high speed filming and SEM pictures of the contact zones are illustrated for both, cutting conditions with controlled chip breakage (depth of cut ap = 1.25 mm) and cutting conditions where no chip breakage occurred (depth of cut ap = 1 mm). Important advantages were achieved in the past years in the simulation of cutting processes with the FEM. Especially the chip flow can be calculated by FEM software in good accordance to the reality [1]. One of the major advantages of the FEM simulation compared to other modelling methods is the possibility of applying tools with complex chip breaker geometries in the cutting simulation. For cutting simulations the FEM software DEFORMTM was applied for the investigation of the chip flow. The simulated chip flows were verified by high speed filming records of the real chip flow. Several cutting simulations with varying cutting conditions were conducted to investigate the influences of the feed rate and the depth of cut on the chip flow. A comparison of different cutting simulations with varying depth of cut is shown in Fig. 2. Two major correlations between the cutting conditions and the chip breakage were identified by the contact zone analysis, the high speed records and the cutting simulations. Firstly a minimum chip thickness, which is mainly deter-
Fig. 2 Simulation of the chip flow with different feed rates
mined by the feed rate, is needed to achieve a certain formability of the chip. The minimum chip thickness depends on the groove geometry of the cutting insert and the workpiece material. Secondly the depth of cut is one of the major influences on the chip flow angel. This angle determines the contact conditions between the cutting tool and the flowing chip. Thereby the depth of cut is mainly determining the elements of the cutting tool and their effectiveness regarding chip forming.
Chip Breakage Prediction by a Web-based Expert System
109
2.2 Relevance and Weighting of Influences Regarding Chip Breakage The relevance and weighting of process, tool and workpiece related parameters are evaluated in cutting experiments. Thereby the widely used chip breakage polygons are applied. These polygons symbolize the borders of controlled chip breakage in the depth-of-cut/feed-plane. The borders are determined by the decrease of the cutting parameters feed rate and depth of cut starting at parameter sets were controlled chip breakage occur until the chips are no longer breaking. In Fig. 3 the borders of chip breakage for different parameters are illustrated. Thereby only one parameter is varied while the other cutting conditions are constant. The investigated parameters were tool geometry, cutting edge radius, cutting speed and workpiece material. If the borders are close to each other, such as for the cutting speed in the upper right side, the influence of this parameter on the chip breakage behaviour is small. Corrective values are derived from the empirically determined borders of chip breakage. Furthermore rules for the extra- and interpolations for the developed software tool are based on these values. Thereby borders of chip breakage
can be computed for applications, which are not fed into the database. Major parts of the software, such as the chip breakage prediction and the search for similar applications are based on these so calculated borders of chip breakage.
3 Database Chip breakage depends on a multiplicity of process, tool and workpiece related influences [2]. Hence it is necessary to collect all process relevant influences of an application. Thereby the concrete cutting conditions, such as the cutting parameters, tool geometry, material and type of cooling as well as the occurring chip shape for the captured applications, are recorded in the database. The chip shape is categorised according to the “Stahl-Eisen-Prüfblatt” into ten different chip shapes. The different categories are illustrated in Fig. 4. Beside the data quality the quantity is limiting the prediction accuracy. This is because several applications are equivalent with more supporting points for inter- and extrapolations. The present database is based on two different sources: on the one hand cutting experiments were conducted at the Laboratory for Machine Tools and Production Engineering
Tool geometry
Cutting speed Chip Breaker 1 Chip Breaker 2 Chip Breaker 3 Chip Breaker 4 Chip Breaker 5 Chip Breaker 6
3 2,5 2 1,5
CNMG120408 vc = 200 200m/min m/min rε == 0,8 0,8mm mm C45E+N dry
1 0,5
3,5
Depth of cut ap / mm
Depth of cut ap / mm
3,5
3 2,5 2 1,5
0,1
0,2
0,3
0,5 0
0,4
0
0,1
Feed f/mm
0,2
0,3
0,4
0,5
Feed f/mm
Corner Radius
Workpiece material
3,5
3,5
3
0,5
2,5 2 1,5
0,4 mm 0,8 mm 1,2 mm 1,6 mm
CNMG1204xx Chip Breaker 5 vc == 200 200m/min m/min C45E+N dry
1 0,5 0 0
0,1
0,2
0,3
0,4
0,5
Depth of cut ap /mm
Depth of cut ap / mm
100 m/min 200 m/min 300 m/min 400 m/min
CNMG120408 Chip Breaker 1 rεε = 0,8 mm mm C45E+N dry
1
0 0
0,5
0,5
3 2,5 2 1,5
CNMG120408 Chip Breaker 1 vc == 200 200m/min m/min rε == 0,8 0,8mm mm C45E+N
1 0,5 0
C45E+N C15 X5CrNi18-10 42CrMo4 16MnCr5
0
0,1
Feed f/mm Fig. 3 Analysis of the parameter influence by comparison of the borders of chip breakage
0,2
0,3
Feed f/mm
0,4
0,5
110
F. Klocke et al.
1
2
3
4
5
6
7
8
9
10
Fig. 4 Chip shape categories according to Stahl-Eisen-Prüfblatt
(WZL) at the Technical University of Aachen and on the other hand existing databases from the involved tool suppliers out of the project accompanying board are comprised.
4 Search for Similar Applications The operator is widely supported by the expert system in the search for similar applications. In a first step the data sets are limited in a selection mask. The selection mask is illustrated in Fig. 5. It is possible to indicate the upper and the lower limit separately for the following parameters: cutting
speed, depth of cut, feed and cutting edge angle. Thereby the range can be determined by each operator for the present application. The limitation can be illustrated by three different methods: Firstly the illustration as a list with all details of an application, secondly the illustration of chip photos in a non-scaled depth-of-cut/feed-plane and thirdly the illustration as coloured symbols in a scaled depth-of-cut/feed-plane. In the displayed selection mask the limitation of application with the method of coloured symbols is chosen in the lower right side of the figure. In Fig. 6 the display method of coloured symbols for certain cutting parameters is illustrated. Each point in the depth-of-cut/feed-plane symbolizes one data set. The
Fig. 5 Limitation of captured data sets for the search for similar applications
Depth of cut ap / mm
Chip Breakage Prediction by a Web-based Expert System Manufacturer: Insert: Groove Geometry: Material: Cutting Speed.: Cooling: Chip Shape Quality:
Chip Shape Class: Breakage Border:
Sandvik CNMG 120408 WF C45E+N vc = 200 m/min dry good feasible poor 3, 4, 9, …
Feed f / mm
Fig. 6 Borders of the area of controlled chip breakage
numbers beside the symbols are the chip categories according to the chip shape categories in Fig. 4. The colour green symbolises good chip shape, yellow symbolises feasible chip shape and red stands for poor chip shape. Each operator can define on his own which chip forms are good, feasible or poor. The borders of controlled chip breakage are drawn as blue lines in this display.
5 Chip Breakage Prediction A further task for the expert system is the prediction of chip breakage. Therefore an algorithm checks, if the point of interest is placed in or out of the area of controlled chip breakage. The borders of this area are described by two different models, depending on the number of captured data. The first model is based on the empirical determination of the boarders. It features high prediction accuracy but additional cutting tests are needed. Firstly the changeover from acceptable chip breakage to bad chip shapes in the depth-ofcut/feed-plane is determined at several points. Therefore the feed and the depth of cut respectively are decreased beginning in the area of controlled chip breakage until poor chips occur. Secondly the borders are approximated by simple geometrical shapes through the determined points. The most of the chip breakage problems occur in industrial applications in the area with low depth of cut and low feed rates respectively. Therefore the chip breakage behaviour in this area is investigated very precisely. The second model for the determination of the area of controlled chip breakage is used, if there is no analytic description of the borders. This is the case, if for a special tool-material-combination the collected data sets are not sufficient to describe the borders exactly. In this case the system places a polygon around the data points with the same chip
111
shape. The chip breakage prediction is also conducted with the described algorithm, which checks if the point of interest is inside or outside of the borders. If a point is outside of the borders of known chip shapes, the chip shape with the closest borders is the most supposable. The chip breakage borders for many tool-materialcombinations are determined empirically and captured in the database. The borders are the base for later inter- and extrapolations. This is done by correction values, which are also captured in the database. The correction values are related to each border points that determine the borders of controlled chip breakage. The correction values include the amount and the direction of the replacement of each border point. This allows the widening in some areas and a constriction of the area of controlled chip breakage in other areas in the depthof-cut/feed-plane at the same time. The influences of the corner radius, the chip breaker geometry of a cutting insert, the cutting speed and the workpiece material are captured in the database as correction values. The data output is combined with a declaration of the output accuracy. The output accuracy decreases if inter- or extrapolations are needed or if the point of interest is close to the border of controlled chips or outside the automatically generated polygon. The best output accuracy is achieved, if the point of interest is on the inside of the area with controlled chip breakage where no inter- or extrapolations are needed.
6 Investigations on the Tool Influence Beside the search for similar applications and the prediction of chip breakage the expert system is also capable to evaluate the effectiveness of chip groove geometries of cutting tools. Therefore the geometrical parameters for the recorded cutting tool geometries are determined by tactile measurements and captured in the database. These parameters are the base for the calculation of effective radii, which is a degree of the bending of the chip at the chip groove geometry. For the calculations other basic rules are consulted into consideration. So the contact between tool and chip is changing depending on the cutting conditions. Only for a chip thickness bigger than a certain minimum value the chip shows enough stability to get leaded by the chip groove geometry. For these correlations rules are defined and captured in the expert system. Using these rules, the effective chip radius R for different cutting conditions can be calculated [3]. In Fig. 7 an equivalent system for the bending of a chip by the tool chip breaker is illustrated. The equivalent system is a cantilever, which is bended with a constant radius R.
112
F. Klocke et al.
σ(y) =
Mb W
(1)
Mb (x) = E · I · w (x) I h / 2
W=
σ(y) =
In this equivalent system the maximum tensile stress in a flowing chip can be calculated. According to the Eqs. (1), (2), (3), (4), (5), (6) the quotient between the effective chip radius R and the chip thickness h’ is proportional to the theoretical tensile stress in the flowing chip.
(3)
E · I · h / 2 · w (x) I
(4)
w (x) = 1 / R
(5)
h · E 2·R
(6)
σ(y) =
Fig. 7 Equivalent system for the bending of a chip by the chip breaker of a cutting tool
(2)
This quotient can be taken into account for the evaluation of the chip breaking characters of different tool geometries. Therefore the tool geometries are captured by tactile measurements and the effective bending radii were calculated for different feed rates and depths of cut. Based on these measurements, the quotient of the effective bending radius could be calculated for different cutting parameters. Clear differences in the chip breaking characters can be found in the comparison of these quotients for different tool geometries. Figure 8 these quotients are displayed for roughing, medium conditions and finishing tools. The larger the value for the quotient is the stronger is the forming of the chip at the chip groove geometry.
Chip groove geometry
M3
Roughing
FF1
Medium conditions
1 0
f / mm
0.
0. 2
0. 3
4
3 2 1 a / mm p
0.3 0.24 0.18 0.12 0.06 0
0.
hch / Rwirk
0
0.
f / mm
1
4 0. 3
0. 2
0
1 0.
0. 2
0. 3
0.
f / mm
3 2 1 a / mm p
0.3 0.24 0.18 0.12 0.06 0
0.
hch / Rwirk
0.3 0.24 0.18 0.12 0.06 0
4
hch / Rwirk
MR3
3 2 1 a / mm p
Finishing
Better chip breaking behaviour 0,03–0,06
0,09–0,12
0,15–0,18
0,21–0,24
>0,27
0–0,03
0,06–0,09
0,12–0,15
0,18–0,21
0,24–0,27
Fig. 8 Chip breaking character of several cutting inserts for different applications
Tool supplier: Tool:
Seco CNMG 120408
Chip Breakage Prediction by a Web-based Expert System
7 Implementation The user-friendly design of the interaction masks was in the focus of the programming of the database. In the generation of a new application, the operator is widely supported by the system, because already captured data, such as materials or tools, can be chosen from a list and selected by a mouse click. The system overtakes all the attributes of these tools or materials for the new application. The software is a modular designed web application and interacts with the operator by a HTML-surface, which can be operated by every web browser. For the data storage a relational database system is applied. The calculation models and the interface between HTML-surface and the database system are implemented with the programming language PHP. This language offers interfaces for many database systems and an extension module for most of the web browser and operating systems. Thereby it is possible to integrate the developed software in the existing IT-infrastructure of a company. The data access as well as the analysis module is observed by security checks. So it is possible that one group can use the analysis modules without having access to critical data. A cost-efficient use of the system is guaranteed due to the modular design of the software, the easy configuration and starting up as well as the use of open-source software. Thereby no additional software has to be installed on the computer of the operators.
113
breakage for planned cutting operations can be predicted. The software tool is based on captured data sets in a database. In this database, data is captured from tool supplying companies as well as data from cutting tests conducted at the Laboratory for Machine Tools and Production Engineering (WZL) in Aachen (Germany). Basic investigations, such as chip flow simulations, high speed filming and analysis of the contact zones between chip and tool were conducted to achieve a better understanding of the chip breaking mechanisms. Basic results of these investigations are influenced by the cutting parameters and chip breaker geometries on the chip breaking behaviour. This knowledge was translated into numerical algorithms for the developed software modules. Thereby the system is enabled to search for similar data sets and to predict the chip breakage for planned turning operations. Furthermore a module for the description of the chip breaking character of complexly shaped chip breaker geometries was developed. A user-friendly interface of the software tool as well as the accessibility of the software via internet was in the focus of the software design. Modular layouts of the programs as well as the use of open-source software solutions guarantee the integrations of the developed software in the existing IT-infrastructure. Acknowledgements The authors gratefully acknowledge the financial support of the Research Association for Machine Tools and Manufacturing Technology (FWF).
References 8 Conclusion and Outlook Chip breakage is highly needed to guarantee a reliable cutting process. Especially for cutting operations with continuous cut, such as cylindrical turning, the chip breakage is essential to avoid injuries of the operator, interruptions of the manufacturing process or damages on the machine or workpiece. With a web-based software tool the chip
1. Klocke, F., Frank, P., Vormann, K., Fiderer, M.: Simulation des Drehprozesses mit DEFORM. In: Abschlussbericht des Verbundforschungsprojekts SindBap, Aachen: Eigendruck des WZL Aachen, 2007, S. 23–34 2. Hintze, W.: Modellgestützte Spanbruchbeurteilung beim Drehen. Hamburg-Harburg, Technische Universität, Dissertation, 1990 3. Nedeß, C., Hintze, W.: Characteristic parameters of chip control in turning operations with indexable inserts and three-dimensionally shaped chip formers. In: CIRP Annals. 38. Jg., Nr.1, 1989, S. 75–79
Parameters with Influence on Burr Formation
Size Effects in Drilling Burr Formation R. Neugebauer, G. Schmidt, and M. Dix
Abstract Despite the existence of numerous competing technologies, drilling is the most common and still highly demanded manufacturing process in production industries. Mostly, the manufacturing of a through bore-hole is attended by the formation of an entrance and an exit burr. Thereby, the dimension of the exit burr exceeds the entrance burr considerably. Due to the hangover on the drill exit surface, the quality of the workpiece is reduced significantly. For this reason, the removal of the burr is often required. But in case of micro-drilling no appropriate tools are available. In addition, the manual removal of the burr, which is often used for conventional drill sizes, is impossible for such small geometries. An efficient method for solving the burr problem in conventional machining is the reduction of the burr size to a tolerable size level by adapting cutting parameters, tool properties or workpiece properties. Numerous investigations have been presented on this field in the last years, but the reached expertise can’t be transferred to micro-machining. Based on the rising influence of the grain structure in micromachining, certain size effects occur when scaling the cutting process from the macroscopic to the micro level. As soon as the drill diameter is smaller than 1000 μm the feed per tooth and the cutting tool geometry comes into the order of magnitude of the grain structure. Thus, the characteristic of the workpiece material switches from homogeneous to anisotropic. Since there only exist investigations to a few discrete diameter ranges, the resulting influences on the burr formation are widely unknown. In this paper the size effects of the drilling burr formation are investigated in systematic scaled drilling series over a wide range of drilling diameters (Ø 0.01 mm–14 mm).
R. Neugebauer, G. Schmidt Fraunhofer Institute for Machine Tools and Forming Technology, Chemnitz, Germany M. Dix () Institute for Machine Tools and Production Processes, Chemnitz, Germany e-mail:
[email protected] url: www.tu-chemnitz.de/mb/iwp/
Keywords Burr formation · Size effect · Drilling · Micro machining
1 Introduction The DIN ISO 13715 standardized the burr as a “material overhang outside of the ideal geometrical shape of the workpiece edge, which remains after the machining. . .”. This overhang reduces the quality of the workpiece significantly. The reduction bases on the geometrical error and also on the undefined stiffness of the burr material. It is strongly deformed and often it holds micro cracks. Another negative attribute of some burrs is the relatively small bonding zone to the rest of the workpiece. Thus, it could happen that burrs get lost during the use of the workpiece and this could cause damages, for example by closing an oil channel. In the past many elementary investigations had been done to the so called orthogonal cutting. There different burr shapes, formation mechanism and also options for the burr minimization were studied [1–7]. However, the orthogonal cut is a two-dimensional burr formation process. Its effects and attributes are only partially valid for three-dimensional chip and burr formations, like drilling. Drilling is the most often used machining operation. It is featured by a decreasing cutting speed with the decreasing of the drill radius. In the area of the centric chisel edge the cutting speed is zero and the workpiece material is displaced by extrusion to the outer main cutting edges. Thus, there exist two different material removal mechanisms in the drilling process. Additional to that the geometry of the cutting edge changes therefore over the radius. The result is a chip and burr formation process which is strongly three-dimensional. Min et al. [8] studied and classified the different drilling burr types for the conventional size level. He divided the burr types into ring burrs with different burr heights, crown burrs, ring burrs with burr cap and wing burrs (Fig. 1).
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_13, © Springer-Verlag Berlin Heidelberg 2010
117
118
R. Neugebauer et al.
Ring burr hb < 0,15mm
Ring burr hb < 0,15mm
Crown burr hb < 0,15mm
Ring burr with cap hb < 0,15mm
Wing burr hb < 0,15mm
Fig. 1 Burr types [8]
Burr type 1 and 2, the so called ring burrs, were relatively flat and uniform burrs with a relatively strong adherence to the rest of the workpiece. Its low burr height and high strength made this kind of burr unproblematic for the manufacturing. A small ring burr is mostly the goal of the burr minimization. Type 4 also was a ring burr, but a burr cap adhered locally. This local bonding was very low, what risk a later lost during the use of the workpiece. Hence, a removal of it is required. The same needs for removal count for the crown burr and wing burr. The burr wings could have been as high as the half of the drill diameter. Therefore, the resulting geometrical error was very large and additional to that the adhering zone was also very small and fragile [8]. Based on the analogies between the wing and the crown burr they are subsumed to the crown burr in the ensuing investigations. A trend to the minimization of parts and components and with it the micro manufacturing accrued in the last decade. The micro machining is distinguished by a large removal rate and good geometrical accuracy. However, some size effects appear by the downscaling of the conventional macroscopic machining to the micro size level. They are the actual focus of experimental and numerical investigations in different research institutes. There, the increase of the specific cutting force by the decreasing of the cutting depth and the appearing of the ploughing effect, were the most significant effects. The ploughing effects occurred when the ratio λfz/rβ between uncut chip thickness and cutting edges radius is in the range of 1 or lower [9–12]. In case of a scaled machining process the uncut chip thickness is scaled also. But for the cutting edge radius a technical minimum exist, which is in the range of the demanded uncut chip thickness. Thus the ratio λfz/rβ could not been scaled. Different investigations of micro burr formation showed an increase of the burr geometries by the downscaling [13, 14]. Additional to that the absence of deburring tools and removing methods enforced the burr problem in micro machining. The manual deburring, which is often used in the conventional size level, is impossible for the micro level. The minimization of the burr by the adaption of cutting parameters or tool geometries was an economical instru-
ment to solve the burr problem in the macroscopic size level. A transformation of these cognitions to the micro machining is actually not possible because with the scaling some unknown size effects accrue and no overall size investigations to the drilling burr formation exist. This work will be a first approach to close this gap in knowledge.
2 Drilling Test In the literature the machining is divided into conventional and micro machining. The word micro relates to the unit of length, what derives the question over the boarder between micro and macro for drilling. In this investigation the border was positioned by d ≤ 1 mm for the micro drilling. Many conventional macroscopic drill series contained the d = 1 mm but for this drill size the feed per tooth fz was in the range of fz = 16 μm (HSS-E standard drill). The material AISI 1045 has an average grain size dgrain ≈ 10 μm under normalized conditions. Thus, there must be influences of the grain structure because the material is no isotropic continuum anymore. Based on this division two different experimental rigs were developed for the tests. These two setups did not differ in the basic principle of the drilling but the used measurement and machine equipment had to be adapted to the size level.
2.1 Experimental Setup of Macroscopic Test The experimental setup for the macroscopic drilling test was installed in a conventional machining centre (Fig. 2). The goal of the setup was to measure the cutting forces and to record the deformation mechanism and to map the thermal distribution on the drill exit surface during the burr formation. The components of the cutting force, feed force Fz and torque M, were determined by a dynamometer
Size Effects in Drilling Burr Formation
119
were smaller than 14 mm [15]. As aforementioned the lower bound was derived by the change to the micro level by d > 1 mm. These drills had a cone-shape grind with an axial spike.
2.2 Micro Drilling Test Setup
Fig. 2 Experimental setup of the macroscopic test
R Kistler9125A. There, the forces were measured using a special tool holder, that enabled a direct measurement of the forces and reduced the disturbance influences. The exit of the tool, which is characterized by the burr formation, was recorded by different camera systems. A high speed camera HSC (FastcamTM SA1) recorded the deformation mechanisms of the exit surface. Additionally a thermal R was used for the docuimaging system InfratecTM ImageR mentation of the thermal distribution on the exit surface during the burr formation. They made it possible to map the deformation mechanisms and thermal procedures with a high chronological and graphically resolution. Like for all other machining processes the cutting parameters were mainly determined by the used tool. Tools usable with high feed rates und fast cutting speeds were available for drilling but these tools were often coated with layer systems. However, the focus of this work was an elementary study of the burr formation which demanded tools as simple as possible. The scheduled adaption of the cutting edge radius by corundum blast also made the use of coated tools impossible. Consequently a HSS-E drill series in the range of 2 ≤ d ≤ 14 mm with identic tool angles and scaled geometries for the different diameter sizes was chosen for the macroscopic tests. Additional to these features the availability of a micro drill series with a high level of geometrical analogy confirmed this tool selection. The upper bound of the d ≤ 14 mm depended on the fact, that 95% of the annual consumption of twist drills
Investigations to size effects require the analysis of the complete presently available drill size spectrum. The influence of the grain structure increases with the decrease of the process size. Therefore, especially the range of small geometrical dimensions was crucial for these investigations. Actual drill sizes down to a diameter of d ≥ 0.05 mm were available by different tool manufacturers. Hence, the range of the micro drilling test was specified for the drill size range of 0.1 mm ≤ d ≤ 1 mm. This allowed an investigation to size effects over more than two decimal power of drill diameter. In correlation to the macroscopic drilling test uncoated HSS-E drills were used for the micro test. The important drill nose angle σ was also σ = 118◦ . Only for the drilling grind some differences existed between the macro- and microscopic tools. Those based on the limited possibilities of manufacturing for micro drills. Thus, the grind for the micro drills was a two-face shape and not a cone shape. In comparison to conventional drilling the equipment for micro drilling had some special features. Not only the requirements on accuracy increased. Identical cutting speeds in macro drilling and micro drilling demanded high revolutions speeds in the range of 40,000 rpm for the micro sizes. Such revolution speed values were not possible for conventional macroscopic machining centres. A micro machining R was used to fulfil these requirecentre Kugler Microgantry ments in accuracy and revolution speed. An additional adaption was necessary for the geometry and surface quality of the workpieces. The used tools required a maximum drill depth of three times of the drill diameter. Thus, very thin workpieces were essential. Furthermore, the drill exit area of the workpiece had to be polished to be able to measure the burr height accurate. The residual stresses, which are based on the machining of the workpiece, also had to be eliminated. This has been realized by a wet grinding and polishing using a preparation system for metallographic specimen. This guaranteed a maximum of surface quality and negligible residual stress values. In the Fig. 3 an overview of the micro drilling test stand is shown. In the rig the workpiece (1) is fixed in the workpiece R Minidyn force holder, which is screwed on a Kistler dynamometer. The drill exit area faced downward and could be monitored by a light microscope (3) via a high precision
120
R. Neugebauer et al.
mirror (2). This mirror was pivot-mounted in the workpiece holder. The used light microscope contained a LCD camera which enabled a digital recording of the micro burr formation.
The feed per tooth was scaled to the drill diameter in case of the test series with a scaled feed per tooth (TS1 and TS3). There, the ratio of drill diameter to the feed per tooth was d/fz = 60. This was common to the recommended cutting parameters from the tool manufacturer for this drill type. The test series with the constant feed per tooth (TS2 and TS4) had a feed per tooth of fz = 0.05 mm for all drill sizes. The cutting edge radius was also adjusted. In the test series with the constant cutting edge radius (TS1 and TS2) the drill was used in delivery condition. There, the radius was constant rβ = 20±5 μm for all drill sizes. In test series TS3 and TS4 the scaled cutting edge radius was modelled by corund blast. The macroscopic test was extended to the micro size level by the micro test series with the drill diameters d = (1 mm, 0.5 mm, 0.25 mm, 0.1 mm). Only a scaled feed per tooth of d/fz = 60, like for the macroscopic test series TS1 and TS3, was possible for the micro drilling. Variations of the feed per tooth away from this recommended value resulted in an early tool brake. An adaption or exact measuring of the cutting edge radius was also actual impossible for micro drill.
2.3 Test Series
3 Results
(Drill)
(2)
(1)
(3)
KISTLER®
Fig. 3 Experimental setup micro drilling
The experiments in the macroscopic range were divided into 4 test series TS for the detection of size effects. In this TS different process parameters were scaled or kept constant during the scale of the drill diameter. The cutting edge radius rβ and feed per tooth fz were the determinant process input parameters. The influence of these parameters on the burr geometries, burr formation mechanism, temperature distribution and cutting force path was tested by the variation of them. The Fig. 4 shows an overview over the test plan.
Input parameters
3.1.1 Burr Geometries The shape and the size of the burr were the main results of the burr formation process. The burrs were classified and their burr heights hb were measured as it is explicated in the Chap. 1. In the following Figs. 5–8 the burr heights and burr types are shown over the increasing drill diameter or cutting
Test
fz
rβ
TS 1
scaled
constant
TS 2
constant
constant
TS 3
scaled
scaled
TS 4
constant
scaled
Test readings Burr geometries
Visualization (HS-Video)
Fig. 4 Overview test plan
3.1 Size Effects in Macroscopic Range
Thermal distribution
Scale effects of the burr formation
Cutting Forces
Size Effects in Drilling Burr Formation
121
Burr height hb [µm]
5000 2500 0 2
4
6
8
10
12
14
Drill diameter d [mm]
Fig. 5 Burr geometries TS1 – scaled fz , constant rβ
Burr height hb [µm]
7500
ød = 6 mm
5000 2500
7500
ød = 10 mm
5000 2500
0 0 UG FK MK GK UG FK MK GK d = 14 mm
7500 5000 2500 0 UG
FK
MK
GK
Fig. 8 Burr geometries TS4 – constant fz , scaled rβ . (UG→no blasted, FG→fine corundum, MK→normal corundum, GK→big corundum)
5000
2500
0 2
4
6 8 10 Drill diameter d [mm]
12
14
12
14
Fig. 6 Burr geometries TS2 – constant fz , constant rβ
7500 Burr height hb [µm]
Burr height hb [µm]
Burr height hb [µm]
7500
7500
Burr height hb [µm]
Ring burr
Crown burr
5000
2500
0 2
4
6
8
10
Drill diameter d [mm]
Fig. 7 Burr geometries TS3 – scaled fz , scaled rβ
edge radius variation. Every shown burr height in the diagrams is based on an average of five measured burrs generated under similar conditions. The burr type must be identic in all five drills to be assigned to this type. If there different burr types for one of the five burrs were identified, no burr type could be assigned (for example in Fig. 5). The test series TS1 was equal to the practical applied drilling with this drill series. In Fig. 5 their burr heights and types are shown. The resulting burr heights showed an increase in the range 2 ≤ d ≤ 7 mm with the increasing drill
diameter. There was always a crown burr detected. In the range of 8 ≤ d ≤ 12 mm crown burrs and ring burrs were identified for one and the same drill diameter. Thus, in this range the burr formation process was classified as instable. Ring burrs were always formed for drill diameters higher than d = 12 mm. Hence, in this test series TS1 a change to the ring burr happened by increasing the drill diameter. An instable range was detected between the different burr types. In comparison to that the bore series TS2 with constant feed per tooth and constant cutting edge radius always resulted in crown burrs. The burr height increased with the increasing drill diameter. The TS3 with scaled feed per tooth and scaled cutting edge radius also formed a crown burr, which height increased with the increase of the drill diameter, all the time. Different values for cutting edges radii were tested for constant feed per tooth and drill size in the TS4 (see Fig. 8). In case of the drill diameters d = 6 mm and d = 10 mm a crown burr was always formed. A change of the burr type from ring to crown burr was detected for the drill diameter d = 14 mm. An increase of the burr height with increasing cutting edge radius was also measured for this diameter. In summary the test results show the following behaviour of the burr formation (see Fig. 9): TS1 → λfz/rβ increased → change from crown to ring burr; TS2 → λfz/rβ constant → constant crown burr; TS3 → λfz/rβ constant → constant crown burr; TS4 → λfz/rβ decreased → change from ring to crown burr for d = 14 mm.
122
R. Neugebauer et al.
3.1.2 Burr Formation Mechanism and Cutting Forces for Different Burr Types Ring burr
Instable Crown burr
0
2
4 Ratio fz /rβ
6
8
Fig. 9 Burr types in comparison to λfz/ rβ
Based on that summary it seems to be that the ratio λfz/rβ between feed per tooth and cutting edge radius has a significant influence on the resulting burr type. In TS4 no change of the burr type arose during the decrease of λfz/rβ for the drill diameter d = 6 mm and d = 10 mm. That indicates that there must have been exceeding a critical value for λfz/rβ where the burr type changed during the tests with a bore diameter of 14 mm. In Fig. 9 the burr types of all macroscopic test series are shown in comparison with its ratio λfz/rβ . The distribution of the values showed a significant influence of λfz/rβ on the resulting burr type. Thus, a ring burr was formed for ratios higher λfz/rβ > 6. A crown burr is always formed for ratios that are smaller than λfz/rβ < 3.8. Between these values an instable range, where both burr types were formed, exists. In case of the ring burr one value with a relative low λfz/rβ was detected. This was with a d = 14 mm drill with a high value for the cutting edge radius. The reason of this outlier will be investigated in future work.
The drilling burr formation was recorded to analyse the different formation mechanisms by a high speed camera (HSC). The videos were synchronised with the measured cutting force signals of the feed force Fz and the torque M. This allowed it to define the burr formation mechanism and history. In the Figs. 10 and 11 the crown and ring burr formation is shown successively with a HSC picture sequence and the cutting force signal. The maximum values of the y-axes are defined by 110% of the average value during the constant drilling. The feed force and the torque during the crown burr formation showed a continuous decrease. The HSC pictures illustrate, that the crown burr formation began with a first distortion of the exit surface in the centre area of the drill. This distortion increased and a so called burr cap was formed. Thereby, it was visible that the cutting edges formed the material in front of them and did not cut anymore. Thus, it was more a forming process than a cutting process. In the area of the chisel edge the drill broke through the surface and radial cracks were initiated into the burr cap. These cracks grew and divided the burr cap into the burr wings which were bent up by the drill, like a one side fixed beam. High deformations were inserted into the feet of the burr wings with this bending. These high deformations could be verified by metallographic specimen (see Fig. 12). Different to the crown burr formation ran the ring burr formation. At the beginning it was nearly the same as for the crown burr formation. After the distortion of the exit surface a burr cap was formed. Then the drill broke through the cap but the initiation of the radial crack was relatively slow compared with the crown burr formation and no burr wings were formed. In case of the ring burr formation the main cutting
Crown burr 1181,1
0 5,111
M
M [Nm]
3,4203
Fz [N]
Fig. 10 Force signal and HSC pictures of the crown burr formation
Fz
Time [s]
0 7,3415
Size Effects in Drilling Burr Formation
Ring burr
0 1,682
Fz
M 8,0952
M [Nm]
2706
Fz [N]
Fig. 11 Force signal and HSC pictures of the ring burr formation
123
Time [s]
0 2,426
At the beginning the cap was formed and the cutting process is superimposed by a forming one. The strength of the burr cap material increases and also the cutting part increases with the deforming of this material. Thus, that instable cutting led to higher amplitudes for the torque, which was the main part of the cutting force for drilling. The high amplitudes end with the cut off of the burr cap and the outer material is formed to the ring burr. That is a forming process with low amplitudes like the whole crown burr formation. Concluding, the ring and crown burr formation bases on different material removal mechanisms which are mirrored by differences in the torque path.
3.1.3 Thermal Distribution During Burr Formation
Fig. 12 Metallographic specimen of the burr feet
edge broke through the surface and cut off the burr cap. After this cut off the remaining outer material was deformed into the relative small ring burr. These different formation mechanisms are reflected in the force signal. The comparison of the signals of the crown and the ring burr formation displays significant differences in the torque time behaviour. There, the ring burr shows higher amplitudes in the first half of the burr formation than the crown burr. During the crown burr formation the burr cap was divided by radial cracks to the burr wings and their bent up. This was more a forming process than a cutting process. Different to this was the ring burr formation. In this case the burr cap was cut off by a break through of the main cutting edges. Therefore, the ring burr formation was a cutting process but it was no continuous cutting like the normal drilling process.
The cutting process was strongly influenced by the temperature distribution in the cutting zone. Hence, the burr formation was also affected. Deformations in the front of the cutting edge and friction in the tool-workpiece-contact caused a local heating of the material which reduced the yield strength in this region. For this investigation the thermal distribution of the surface of the workpiece, where the tool exits, was measured by a thermographic camera system with a high frequency. In Fig. 13 a picture series is shown for each burr type. This gives a short overview but could not substitute a quantitative comparison. As a quantitative comparison five measuring rings were uniformly distributed over the radius. For each ring the average temperature was determined with a measuring rate of 100 Hz during the burr formation. Diagrams for the course of the average temperature value for each ring are shown in the Fig. 14. At the beginning the average temperature increased constantly in both cases. The average temperature for the inner
124
R. Neugebauer et al.
ring started to decrease with the breakthrough of the drill in the middle of the drilling hole. After the breakthrough the temperature of the drill was measured in this region. Overall the crown burr showed a constant increase of the temperature and after the breakthrough a constant decrease. In contrast to this was the ring burr. There, the average temperature paths, especially the path for the middle and outer rings (3–5), had an abruptly decrease. In combination with the HSC-pictures it became clear that this was the cut off of the burr material. After the cut off mainly the drill temperature was measured. The temperature level for the ring burr was significantly higher than for the crown burr, which was based on the significantly larger cutting depth.
3.2 Micro Drilling Burrs in Comparison to Macroscopic Drilling Burrs 3.2.1 Burr Heights and Burr Types in Micro Drilling As described in Chap. 1 the micro machining exhibited higher specific cutting forces than the conventional
Temperature [°C]
Average temperature, Ring burr
Burr height [μm]
Fig. 13 Qualitative comparison of the thermal distribution
machining. The material characteristic also changed from isotropic in macro machining to anisotropic in micro machining. These changes must have had influences on the burr formation. Significant differences of the burr height and of the burr type during the increase of the drill diameter were measured by the TS1 in the macroscopic size level. This test series was in the range of 2 mm ≤ d ≤ 14 mm and had a scaled feed per tooth and a constant cutting edge radius. Using these parameters the drilling test series were extended to the micro level, as illustrated in Sect. 2.3. Every shown value for the burr height is based on the average of 5 values and also the burr type was classified synchronic to the macroscopic test. Only the measurement setup was different. A high solution and a relative high depth of sharpness were necessary for the graphic measurement of the burr height. Thus, a scanning electron microscope SEM was used for measuring the burr height and not a normal light microscope like for the macroscopic size level. Thereby, the following average values were measured. The results clarified that with the decreasing of the drill diameter the burr height also decreased (Fig. 15). There, the
900 800 700 600 500 400 300 200 100 0 0,1
Fig. 15 Burr heights in micro drilling
Average temperature, Crown burr 450
400
400
350
350
300
300
250
250
200
200 150 Time [s]
Fig. 14 Quantitative comparison of the thermal distribution
0,5
Drill diameter d [mm]
450
150
0,25
Time [s]
1
Size Effects in Drilling Burr Formation
125
Absolute burr height
Relative burr height
d = 0.5 mm
d = 0.25 mm
d = 0.1 mm
1,4
5000
1,2 4000
1
3000
0,8 0,6
2000
0,4 1000
Relative burr height [mm/mm]
Absolute burr height [µm]
d = 1 mm 1,6
6000
0,2
0
0 0
1
2
3
4
5
6
7
8
9
10 11 12 13 14
Drill diameter [mm]
Fig. 16 Burr heights of the full drill diameter range Fig. 17 Burr shapes in micro drilling
measured heights were shown in absolute values. In Fig. 16 the burr heights for the full drill diameter range are illustrated absolute and relative to the burr height. The path of the absolute burr height showed an increase of the burr height in the range from 0.1 mm ≤ d ≤ 7 mm. Afterwards, the burr type changed to ring burr what resulted in a decrease of the burr height. This change was combined with an instable drill diameter range (0.1 mm ≤ d ≤ 7 mm) where both burr types existed. Oppositional to this is the path of the relative burr height (hb,r =hb /d). There, the highest values were measured in the micro drilling range. The relative burr height decreased exponential with an increase of the drill diameter for d ≤ 5 mm. This course had a high analogy with the course of the specific cutting energy [12] for these diameters. The range 5 mm > d ≥ 12 mm showed any local maxima and minima which was based on the burr type change in the instable range. Another decrease of the relative burr height appeared for drill diameters higher than d > 12 mm. An approximation of the relative burr height to zero for a higher drill diameter could be supposed. Additional to the burr height major changes for the burr type were detected in the micro drilling range. In the SEM pictures (Fig. 17) a change of the burr shape was visible. Thus, the burr type converted from crown burr to ring burr with a cap during the decrease of the drill diameter. The burr of the drill diameter d = 1 mm was a high crown burr with two or more burr wings. With the decrease of drill diameter the burr wings became bigger. A ring burr with an adhesive complete burr cap was detected for the drill diameter d = 0.1 mm. Thus, there were two changes of the burr type for the full drill diameter range, one in the microscopic and one in the macroscopic range. Additionally the relative burr height increased with the decrease of the drill diameter in the whole range. This enlarged the burr problem for the micro machining.
Fig. 18 SEM pictures of a micro drill d = 0.25 mm
Investigations of the influence of the ratio λfz/rβ on the burr formation for micro drilling were actual not possible because the measurement of the cutting edge radius of micro drills was not technically feasible. SEM pictures of the micro drills showed a cutting edge radius of approximately rβ = 1–2 μm (Fig. 18). This value was only an optical estimation and not a measurement value.
3.2.2 Micro Burr Formation Additional to the investigation of the burr formation mechanism with the HSC in the macroscopic range the micro burr formation was recorded by a light microscope with a LCD camera. The picture capture frequency was with 25 Hz relatively slow but it enabled 5 to 6 pictures over the burr formation time. This offered an overview on the formation mechanism. In the following Fig. 19 a picture series of the drilling burr formation for a d = 0.25 mm is shown. The picture sequence shows that the micro burr formation had high analogies to the macroscopic burr formation in the first steps. First, the surface in the centre of the drill was deformed. Then a burr cap was generated by the pushing of the drill. Oppositional to the macroscopic burr formation the first break through of the drill was in the outer radius
126
R. Neugebauer et al.
Fig. 19 Burr formation mechanism for d = 0.25 mm
Burr cap
and not in the centre of the drill. Thus, the frame of the cap was partially cut off and had only a small adherent area to the workpiece. The bent off cap had no break trough in the chisel edge area (see Fig. 17). This resulted in a big value for the burr height. Additional to the adhering cap a relative high ring burr was formed. This big ring burr is well shown in Fig. 17.
4 Conclusion and Outlook Size effects occurred with the scaling of conventional machining operations to the micro size level. An increase of the specific cutting force and the under-run of the minimum uncut chip thickness were some examples for size effects [9–12]. These size effects also influenced the burr formation process at the end of the cutting process. Thus, an influence of the ratio λfz/rβ between feed per tooth and cutting edge radius on the resulting burr type was shown. This ratio also mainly controlled the minimum uncut chip thickness. Analyses of the formation mechanisms, force paths and thermal distributions during burr formation clarified that the different burr types are based on different material removal mechanisms. Test series over the range of 0.1 mm ≤ d ≤ 14 mm showed an exponential increase of the relative burr height and an additional burr type change during the scaling of the conventional process to micro drilling. Its course had high analogies to the course of the specific cutting energy in this range. Over the full drill diameter range a change of the burr type from ring burr with adhering burr cap to crown burr and an additional change from crown burr to small ring burr occur for the test series with scaled feed per tooth.
This investigation clarifies that with minimization of the drilling process some size effects for the burr formation appear. The relative large burrs in micro machining produce a quality problem. It is harder to solve the burr removing or minimization for micro drilling than for conventional machining because no deburring tools exist and the possibilities for adaption of the cutting parameters are significantly smaller. Acknowledgements The authors acknowledge the DFG for supporting this work carried out within the framework of the Priority Program 1138 “Modelling of size effects in production processes” and want to thank the members of the framework consortium for their discussions and feedback.
References 1. Stoll, A., Leopold, J., Influence of Cutting Speed and Feed on Burr Formation, 6th International Conference High Speed Machining, San Sebastian, March, 2007. 2. Freitag, A., Sohrmann, C., Leopold, J., Simulation of the Burr Formation, 8th CIRP International Workshop on Modeling and Machining Operations, Proceedings, 641–650, 2005. 3. Schäfer, F., Entgraten in der Fertigungstechnik, Grundlagen des Entgratens, wt-Z. ind. Fertigung Volume 63, 692–696, 1973. 4. Severt, W., Gratminimale Auslegung von Zerspanprozessen auf Basis rechnergestützter Datenauswertung und FEM-Simulation, Dissertation, RTWH Aachen, Shaker-Verlag, 1997. 5. Pekelharing, A. J., The Exit Failure in Interrupted Cutting, Annals of the CIRP, Volume 27(1), 5–10, 1978. 6. Hashimura, M., Chang, Y. P., Dornfeld, D., Analysis of Burr Formation Mechanism in Orthogonal Cutting, Journal of Manufacturing Science and Engineering, Volume 121, S. 1–7, 1999. 7. Link, R., Gratbildung und Strategien zur Gratreduzierung bei der Zerspanung mit geometrisch bestimmter Schneide, Dissertation, RTWH Aachen, 1992. 8. Min, S., Kim, J., Dornfeld, D., Development of a Drilling Burr Control Chart for Low Alloy Steel, AISI 4118, Journal of Material Processing Technology, Volume 113, 4–9, 2001.
Size Effects in Drilling Burr Formation 9. Kotschenreuther, J., Empirische Erweiterung von Modellen der Makrozerspanung auf den Bereich der Mikrozerspanung, Dissertation Universität Karlsruhe (TH), Forschungsberichte aus dem wbk Institut für Produktionstechnik, Volume 141, 2008. 10. Weinert, K., Kahnis, P., Untersuchung von Größeneinflüssen auf den Mikro-Fräsprozess, Prozessskalierung im Rahmen des SPP 1138 “Modellierung von Größeneinflüssen bei Fertigungsprozessen”, Hrsg. F. Vollertsen, Strahltechnik, Volume 27, BIAS Verlag, ISBN: 3-933762-17-0, 2005. 11. Kahnis, P., Weinert, K., Analysis of Tool Influence on Downscaled Milling Process, Proceedings of the 2nd International Conference on New Forming Technology, Bremen, BIAS-Verlag, 481–489, 2007.
127 12. Abiouridouane, M., Skalierungseffekte beim Mikrobohren mit Wendelbohrern in Stahl, Arbeitskreistreffen des SPP1138 “Modellierung von Größeneinflüssen bei Fertigungsprozessen”, Chemnitz, Germany, 2008. 13. Filiz, S., Conley, C. M., Wasserman, M., Ozdganlar, O. B., An Experimental Invastigation of Mirco-machinability of Copper 101 using tungsten Carbide micro-endmills, International Journal of Machine Tools and Manufacture, Band 47, Volume 7(8), S. 1088–1100, 2006. 14. Lee, K., Dornfeld, D., Micro-burr Formation and Minimization through process control, Precision Engineering, Volume 29(2), S. 246–252, 2005. 15. Hoff, M., Analyse und Optimierung des Bohrprozesses, Dissertation, RWTH Aachen, 1986.
Burr Formation and Surface Characteristics in Micro-End Milling of Titanium Alloys G.M. Schueler, J. Engmann, T. Marx, R. Haberland, and J.C. Aurich
Abstract Titanium base alloys are used for aircraft structures, turbine blades and medical implants. The surface of these parts can be functionalized by micro-structuring e.g. to get a more energy efficient turbine or to get a better biocompatibility for medical implants. Micro-end milling with end mills down to 300 μm is already efficient and reliable for microstructuring ductile materials. We developed and manufactured carbide micro end mills with a smallest diameter of 7 μm to improve the spectrum of functional structures. This paper describes micro end milling experiments of the difficult to machine titanium alloys Ti-6Al-4V and Ti-6Al7Nb with 48 μm micro-end mills. Large areas are machined to observe the micro structure on the surface and the influence on surface quality. The burr formations in slot milling is observed and compared to still existing models. Up milling and down milling at the sidewalls were compared. Keywords Micro-end milling · Micro-end mill · Burr · Burr formations · Titanium alloy
1 Introduction Titanium base alloys are mostly used for aerospace application both for aircraft structures and turbine blades, because of high strength and low weight and especially for its corrosion resistance. For medical applications titanium alloys are often used because of the corrosion resistance and biocompatibility. Disadvantages are the difficult machinability and the high prize [1, 2]. Micro-milling is still increasing as a reliable manufacturing process. Main advantages are high material
G.M. Schueler (), J. Engmann, T. Marx, R. Haberland, J.C. Aurich Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany e-mail:
[email protected] url: www.fbk-kl.de
removal rates and a wide range of processable materials [3]. This predestinates micro-milling for micro-structuring titanium-alloys. Micro-milling of titanium for medical applications especially for implants can create special shaped surfaces, which can improve biocompatibility [4]. Two titanium alloys Ti-6Al-7Nb and Ti-6Al-4V were machined with ultra small micro, end mills with a shank diameter of 48 μm, in order to characterize the performance, to improve the surface quality and to minimize burrs on the workpieces. The surfaces on slot ground and sidewalls after micro-end milling were characterized. A second focus is set on the burr formations on the slots and the generation of the burrs through micro-milling. A tool breakage test gives information about the chip appearance and the surface generation.
2 Micro-Cutting and Burr Classification In micro-milling similar effects on the work results emerge compared to conventional machining. In that case, insufficient surface quality, shape failures, poor edge finish, and burrs are resulting. These problems are more significant in micromachining because of the size effects. E.g. the cutting edge radius is very big compared to the uncut chip thickness. The surface roughness and the burr formations are big compared to the structure sizes [5]. Although deburring and surface finishing are limited due to expensive techniques or the resulting damages on the microstructures.
2.1 Micro-Cutting In micromachining the cutting edges are often bigger than the chip thickness compared to conventional cutting. Microcutting is influenced by the ratio of the depth of cut to the effective cutting edge radius of the tool [6]. The effective cutting edge radius causes a dramatic change of the effective rake angle, illustrated in Fig. 1. As a result of the changed cutting conditions the cutting forces increase and
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_14, © Springer-Verlag Berlin Heidelberg 2010
129
130
G.M. Schueler et al.
Fig. 1 Change of effective rake angle through micro-cutting operations
the surface is more elastically deformed than before. Due to elastic recovery ploughing of the cutting edge on the surface occurs, resulting in surface defects and poor roughness. In that case the tool wear increases and the cutting edge gets blunt, the ploughing effect is intensified [7].
2.2 Burrs in Micro-Milling Different burr descriptions depending on application, manufacturing process, shape, formation mechanism and material properties still exist. The following description focuses of burr in milling and micro-milling. The general burr description is based on cause of formation [8]. Four types of machining burrs were detected: Poisson burr, rollover burr, tear burr and cut-off burr, Fig. 2. The Poisson burr is a result of the material’s tendency to bulge to the sides when it is compressed until permanent plastic deformation occurs [9]. The rollover burr is essentially a chip which is bent rather than sheared resulting in a comparatively large burr. This type of burr is also known as an exit burr because it is usually formed at the end of a cut. The tear burr is the result of material tearing loose from the workpiece rather than shearing clearly. It is similar to the burr formed in punching operations. The cut-off burr is the result of workpiece separation from the raw material before the separation cut is finished [8]. In milling the type of burr highly dependent on the in-plane exit angle, explained by Chern [10]. Five types of burrs were observed illustrated in Fig. 3: (a) the knife-type burr; (b) the curl-type burr; (c) the wave-type burr; (d) the edge breakout; and (e) the secondary burr.
Fig. 2 Schematics of poisson, tear and rollover burr [8]
In face milling burrs were classified by Hashimura et al. [11] according to burr locations, burr shapes and burr formation mechanisms. The burr attached to the surface machined by the minor edge of the tool is name exit burr. A side burr is defined as a burr attached to the transition surface machined by the major edge and a top burr is defined as a burr attached to the top surface of the workpiece. The burr-types are shown in Fig. 4. Weinert et al. [12] described the resulting surface features due to micro-milling of the TiNi shape memory alloy. The ductility of the material causes a poor surface quality of the finished parts, as seen in Fig. 5. Different types of wavy-burrs occur on the edges (B). The slot base and the slot side wall show a rough surface with adhered particles (A).
Burr Formation and Surface Characteristics in Micro-End Milling of Titanium Alloys
131
Fig. 3 Five types of burrs observed in face milling [10]
3 Micro-End Mills First step is the development and fabrication of micro-end mills. Carbide microtools have been fabricated by micro grinding. With this method micro-tools down to 7 μm in diameter can be fabricated with high aspect ratios. Grinding delivers a high accuracy of the tool shape and sharp cutting edges. After the micro-tool fabrication, they can be used for micro-end milling of a wide range of ductile workpiece materials. In Fig. 6 a 10 μm micro-end mill is shown. With this tool the instituts brand is milled in a human hair.
3.1 Micro-End Mill Properties Fig. 4 Types of milling burrs
Micro-grain tungsten carbide micro-end mills with diameters ranging between 7 μm and 100 μm were developed and manufactured. Because of the high chemical wear during the machining of titanium diamond tools are not suitable. For these tools solid tungsten carbide shanks with a diameter of 3.175 mm are used as tool carriers with a high bending strength of 4000 N/mm2 and a hardness of 1920 HV 30. This
Fig. 5 Surface features of micro-end milled NiTi alloy from Weinert et al. [12]
Fig. 6 Left: 10 μm micro-end mill; right: machined human hair
132
G.M. Schueler et al.
enables the manufacturing of microtools with very low diameters, high aspect ratios and long tool life. A coating is not applied to keep the cutting edge as sharp as possible. The end mills have only a single flute to get a positive wedge angle and a clearance angle of 6◦ . A high sharpness of the cutting edge with a radius below 0.1 μm and an accurate corner are reached.
3.2 Manufacturing of Micro-End Mills The manufacturing of the ultra small micro-end mills is done on a self developed micro-grinding machine specialized for an economic tool production time of only 10 min. The CNC-controlled machine is built up on a solid granite base with vibration damping. In Fig. 7 the working space of the machine is shown with the precision clamped micro-end mill on the rotational axis run on an air bearing. Two grinding spindles are mounted on the horizontal X-Y-axes. The axes have a resolution of 0.1 μm and a repeating positioning accuracy of 1 μm. One spindle is used for pre-grinding and a second self developed air bearing spindle is used for the finishing of the micro-end mill. For grinding thin diamond grinding wheels with a diamond grit of 1 μm were used. Small grain sizes and a low thickness of the grinding wheel provide low cutting forces. So very small tool diameters of only 7 μm are possible to manufacture [13]. The presented grinding machine gives the possibility to vary the shapes of the cutting tool to improve the efficiency, reliability, surfaces roughness, edge sharpness and to minimize burrs.
4 Titanium Alloys Ti-6Al-7Nb and Ti-6Al-4V Titanium alloys are used in different industries like aeronautical, medical and chemical plant engineering. The high strength and low density and the high corrosion resistance are the most common advantages of this kind of alloys. The titanium alloy Ti-6Al-4V is widely used for aircraft industry. Ti-6Al-7Nb is mostly used for medical implants science based of the better biocompatibility of the Niobium than the Vanadium [14]. Both are α+β-alloys which are classified as difficult to machine materials [15]. Table 1 represents different properties of both alloys. Other important physical characteristics are a remarkable corrosion resistance, a high heat resistance and high strength. Some properties influence the machining process. In fact, especially the low thermal conductivity is responsible for difficult machinability [1]. The comparatively low coefficient of elasticity causes an elastic deformation while chip formation. In the contact zone, this leads to a high flank wear which could be minimized by an upper clearance angle. The ductile behavior and the low coefficient of elasticity require a positive chip angle. While machining, it is recommended to use low cutting speeds and low feed per tooth because of the low thermal conductivity. In general, down milling achieves better machining results compared to up milling whereas brittle materials are also machined by up milling [2]. Some experimental and scientific results propose the following process optimizations: • Smaller cutting edge radius for better disconnection of material • Smaller cutting radius for better finishing • Higher cutting speeds and • Lower feeds without falling below the minimal chip thickness [16] In case of considering these advices metal materials achieve a better surface-quality due to micro-milling. The quality of workpiece boundary makes also demands to the milling result because the chipping leads into strong variances in the physical characteristics of the material surface. During the research of the boundary, the variances of the structural conditions and hardness as well as the conditions of residual stress have to be observed. The modification of structural conditions represents a restriction in machining ductile and difficult to machine materials. Research results
Fig. 7 Precision grinding machine
Table 1 Mechanical properties of Ti-6Al-4V and Ti-6Al-7Nb [1] Alloy Density Young’s modulus Tensile strength yield Ti-6Al-4V Ti-6Al-7Nb
4.42 g/mm3 4.52 g/mm3
114 GPa 120 GPa
865 MPa 800 MPa
Tensile strength ultimate
Hardness
930 MPa 900 MPa
HV 392 HV 365
Burr Formation and Surface Characteristics in Micro-End Milling of Titanium Alloys
allocate, that lower feeds execute proportionate more work in chipping. In the theory this kind of observation is untended. Furthermore, different researches analyzed, that regarding the cutting forces, only the force to press the material in the surface which depends mostly on the very little relation between chip thickness and cuttings edge radius accomplished the chipping. A significant variation of the boundary could not be analyzed by using different feeds and axial depth of cut. In fact, a remarkable deformation of the boundary results from using small cutting edge radius and an increasing cutting speeds which could be explained by the increasing chipping temperature and the better material deformation. In comparisons to steel, increasing cutting speeds leads to chip compression which could be minimized by choosing a cutting fluid. The manageable models of boundary do not picture all the phenomena of chipping. In regarding the micro-machining, an explanation of variance in the quality of boundary is not given by literature. Also the influence of residual stresses could not be explained scientifically for micro-milling.
133
Fig. 9 Manufactured 48 μm micro-end mill; left side: Overview; right side: Detailed view of the cutting edge and corner
of 60,000 rpm. By the use of a 48 μm end mill only a cutting speed of 9.05 m/min is reached. The value is below the recommended cutting speed for titanium of 40–70 m/min in that case a variation of cutting speed is passed [2]. All tests were done in dry machining. To characterize the surfaces and the burrs, the slot-base and the side walls were analyzed by SEM and a white-light interferometer.
5 Experimental Set-Up
6 Results
The micro-end mills are used for machining Ti-6Al-7Nb and Ti-6Al-4V in a self developed 3-axis precision-milling machine, shown in (Fig. 8, right) [17]. This desktop sized machine tool is based on a granite base with additional vibration damping. The machine has a resolution of 20 nm and a repeating positioning accuracy of 1 μm. A vertically mounted air bearing spindle provides 60,000 rpm. The tools are clamped in the spindle in an adjustable clamping device. The run out is aligned manually by monitoring with a long distance microscope. For further experiments the machine is equipped with a dynamometer Kistler Minidyn 9256C2 for the measurement of the cutting forces (Fig. 8, left). For all experiments a microtool with 48 μm in diameter is used, shown in Fig. 9. The spindle speed is set to a maximum
For the characterization of micro-end milling of titanium alloys five tests were performed. With these experiments the fundamental behavior of micro tool in titanium alloy was observed, with a focus on burrs and surface quality.
Fig. 8 Right: Milling machine; left: Work space with dynamometer and workpiece clamping
6.1 Surface Characteristics of Large Area Machining In the first test series a surface is pre-cut to create a plain surface on the workpiece by machining 14 parallel lines. Then the tool was changed and 3 different feed rates were tested for the fine machining of the workpiece surface. The schema the slot-end milling is shown in Fig. 10.
134
G.M. Schueler et al.
Fig. 12 SEM pictures of the slot surface of Ti-6Al-4V Fig. 10 Schema of large area of slot-end milling
In Table 2 the cutting parameters and experimental conditions are summarized. A low feed rate of 0.084 μm/tooth was chosen corresponding to the cutting edge radius. The SEM pictures of Ti-6Al-7Nb in Fig. 11 and of Ti-6Al-4V in Fig. 12 show that different feed rates did not result in different surface roughness. The results are very similar for both titanium alloys. Explained by Weinert et al. [12] and Schaller et al. [18] material particles adhere on the
Table 2 Experimental conditions for large area machining Tool diameter Cutting speed vc Feed per tooth fz Axial depth of cut ap
48 μm 9.05 m/min at 60,000 rpm 0.084; 0.168; 0.252 μm/tooth 2 μm
Fig. 11 SEM pictures of the milled surface of Ti-6Al-7Nb
slot ground surface. The adhesion of particles is only placed at the side of the tool path. This effect is more distinct Ti6Al-7Nb. Comparing the surfaces milled by different feed rates, all tool paths have the same shaped regular scratches generated through the circular motion of cutting edge. It is observed that the distance between two scratches is always equal, independent of the feed rate. The average roughness of the bottom of the slot is Ra 100 nm, shown in Fig. 13 for both alloys. By applying only 2 μm depth of cut massive burrs formations between the pre-cut and the fine-cut area occurred. The edge is placed on the up milling side oft the tool path. This burr formation is classified as wave-type burr, explained in [10]. It seems that the whole material is plastically deformed and pushed aside, without cutting. Because of the very high burr formation, a different possible cause is that the burr consists of accumulated small burrs resulting of the cutting
Fig. 13 Slot surface roughness
Burr Formation and Surface Characteristics in Micro-End Milling of Titanium Alloys
135
Fig. 15 Exit and entrance burrs by milling of Ti-6Al-7Nb axial depth of cut 20 μm Fig. 14 Wave-burr formation of the slot surface of Ti-6Al-4 V
of the different path before. This is shown in Fig. 14. The top of the burr looks very similar to the adhesive particles of the slot ground, while the bottom side of the burr has a typical wave form of plastically deformed material, like described in [10].
6.2 Burrs on Microslots To get information of the resulting burr on the slots grooves were milled into the workpiece by a single step. The chosen cutting parameters are shown in Table 3. SEM pictures of the slot entrance and slot exit of an exemplary groove in Ti-6Al-7Nb is presented in Fig. 15. All kind of burrs like explained of Hashimura et al. [11] were observed. On both top sides a similar wave burr formation occurred, independent of down milling or up milling side. The top burr formations are about 15 μm high. A reason for the top burr formations in hardened steel is explained of Bissacco et al. [19]. He explained that top burrs are large in micromilling due to the size effect. Very high biaxial compressive stress pushes material toward the free surface and generates large top burrs. Typical entrance and exit burrs emerged at the slot bottom. At the groove exit big exit side burrs were noticed. On the slot entrance side exceptional entrance side burrs occurred on this differs to the classification of Hashimura et al. [11]. A possible explanation is given in Sect. 4.5. Overall the values of the entrance and the exit burrs are equal, while the exit side burrs are significant bigger than the entrance side burrs.
6.3 Up Milling and Down Milling Effects on Side Walls in Microslots The side walls of the same slot like seen in Sect. 4.3 were observed. In up milling the cutting direction of the end mill is against the feed direction oft the workpiece. In down milling the cutting direction has the same direction of the feed. It is fundamental, that the surface in up milling is generated in at the beginning of the cut. In down milling the surface is generated at the end of the cut [20]. This results in different surface characteristic on the groove side walls. In Fig. 16 SEM pictures of both slot sides are shown. On the right picture the micro-end mill generated the surface is up milling action, on the left side the surface is generated due to down milling. The down milled surface is smooth, while the up milled surface is very rough and fissured. The fissured particles are oriented against the cutting edge which moved over the side wall surface. Like conventional milling the surface generated through down milling is better than the up milled surface [20]. An explanation for the fissured up milled surface is given in Sect. 4.5. As seen the milling method has no influence on the top burr formation.
Table 3 Experimental conditions for slot milling of Ti-6Al-7Nb Tool diameter Cutting speed vc Feed per tooth fz Axial depth of cut ap
48 μm 9.05 m/min at 60,000 rpm 0.084 μm/tooth 20 μm
Fig. 16 Slot milling: Up milling and down milling of Ti-6Al-7Nb; axial depth of cut 40 μm
136
G.M. Schueler et al. Table 4 Experimental conditions for side wall machining of Ti-6Al-4 V Tool diameter Cutting speed vc Feed per tooth fz Axial depth of cut ap Radial depth of cut
48 μm 9.05 m/min at 60,000 rpm 0.084 μm/tooth 40 μm 2 μm
6.4 Up Milling and Down Milling of Side Walls In the test series before different side wall surfaces were observed. If complex microparts like implants are machined, not only the surface generated of the tool face is important, although the side walls are important. In that case the surfaces of side walls were analyzed. In this test series the up and down milling on the side walls is realized in order to characterize the surface structure. The focus of those experiments is the generation of vertical side walls, the bottom surface is not considered. For a 48 μm micro-end mill the axial depth cut is 40 μm and the radial depth of cut is of 2 μm. The applied cutting parameters are summarized in Table 4. In Fig. 17 SEM pictures of the up milled and down milled side walls of Ti-6Al-4 V are presented. Down milling results in plain surfaces, while up-milling generates very rough surfaces with fissures. The results are equal to the slot milling results in Sect. 6.3. Similar results were observed in Min et al. [21]. Reasons for the fissured up milled surface are given in Sect. 6.5.
6.5 Chip Formation and Sidewall Generation In micro-end milling the generation of chips is important to understand the limitations on the ongoing miniaturization of micro end mills and its application. To get some information about the chip formation the tool breakage of 48 μm single fluted micro-end mills is caused. The axial depth of cut is set on 20 μm. In Table 5
Fig. 17 Side walls: Up milling and down milling of Ti–6Al–4V, axial depth of cut 40 μm
Table 5 Experimental conditions for tool breakage test in Ti-6Al-7Nb Tool diameter Cutting speed vc Feed per tooth fz Axial depth of cut ap
48 μm 9.05 m/min at 60,000 rpm 0.084–0.57 μm/tooth 20 μm
the furthermore cutting parameters are summarized. In this experiment the feed rate is increased linear while milling the grooves. The tool broke in both tests at a feed rate of 0.57 μm/tooth. In Figs. 18 and 19 the results of the tool brakeage in Ti-6Al7Nb are shown. In Fig. 18 the broken tool still remains in the groove. It is difficult to measure the cutting edge radius, but the results of Fig. 18 shows that the cutting edge radius is lower than 0.1 μm. On the very sharp cutting edge of the tool a segmented chip formation is seen. The measurement of the segmented chips shows that the chip thickness is about 0.6 μm. That compares equal to feed rate when the tool breakage occurs. Noticeable is the big amount of chips in the milled groove, as seen in Figs. 18 and 19. Figure 19 gives valuable information about the surface generation of the up milled slot side. The tool broke at point (A) some chips are still adhered at this point. The presence of these chips suggests that they are attached on the cutting edge and that they were rotating with tool. In point (B) the chips were reattached on the side wall due to squeezing and plastical deformations. This could be the reason, why the fissures on the side wall aimed in the same direction, collinear to the chips as see in point (B). Due to inspection of the SEM pictures, it is unclear, on which point of impact the wave burr formation emerges.
7 Conclusions and Outlook Two Titanium-alloys (Ti-6Al-7Nb and Ti-6Al-4V) have been successfully machined with 48 μm diameter micro-end mills.
Burr Formation and Surface Characteristics in Micro-End Milling of Titanium Alloys
137
Fig. 18 Chip formations and a broken end mill in a 20 μm slot
Further experiments have to be carried out with ultra highspeed spindles to provide higher cutting speeds. This is necessary to get more common cutting conditions. In that case a newly developed ultra high speed spindle will be used for the micro-tools (diameter down to 10 μm). Acknowledgments This research was funded by the German Research Foundation (DFG). We gratefully acknowledge for financially supporting within the DFG Research Training Group 814:“Engineering Materials at Multiple Scales: Experiments, Modeling and Simulation”
References
Fig. 19 Chip adhesion on the slot side walls
The slot bottom surface has been characterized. An average surface roughness of Ra 100 nm is achieved in both alloys. The roughness did not vary significantly due to feed variation. Massive wavy burr formations were observed by applying a low depth of cut. In a second series the burrs of micro-end milled grooves were observed. The results agree to former studies. In the micro-milled slots differences on the side walls were observed. Down milling generates smooth surfaces, while up milling causes very poor surfaces. Structured side walls have to be machined or finished by down milling tool paths. A tool breakage test gives information about the chip formation and reasons for the poor surfaces generated due to up milling. A new approach will be the test of different shaped microend mills and the influence of the modified tools on surfaces and burrs.
1. Peters, M., Leyens, C., 2003, Titanium and Titanium Alloys. Fundamentals and Applications, Wiley-VCH, Weinheim 2. Kisselbach, A., Stühmeier, W., 1994, Titanlegierungen wirtschaftlich zerspanen, Spanende Fertigung, Weinert, K. (Hrsg.), Essen 3. Dornfeld, D., Lee, D., 2008, Precision Manufacturing, Springer, New York 4. Park, J.B., Kim, Y.K., 2000, Metallic Biometerials, The Biomedical Engineering Handbook, 2nd ed., CRC Press, Boca Raton, FL 5. Dornfeld D., Min S., Takeuchi Y., 2001, Recent Advances in Mechanical Micromachining, CIRP Annals – Manufacturing Technology 55(2):745–768. 6. Weule, H., Hunrup, V., Trischler, H., 2001, Micro-Cutting of Steel to Meet New Requirements in Miniaturization, CIRP Annals – Manufacturing Technology 50(1):61–64. 7. Ikawa, N., Shimada, S., Tanaka, H., 1992, Minimum Thickness of Cut in Micromachining, Nanotechnology 3(1):6–9 8. Gillespie, L.K., Blotter, P.T., 1976, The Formation and Properties of Machining Burrs, Transactions of the ASME, Journal of Engineering for Industry 98:66–74 9. Gillespie, L.K., 1976, Deburring, an Annotated Bibliography, Volume III, Technical Paper. Society of Manufacturing Engineers, 1–59, Paper-Nr. MRR76-07 10. Chern, G.L., 1993, Analysis of Burr Formation and Breakout in Metal Cutting, Ph.D. Thesis, University of California at Berkeley, Berkeley, CA 11. Hashimura, M., Hassamamontr, J., Dornfeld, D.A., 1999, Effect of In-plane Exit Angle and Rake Angles on Burr Height
138
12.
13.
14.
15.
16.
G.M. Schueler et al. and Thickness in Face Milling Operation, Transactions of the ASME, Journal of Manufacturing Science and Engineering 121(1): 13–19 Weinert, K., Kahnis, P., Petzoldt, V., Peters, C., 2005, MicroMilling of Steel and NiTi SMA, 55th CIRP General Assembly, STC-C Section Meeting Presentation File, Antalya, Turkey Schmidt, K.H., 2006, Mikrofräswerkzeuge aus Hartmetall, Systeme für Herstellung und Anwendung, Dissertation, Technische Universitaet Kaiserslautern Böhm, B.T., 2005, Klinische Verlaufsstudie zur Verbundfestigkeit von keramisch verblendeten TiAl6Nb7-Kronen und -Brücken. Tübingen Ezugwu, E.U., Wang, Z.M., 1979. Titanium Alloys and their Machinability – a Review, Journal of Materials Processing Technology 68:262–274 Altmüller, S., 2001, Simultanes fünfachsiges Fräsen von Freiformflächen aus Titan, Aachen
17. Aurich, J.C., Haberland, R., Schueler, G.M., Engmann, J., 2008, A New Approach for Using Micro-End Mills at High Rotational Speed and Ultra Low Run-Out, Proceedings of the 3rd International Conference on High Performance Cutting, Dublin, Ireland, pp. 189–197. 18. Schaller, T., Bohn, L., Mayer J., Schubert, K., 1999, Microstructure Grooves with a Width of Less than 50 μm Cut with Ground Hard Metal Micro-End Mills, Precision Engineering 23:229–235. 19. Bissacco, G., Hansen, H., De Chiffre, L., 2005, Micromilling of Hardened Tool Steel for Mould Making Applications, Journal of Materials Processing Technology 167:2–3 20. Klocke, F., König, W., 2008, Fertigungsverfahren, Drehen, Fräsen, Bohren, 8. Auflage, Springer, Heidelberg, 21. Min, S., Sangermann, H., Mertens, C., Dornfeld, D., 2008, A Study on Initial Contact Detection for Precision Micro-Mold and Surface Generation of Vertical Side Walls in Micromachining, CIRP Annals – Manufacturing Technology 57:109–112
Influence of Minimum Quantity Lubrication on Burr Formation in Milling U. Heisel, M. Schaal, and G. Wolf
Abstract Workpieces that are manufactured by machining processes often have to be deburred with considerable effort. This rework involves time and costs. The avoidance or at least minimisation of burr formation offers potential for a more economical production. Another possibility to save costs is to use minimum quantity lubrication. The investigation into the burr formation in milling with minimum quantity lubrication combines these two approaches. The test results presented in this paper show the influence both procedures have on burr formation. Keywords Burr analysis · Milling · Minimum quantity lubrication · Dry machining
1 Introduction Milling is one of the most important manufacturing processes applied in industry and can also be regarded as a valueadded process, in which complex, enhanced parts are manufactured from simple workpieces in each production system. This added value can, however, only be achieved if the manufacturing target values with regard to cost, time and quality are considered. This is especially true in a time when companies are confronted with increasing pricing pressure, intense market competition and growing time pressure. One aspect that equally considers these target values, is the minimisation of burr formation in machining. Thus deburring operations can be simplified or even omitted. Another option is dry machining and hence the abandonment of cooling lubricants as well as minimum quantity lubrication (Fig. 1). Usually one refers to minimum quantity lubri-
U. Heisel (), M. Schaal, G. Wolf Institute for Machine Tools, Universität Stuttgart, P.O. Box 106037, 70049 Stuttgart, Germany e-mail:
[email protected] url: www.ifw.uni-stuttgart.de
Fig. 1 Costs related to cutting fluids in the machining of metals (Federal Statistical Office) [2]
cation when less than 50 ml/h fluid is used. Often minimum quantity lubrication is also referred to as near-dry or dry machining, because the chips for the removal are regarded as dry [1]. Lamers investigated [3] the advantages and disadvantages of wet machining, machining with minimum quantity lubrication and dry machining on the basis of different quality characteristics. The investigations reveal that minimum quantity lubrication consistently shows good to excellent results, except for the cooling effect, the chip transport and the investment costs. Although the introduction of minimum quantity lubrication and dry machining would offer many advantages, only few companies choose these machining options. The reason for that is primarily lack of experience, lot sizes that are too small, testing that is extremely time-consuming, oftenchanging materials as well as manufacturing security and not least resistance within the company [4].
2 Investigations into Burr Formation in Milling Schäfer conducted preliminary extensive investigations into the influence of cutting parameters on the burr formation in milling [5]. Statements about the lubrication were not made. But it can be assumed that a hydrodynamic lubrication was applied, since the tools were made of tool steel and would not have thermally withstood a dry machining or machining with
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_15, © Springer-Verlag Berlin Heidelberg 2010
139
140
minimum quantity lubrication. Schäfer detected that burr root thickness increases with growing feed per tooth. Furthermore, he examined how the cutting edge exit angle influences burr root thickness in face milling and could determine an optimum for a cutting edge exit angle of ca. 85◦ . In addition, Schäfer detected in his investigations that the depth of cut influences burr height and burr thickness. This is, however, only the case as long as the corner radius of the indexable insert is not exceeded. If the depth of cut reaches the value of the corner radius, the burr height drops and then increases moderately from this time with growing depth of cut. Dornfeld makes a distinction between two mechanisms of burr formation in milling, which are independent of the chip formation process. The primary burr is formed by the cutting edge. The secondary burr is generated by the tool flank [6]. The primary burr develops, when the cutting edge exits the material and the workpiece material cannot be cut off anymore. The secondary burr develops when the plasticity of the material is so low that a burr is already formed before the chip. According to his investigations, the negative effects of dry machining on burr formation have to be attributed to the higher temperatures due to the lack of cutting fluid and the higher ductility of the material because of that. No statements were, however, made about the influence of minimum quantity lubrication on burr formation [7]. Furthermore, Dornfeld examined the influence of the pressure angle and detected that burr height increases with growing pressure angle [8].
U. Heisel et al.
Fig. 3 Measured values of a burr [11]
The standard ISO 13715 defines the edge of a workpiece as burred if it has an overhang greater than zero. With regard to a burr avoidance or reduction this definition is not very meaningful though, as no measure is given for the machinability of the burr. For this reason, Schäfer [4, 11] defined the so-called burr value (Fig. 3). By the burr value g the four geometric parameters of burr root thickness bf , burr root radius rf , burr thickness bg , and burr height h0 are combined to a comparative value. The different weighting factors result from the effect with which the individual burr parameters influence the deburring process. In the following, the burr value will be used to describe burr formation. The burr value or rather the burr parameters were determined with micrographs by means of a microscope at least two points. For each parameter combination the tests are repeated twice.
3.1 Measuring Points in Angle and Face Milling 3 Definition of Burr and Burr Parameters The assessment whether a burr is critical or not is mostly laid down internally in company standards. Many companies even check and deburr all components as a precaution. A small number, however, does without the analysis and evaluation of burrs [4]. To characterise a burr, the burr height or the five burr types defined by CODEF (Consortium on Deburring and Edge Finishing) for drilling are used frequently [9]. Figure 2 presents an overview of different burr forms.
Fig. 2 Burr forms (according to [10])
Two kinds of burrs arise in milling: the entrance or Poisson burr and the exit burr. The entrance burr develops perpendicularly to the cutting force direction. The exit burr develops in the direction parallel to the cutting force. Figure 4 illustrates the burrs for angle up-milling.
Fig. 4 Measuring points in angle milling (up-milling)
Influence of Minimum Quantity Lubrication on Burr Formation in Milling
141
Fig. 7 Indexable insert, face and angle milling cutters
Fig. 5 Measuring points in face milling
Fig. 6 Burr formation depending on measuring position
In down-milling the formation mechanisms of the burrs 1–7 are changed from exit burrs to Poisson burrs and vice versa except for burr 3. A pure exit burr and a pure entrance burr as well as two mixtures of these two formation types are produced in face milling (Fig. 5). The definition of the measuring points in angle and face milling in Figs. 4 and 5 also shows the burr forms of the respective edges. These burr forms are constant over each entire edge, but their sizes are not. The size rather varies partly heavily depending on the measuring position (Fig. 6).
4 Test Procedure The tests were conducted on an EX-CELL-O XHC 241 machining centre. A single channel unit by the company Lubrix was used as minimum quantity lubrication system. Fifteen different programs can be set with its three pressure stages. These programs range from program 1 with pure compressed air to program 15 with a fluid saturation of the air of 100%. Furthermore, the agent can be delivered by the rotary feedthrough of the spindle and the internal cooling of
the tool as well as by external supply to the cutting region with the nozzle. A standard nozzle for high material removal rates by the company Microjet was used as external nozzle (Fig. 16). In the tests the pressure in the pipeline was pF = 3.6 bar on average for the variants 1, 3 and 5, and pF = 4.1 bar for the variants 12 and 15. Serving as agent was the lubricant Ecocut Mikro Plus 82 by the company Fuchs Schmierstoffe. This lubricant was developed especially for minimum quantity lubrication machining and is based on special fatty alcohols. A face milling cutter and an angle milling cutter are used as test tools (Fig. 7). The face milling cutter has a diameter of d = 50 mm, a helix angle of δ = 12◦ and six indexable inserts. It can be used for a cutting speed of up to vc = 250 m/min and a depth of cut of ap = 11 mm. The angle milling cutter has a diameter of d = 25 mm, a helix angle of δ = 8◦ and three indexable inserts. It can be used for the same parameters as the face milling cutter. The indexable inserts used are identical for both tools and merely differ in corner radius, which is rε = 0.4 mm (average cutting edge radius rβ = 34 μm), rε = 0.8 mm (rβ = 30 μm) and rε = 1.2 mm (rβ = 29 μm) for the tests. These CVD-coated indexable inserts have a layer structure of TiCN + Al2O3 (+TiN) and a cutting edge length of b = 1.2 mm. The tool cutting edge angle is κ = 90◦ for both milling cutters. The heat-treatable steel C45E was used as reference material. Comparative tests were carried out with 42 CrMo 4 V and AlCuMgPb-F34. The cutting speed was vc = 225 m/min for the comparative tests, and the feed per tooth was fz = 0.11 mm per revolution (Fig. 8). For the further tests the cutting speed was varied in the range from vc = 150 m/min to vc = 225 m/min, and the feed per tooth was varied in the range between fz = 0.05 mm and fz = 0.11 mm. The tests were performed with a constant depth of cut of ap = 3 mm. In addition to that, the width of cut ae was varied. Concerning the face milling cutter, the milling was conducted in the middle of the workpiece with a width of cut of ae = 12.5, 25 and 37.5 mm. Regarding the angle milling cutter, widths of cut of ae = 6.25, 12.5, 18.75 and 23.5 mm were investigated.
142
U. Heisel et al.
Fig. 10 Measurement of minimum quantity Fig. 8 Experimental plan
were measured as well as dry machining, corresponding to variant 0.
5 Burr Formation in Angle Up-Milling 5.1 Influence of Minimum Quantity
Determining the minimum quantity lubrication is elementary for examining the connections in the investigation into how minimum quantity lubrication influences burr formation. The investigation is carried out at different rotational speeds and the different stages of the Lubrix system. A receptacle has proven to be a simple and economical measuring instrument for determining the minimum quantity. The workpiece is inserted into one side of this receptacle and sealed. In its opposite side, there are openings for the air to escape. Layers of cellulose and a sponge have proven to be a favourable absorber (Fig. 9). A spray duration of 10 min is set for the tests. The blownout lubricant is collected, and its mass is determined from the difference between empty and loaded receptacle. From this the quantity of the fluid can be determined via its density. The curve of the lubricant quantity in Fig. 10 results from the different pressure stages. The variants 1, 3, 5, 12 and 15
To determine how the factor of minimum quantity lubrication influences the burr value, tests with the reference material C45E were conducted at a width of cut of 12.5 mm, a cutting speed of vc = 150 m/min and a feed per tooth of fz = 0.05 mm. Figure 11 shows that the burr value decreases with growing minimum quantity lubrication for the burr 4. In the case of the burrs 1 and 2, a slight increase can be noted before the burr value decreases here as well, as the minimum quantity lubrication increases. The greatest difference in burr value can be detected for the variants MQL 0 (dry) and MQL 5 (minimum quantity). Furthermore, a significant influence of minimum quantity lubrication can not be detected. For this reasons the tests with the settings MQL 0 and MQL 5 of the Lubrix system were continued in the following.
Fig. 9 Measurement of lubricant amount in milling
Fig. 11 Influence of minimum quantity
Influence of Minimum Quantity Lubrication on Burr Formation in Milling
Fig. 12 Influence of cutting speed with minimum quantity lubrication
143
Fig. 13 Influence of feed with minimum quantity lubrication
5.2 Influence of Cutting Speed The tests were carried out at a constant feed per tooth of fz = 0.07 mm and fz = 0.11 mm respectively, and stage 5 (MQL 5). As it is already known from literature, cutting speed has only a small yet positive effect in milling (Fig. 12). In dry milling, the machining provides a similar result. The influence of cutting speed on burr formation is minor here as well. A significant difference between dry machining and the machining with minimum quantity lubrication could not be detected. When varying the cutting speed, the feed per tooth was kept constant. As a consequence of this, all cutting tests have the same chip thickness. According to Victor and Kienzle, the resultant force theoretically does not depend on the cutting speed, but on the chip thickness. Since burr formation is decisively influenced by the predominant forces, which are theoretically constant, it follows that the burr value also remains approximately constant when the cutting speed is varied.
5.3 Influence of Feed By analogy with the previous section, the influence of feed is investigated at a constant cutting speed of vc = 225 m/min and stage 5 in the following. While the entrance burr 3 increases at first and then decreases constantly in machining with and without minimum quantity lubrication, the exit burr 1 behaves in the opposite way in dry machining and the machining with minimum quantity lubrication (Fig. 13). In the machining with minimum quantity lubrication, the burr value drops at first and then rises again from a feed of f = 0.07 mm up, before its course remains nearly constant from f = 0.09 mm up. In dry machining the burr 1 behaves in exactly the opposite way. The burr value here increases slightly at first, drops from
Fig. 14 Influence of feed without minimum quantity lubrication
f = 0.07 mm to f = 0.09 mm and then increases moderately with growing feed. Alternatively, the course of the curve in dry machining can be considered as a shift of the curve towards higher feed values, in comparison with the machining with minimum quantity lubrication (Fig. 14). The reason for this could be a change in the friction conditions during cutting or different temperatures in the burr forming area. A positive influence of low feeds per tooth on burr formation could not be confirmed, like it is the case in short hole drilling [12]. Hence, the statement that more material is available for deformation with increasing feed per tooth cannot be applied to milling. Moreover, the finding made regarding the variation of cutting speed cannot be applied to the variation of feed, for otherwise a rise in resultant force and thus in burr formation would result from an increase in feed per tooth.
5.4 Influence of Width of Cut It showed that the tendency to burr formation gets lower for the exit burr 1, as width of cut increases (Fig. 15). The burr value with minimum quantity lubrication is less than in dry machining from a width of cut of ae = 12.5 mm up.
144
U. Heisel et al.
Fig. 15 Influence of width of cut Fig. 17 Influence of cutting edge corner
The lower tendency to burr formation in the cutting with minimum quantity lubrication and great width of cut could result from the lower temperature during machining. This was recorded with a high-speed thermographic camera at 50 Hz. The individual pictures were combined in a film by means of the evaluation software. Then five measuring points were positioned on the exit edge. Their maximum temperature was established, and the average was calculated from that (Fig. 16). The maximum temperature arises each time when the cutting edge exits. As a result of the lower temperature, the yield stress of the workpiece material increases and smaller burrs are formed [13].
5.5 Influence of Corner Radius As in the case of the previous influencing factors, the corner radius has a minor influence on burr formation as well. It showed, however, that small corner radii tend to have a little advantage over large corner radii concerning the burr formation for C45E (Fig. 17). Olvera and Barrow [14] attribute this effect to a zone where no cutting process occurs, but a pure deformation process. This region increases with growing corner radius. The authors do, however, not give any details about the extent of this effect. When the cutting edge exits the workpiece, the cutting process is steadily turning into a deformation process as soon as the material still to be machined cannot offer enough resistance to the forces any more. As corner radius increases, the force perpendicular to the direction of feed motion or rather in the direction of the workpiece rises and thus encourages that the remaining material is deformed into a burr. Furthermore, the tests showed that the burr values with minimum quantity lubrication are higher than those of dry machining. Tests with the workpiece materials 42CrMo4V and AlCuMgPb-F34 revealed a similar result.
5.6 Influence of Supply
Fig. 16 Influence of width of cut on temperature
Another possibility for delivering the minimum quantity lubricant to the region of cutting is to attach an external nozzle. The angle between nozzle and angle milling cutter was varied in the tests conducted (Fig. 18). Again, the cutting parameters were vc = 225 m/min and fz = 0.11 mm. The depth of cut was ap = 3 mm, and the width of cut of the milling cutter was ae = 12.5 mm. The examined positions, relative to the direction of primary motion, were 0◦ , 90◦ and 180◦ . The inclination to the component surface was 45◦ .
Influence of Minimum Quantity Lubrication on Burr Formation in Milling
145
Fig. 18 Position of nozzle for external supply
Fig. 20 Comparison between down-milling and up-milling
Fig. 19 Influence of supply
The nozzle itself was fastened to the spindle by means of a mounting. Hence, the relative position of the nozzle, related to the milling cutter, has not changed during the milling process. The spray positions of 90◦ and 180◦ proved to be favourable with regard to a lower burr formation (Fig. 19). Internal supply and dry machining, however, provide better results. The entrance burr 3 is an exception. It is smaller in the tests with minimum quantity lubrication than in dry machining. One reason for this could be that the minor cutting edge or the workpiece in this area is moistened with minimum quantity lubrication in the case of external supply, whereas the fluid has already evaporated in the case of internal supply.
6 Burr Formation in Angle Down-Milling In contrast to up-milling, the chip thickness at the cutting edge entrance is highest and greater than at the cutting edge exit at a width of cut of ae = 12.5 mm in down-milling. Hence, a lower burr formation in down-milling should be resulting. The tests were conducted with the angle milling cutter at the cutting parameters of vc = 225 m/min and
fz = 0.11 mm in each instance with dry machining and minimum quantity lubrication. Figure 18 shows the expected course of the burr values. Thus the burr values in down-milling are only half of the values in up-milling. However, it has to be taken into account that, in contrast to up-milling, the burrs 1 and 2 are entrance burrs, but the burr 4 is an exit burr in down-milling (Fig. 4). The burr 3 is an entrance burr in both processes. In addition, the burr values increase in both processes if minimum quantity lubrication is used. When looking at the burr thickness, it can be detected that the burr thickness of the entrance burrs is smaller in down-milling than the one in up-milling. This conflicts with investigations by Rangarajan and Dornfeld [15], who could not find an influence of the milling process on burr thickness. A significant influence of corner radius on burr formation cannot be detected in down-milling (Fig. 20).
7 Burr Formation in Face Milling By analogy with the tests in angle milling, the tests in face milling were conducted with the cutting parameters of vc = 225 m/min and fz = 0.11 mm at a depth of cut of ap = 3 mm and workpiece width of ae = 37.5 mm. The minimum quantity lubrication was carried out through an external nozzle and carried along with the spindle. The indexable inserts used correspond to those of the angle milling cutter. The mixed burr 2 consists of two areas: one area with entrance burr and one with exit burr (Fig. 5). The point P separating the two areas from each other can be calculated from the cutting parameters as well as the geometry of the milling cutter and of the component. The distance from the centre line is
p≈
2 · π2 · d z · 360◦
2 − fz2 = 0.44 mm
(1)
146
U. Heisel et al.
The supply of the fluid to the cutting region is another parameter that was varied within the framework of these investigations. In this connection the supply of the fluid through an external nozzle proved to be disadvantageous. The burr values of the exit burrs were higher than those of the internal supply and those of dry machining. Moreover, it was shown that the burr formation tendency decreases with increasing width of cut. Acknowledgments This research was conducted within the framework of a project supported by the German Research Foundation (DFG).
Fig. 21 Influence of cutting edge corner in face milling
References
for the given cutting values and the milling cutter geometry. This is clearly less than the measured distance of up to p = 8 mm. In face milling it can be detected that the burr value decreases with increasing corner radius. This results from an increase in burr root thickness and burr root radius with decreasing corner radius. The greatest decrease in burr value shows AlCuMgPb-F34 (Fig. 21). The influence of minimum quantity lubrication on burr formation is not obvious. Compared with dry machining, the burr value of burr 2 decreases noticeably when minimum quantity lubrication is used in the case of a corner radius of rε = 1.2 mm for C45E as well as rε = 0.8 mm and rε = 1.2 mm for AlCuMgPb-F34. However, the difference is little in a lot of cases.
1. Weinert, K., 1999, Trockenbearbeitung und Minimalmengenkühlschmierung. Einsatz in der spanenden Fertigungstechnik, Springer-Verlag 2. N.N., 2004, Trockenbearbeitung in der Praxis, Landesgewerbeamt Baden-Württemberg 3. Lamers, R., 1999, Lohnt der Umstieg auf Minimalmenge? WB Werkstatt und Betrieb 138(12):79–81 4. Aurich, J. C., 2006, SpanSauber, Untersuchung zur Beherrschung der Sauberkeit von zerspanend hergestellten Bauteilen, Ergebnisworkshop, Lehrstuhl für Fertigungstechnik und Betriebsorganisation, Technische Universität Kaiserslautern 5. Schäfer, F., 1976, Untersuchungen zur Gratbildung und zum Entgraten insbesondere beim Umfangsstirnfräsen, Dissertation, Universität Stuttgart 6. Dornfeld, D. A., Avila, M. C., 2004, On the Face Milling Burr Formation Mechanisms and Minimization Strategies at High Tool Engagement. Consortium on Deburring and Edge Finishing, University of California, Berkeley, CA 7. Dornfeld, D. A., Shefelbine, W., 2004, The Effect of Dry Machining on Burr Size, Consortium on Deburring and Edge Finishing, University of California, Berkeley, CA 8. Dornfeld, D. A., Avila, M., 2003, The Effect of Kinematical Parameters and Tool Geometry on Burr Height in Face Milling of Al-Si Alloys, Consortium of Deburring and Edge Finishing, University of California, Berkeley, CA 9. Dornfeld, D. A., Min, S., Reich-Weiser, C., 2004, Burr Formation, Deburring & Surface Finishing, 7th International Conference on Deburring and Surface Finishing, Berkeley 10. Dornfeld, D. A., Kim, J. S., Dechow, H., Hewson, J., Chen, L. J., 1999, Drilling Burr Formation in Titanium Alloy, Ti-6Al-4 V, Annals of the CIRP, 48(1):73–76 11. Schäfer, F., 1992, Grundlagen zur Lösung von Entgratproblemen, Entgrat-Technik, Entwicklungsstand und Problemlösungen, Reihe Kontakt und Studium, Oberfläche, 392:33–42 12. Heisel, U., Luik, M., Schaal, M., 2005, Berechnung der Gratgrößen beim Kurzlochbohren, Spanende Fertigung, 4:123–131 13. Luik, M., 2007, Gratbildung und Gratminimierung bei asymmetrisch mit Hartmetall-Wendeschneidplatten bestückten Bohrwerkzeugen, Dissertation, Universität Stuttgart 14. Olvera, O., Barrow, G., 1998, Influence of Exit Angle and Tool Nose Geometry on Burr Formation in Face Milling Operations, Proceedings of the Institution of Mechanical Engineering B, Journal of Engineering Manufacture, 212(1):59–72 15. Rangarajan, A., Dornfeld, D. A., 2004, Back Cutting and Tool Wear Influence on Burr in Face Milling – Analysis and Solutions, Consortium on Deburring and Edge Finishing, University of California, Berkeley, CA
8 Conclusion Apart from the influences of the procedural parameters cutting speed and feed on burr formation in the machining with minimum quantity lubrication, the effects of the width of cut, the beam control, the lubricant quantity and the cutting edge geometry were examined. The tests showed that the burr value increases in the machining with minimum quantity lubrication compared to dry machining, but does not change when varying the minimum quantity. Hence, the cutting speed does not have any significant influence on burr formation in the machining with a minimum quantity. However, when varying the feed per tooth in angle milling, the exit burr curve of the lateral face shifts towards higher values in dry machining compared with minimum quantity lubrication. Compared to angle milling, the cutting edge radius has a negligible influence on burr formation in face milling. In the angle milling cutter the burr value increases with increasing corner radius.
Burr Formation and Removal at Profile Grinding of Riblet Structures B. Denkena, L. de Leon, and B. Wang
Abstract Compared to smooth surfaces, ideal riblet structures have proven to reduce skin friction and wall shear stresses in turbulent flow by up to 10%. Investigations in the wind tunnel show that the tip radius of the riblets has a significant influence on the reduction of the skin friction. One promising process for the production of riblets on large surfaces is profile grinding. In the first experiments conducted only a minimum tip radius of 8 μm was reached. Micrograph sections of the workpiece show that the radius results from a rolled burr. The contact area between the grinding wheel and the workpiece is simulated in order to identify the mechanisms of the burr formation. It is observed that the contact area is asymmetric. Thus the removed material mainly flows to one side of the groove and forms a burr. In the following, the influence of the grinding strategy, the grinding wheel specifications and the grinding parameters on the size of the burr is investigated. The experiments have shown that the burr formation in up grinding process is quite unsteady in comparison to down grinding. The cutting speed and the feed rate as well as the grinding wheel specifications have no significant influence on the burr size. However, the burr size develops linear to the depth of cut, i.e. an increase in the depth of cut leads to an increase in the burr size. The burr can be removed by a spark-out process after the grinding process. Furthermore, the burr size decreases with an increasing feed rate during the spark-out process. Thus a riblet tip radius below 2 μm is achieved. Keywords Burr removal · Profile grinding · Burr formation mechanisms
B. Denkena, L. de Leon, B. Wang () Institute of Production Engineering and Machine Tools, Leibniz Universität Hannover, An der Universität 2, 30823 Garbsen, Germany e-mail:
[email protected]
1 Introduction Surface relief structures at micro and nanoscales are currently focused in different engineering areas. These structures enable special optical, mechanical or chemical properties of the workpiece surfaces [1]. One example is the riblet surface with microgrooves in the flow direction. Compared to smooth surfaces, ideal riblet structures have proven to reduce skin friction and wall shear stresses in turbulent flow by up to 10% [2]. The basic principle of the drag reduction can be ascribed to the displacement of the virtual derivation of the fluid flow in both the lateral and vertical direction. Thus the lateral fluctuation in the viscose underlayer of the turbulent boundary layer can be reduced [3]. For most technical applications of riblets, microgroove structures with a width of less than 100 μm and a depth of the half of the width are required on extensive surfaces. Besides the geometric dimension of the riblets, the radius on the profile tip rtip,riblet has to be as small as possible in order to achieve a high effectiveness of the riblet structures [3]. Within the joint research work “Riblets for Compressor Blades”, which is supported by the German Research Foundation (DFG), profile grinding has been investigated as a micromachining process for the production of riblets directly on compressor blades [4]. Figure 1 shows the predefined ideal riblet geometry and the relevant riblet forms used for this work. Microgrooves with a width of 40 μm and a depth of 20 μm are to be produced. Furthermore, in order to achieve a high aerodynamic efficiency of the riblet-structures, the tip radius of the profile has to be as small as possible [3]. In this case, the ideal radius on the profile tip of the riblets rtip,riblet is 0.2 μm. However, due to the applied material X20Cr13 (1.4021, typical steel for compressor blades) and the desired micro profile geometry of the workpiece, burr formation during the grinding process has a significant influences on the profile geometry, especially in the profile tip area.
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_16, © Springer-Verlag Berlin Heidelberg 2010
147
148
B. Denkena et al.
Fig. 1 Ideal riblet-geometries
2 Burr Formation at Riblet-Grinding 2.1 Experimental Setup The grinding experiments were carried out on a high precision surface grinding machine (Blohm Profimat 407). Vitrified bonded wheels were selected due to the advantageous dressability and profile holding properties compared to other bonding systems [5]. In order to achieve a high cutting speed and a small wear rate, a wheel diameter of 300 mm was chosen. Multiple V-shaped wheel profiles have been generated by a profile dressing process. A conventional SiC wheel SiC400 (dG = 17 μm) as well as a superabrasive
Fig. 2 Grinding wheel profile and grinding strategy
CBN wheel MB16 (dG = 16 mm) have been applied, which offered the highest profile form accuracy during the dressing experiments [6]. The upper part of Fig. 2 shows the profile tip geometry of the grinding wheel after the dressing process. The wheel profile geometry was generated by applying a profile dressing process with a special profile shift kinematic [7]. Due to breakouts at the profile tip there is a profile height deviation between the ideal sharp and the actual wheel profile of h = 30 μm. In the following profile grinding process, microgrooves have been ground onto the workpiece surface. Since the distance between two profile tips on the grinding wheel is larger than the desired riblet-width sriblet (due to
Burr Formation and Removal at Profile Grinding of Riblet Structures
the dressing tool geometry), grinding strategies with profile overlapping have been applied to achieve continuous riblets. The lower part of Fig. 2 illustrates the profile grinding strategy. At first, single profiles are ground. In the second step, the profile grinding is done with an axial profile offset. Thus the desired riblet profiles can be machined through offsets of the single profiles.
2.2 Characterization of the Burr Formation at Riblet-Grinding After the profile tips were generated on the grinding wheels by a dressing process, profile grinding experiments have been carried out. Figure 3 shows the SEM-images and the micrographs of the microgrooves ground by the SiC400 wheel. In order to reduce the profile wear due to the grinding process, a small material removal rate was chosen. The upper part of Fig. 3 shows a single ground profile. It can be observed that the plastic material removal leads to side burrs due to the edge bulging effect. The lower part of Fig. 3 shows the riblets produced by grinding with an axial profile offset z of 60 μm. Due to the micrograph, the profile width is 60 μm and the profile height is 20 μm, while the infeed at the grinding process ae is set at 30 μm. The asymmetric profile flanks and the significant spiral side burrs at the profile tips are especially remarkable. Furthermore, the burr size is much larger compared to the burrs produced when a single profile is ground (Fig. 3, upper). In order to increase the aerodynamic efficiency of the riblet structures, the tip radius of the spiral burrs has to be reduced [3].
Fig. 3 The riblet-grinding process
149
According to the definition of burr formation at the surface grinding process [8], resulting burrs are located at the edges of the workpiece and can be divided into entrance burr, side burr and exit burr (Fig. 4, left). According to this definition the spiral burrs produced at the riblet-grinding process can be classified as side burrs. In Fig. 4, the spiral formed side burr is characterized by a set of geometrical parameters. Besides the burr thickness at the root bf and the unwound burr length bl , the geometry of the spiral burr is described by the spiral burr height bhr and the spiral burr width blr . In most of the previous works of burr formation during grinding processes, surface grinding of steel was investigated [9–11]. It was observed that among the three different burr types (entrance burr, side burr and exit burr) the exit burr is the largest in size. Hence most of the investigations concentrated on the formation of the exit burr. The formation mechanism of the exit burr is explained by the plastic material deformation at the exit edge. Since the edge of the workpiece does not offer sufficient resistance against plastic deformation due to acting forces, the remaining material on the workpiece edge is deformed and bent over the edge [12]. Furthermore, it was observed that among the process input parameters, the angle on the workpiece edges γ, the grinding wheel specifications and the cutting parameters have significant influences on the burr formation. In contrast to the exit burr, the side burr was not investigated in any detail in the previous research works. In the work of Warnecke et al. [9], spiral burrs as side burrs were observed at surface grinding. However, the experimental results showed no direct correlation between the size of the side burr and the process input parameters like the wheel specifications and the grinding parameters.
150
B. Denkena et al.
Fig. 4 Characterization of the side burr at riblet-grinding
The riblet-grinding process investigated in this work is a micro profile grinding process with axial profile overlapping. The process kinematics differs considerably from those of surface grinding processes. The investigations conducted in this work focus on side burrs. The burr formation is not only determined by the current grinding process itself, but also by the existing groove structure before grinding. In the following grinding experiments the influence of different process input parameters on the burr formation will be investigated.
Fig. 5 Grinding strategy and burr formation
3 Influence of the Grinding Process on the Burr Formation 3.1 Grinding Strategy The burr formation behaviour has been investigated for different grinding strategies (up/down grinding). Figure 5 shows the REM-image for a direct comparison of the ground ribletstructures. In the up grinding mode, the burr formation
Burr Formation and Removal at Profile Grinding of Riblet Structures
is inhomogeneous along the profile peaks. This can be explained by the different cutting kinematics of the single grains. In down grinding, the single grain chip thickness reaches its maximum within a very short time, while a chip is formed. On the contrary, the single grain chip thickness slowly rises up to its maximum in the up grinding mode. Hence the period after which the effective chipping thickness is exceeded is much longer compared to that in down grinding and the effectiveness of the chip formation is lower. As a result, there are strong interactions between the current chip forming process and the subsequent one, which leads to an inhomogeneous chip formation. According to this observation only the down grinding mode was applied in the following grinding experiments.
3.2 Grinding Wheel Specifications Besides the grinding strategy, the influence of the grinding wheel specifications on the burr formation was investigated. Both the conventional SiC wheel and the superabrasive CBN-wheel have been applied under the same cutting conditions. The CBN and the SiC grains have different properties. Generally, CBN has a cool grinding effect due to its high thermal conductivity compared to conventional grains [13]. Thus the expected workpiece temperature in the cutting zone is lower. Since the plastic material flow is influenced not only by mechanical but also by thermal factors, the effect of the temperature on the burr formation can be observed by comparing the grinding results. Figure 6 shows the comparison of the ground riblet structures and the burr characterization factors. All geometric parameters of the burr were measured by means of micrographs of the workpiece and a graphical software. The value in the table is the average of three measured values. No significant difference at both of the spiral burr height bhr and
Fig. 6 Grinding wheel specifications and burr formation
151
the unwound burr length bl could be observed. However, the characteristic of the burr roll-in effect is stronger at the SiCgrinding than at the CBN-grinding, which can be observed by the difference in the spiral burr width blr . In conclusion, the influence of the thermal effect on burr formation is relatively low and is not the main reason for the formation of the spiral side burr.
3.3 Grinding Parameters In the next step, the influence of the grinding parameters on the side burrs at riblet-grinding will be investigated. Each of the parameters cutting speed vc (between 5 and 35 m/s), cutting infeed ae (between 30 and 80 μm) and feed rate vft (between 30 and 1200 mm/min) was varied in several steps. The grinding results show that both the cutting speed vc (between 20 and 35 m/s, Fig. 7, left) and the feed rate vft (between 30 and 240 mm/min, Fig. 7, right) have no remarkable influence on the side burr formation. Furthermore, at a low cutting speed vc = 10 m/s and a high feed rate vft = 1200 mm/min, the resulting profile height of the riblets is very low due to the high profile wear. This can be explained by the high single grain chip thickness at the grinding process, which leads to an overloading of the wheel profile. It was observed that there is a linear correlation between the he (height difference between the infeed ae and the profile root height hf ) and the unwound burr length bl when the cutting infeed ae is varied (Fig. 8). Since the profile root height hf was kept constant during the tests, the unwound burr length bl increases proportionally with increasing cutting infeed ae . The calculated ratio of the length stretching between the bl and the he is about 1.6. This behaviour can be explained by the plastic material deformation of the burr during the grinding process.
152
B. Denkena et al.
Fig. 7 Influence of the cutting speed and the feed rate on the burr formation
Fig. 8 Influence of the cutting infeed on the burr formation
4 Burr Formation Mechanisms In order to identify the mechanism of the burring formation at riblet-grinding, the contact surface geometries between the grinding wheel profile and the existing workpiece profiles have been calculated using a material removal simulation (Fig. 9). Due to the profile overlapping with the existing groove the resulting riblet-height hriblet is lower than the cutting infeed ae . Along the contact length in the cutting direction, the actual infeed decreases from the set cutting infeed ae to zero. An analysis of the profile sections (A, B, C) shows that the geometry line in the exiting area of the grinding wheel is asymmetric (section A, B). Thus the removed material
mainly flows from the side of the profile flank in grinding to the side of the existing profile flank and forms a burr. Furthermore, the feed movement of the wheel profile in the cutting direction causes the roll-in effect of the formed burr. If the spiral burr is unwound, the burr thickness increases continuously from the tip to the root area (Fig. 8).
5 Deburring Process In order to achieve a high functional efficiency of the riblet structures, it is very important to keep the size of the side burr as small as possible. One possibility is to adjust the profile overlapping to a value of exactly one (ae = hriblet ).
Burr Formation and Removal at Profile Grinding of Riblet Structures
153
Fig. 9 Burr formation mechanisms
Fig. 10 Deburring by means of a spark-out process
However, due to the profile flank wear the achieved profile width decreases during the grinding process. In this case there is no more profile overlapping and the produced ribletprofiles are not continuous. Hence, another method for the deburring process has to be applied. Due to the process kinematics at profile grinding the burrs can be cut off by applying a spark-out grinding process, while the whole profile grinding process is executed once more without an additional cutting infeed ae . Since the material removal volume is low at the spark-out process, different feed rates between 60 and 1200 mm/min were investigated. The achieved results are presented in Fig. 10. With an increasing feed rate vft,sp at the deburring process all burr characterizing factors decrease.
The spiral burrs are almost completely removed at vft,sp = 1200 mm/min and the minimum achieved riblet profile tip radius rtip,riblet is about 1 μm.
6 Summary Within this paper, the burr formation at profile grinding of microgroove structures on metal surfaces was investigated. The ground riblet structures have shown asymmetric profile flanks and significant spiral burrs at the profile tips. In
154
the following, grinding experiments have been carried out to determine the influence of the grinding strategies and the grinding parameters on the burr formation. Among the process input parameters, only the cutting infeed shows a strong influence on the unwound length of the spiral burr. The contact area between the grinding wheel and the workpiece was analyzed by means of a material removal simulation. The spiral burr formation can be explained by the asymmetric contact profiles. A spark-out grinding process was applied to remove the spiral burr. A high feed rate has proven to be more efficient in this deburring process. Acknowledgements The investigations presented in this paper are kindly supported by the German Research Foundation (DFG) within the scope of the project “Riblets on Compressor Blades”.
References 1. Bläsi, B., Boerner, V., Gombert, A., Heinzel, A., Kübler, V., Wittwer, V., 2000, Combined functional surface-relief structures for various applications, International Colloquium on Surfaces, Vol. 10, pp. 222–229 2. Bruse, M., 1998, Zur Strömungsmechanik wandreibungsvermindernder Riblet-Oberflächen, Dr.-Ing. Dissertation, TU Berlin, Germany 3. Hage, W., 2005, Zur Widerstandsverminderung von dreidimensionalen Riblet-Strukturen und anderen Oberflächen, Dr.-Ing. Dissertation, TU Berlin, Germany
B. Denkena et al. 4. Seume, R., Ostendorf, A., Denkena, B., Reithmeier, E., Meyer, R., 2007, Exploratory Experiments on Machined Riblets for 2-D Compressor Blades, ASME, Seattle, USA 5. Hessel, D., 2003, Punktcrushieren keramisch gebundener Diamantschleifscheiben, Dr.-Ing. Dissertation, Universität Hannover, Germany 6. Denkena, B., Reichstein, M., Wang, B., 2007, Manufacturing of Functional Micro-structured Surfaces by Grinding, International Symposium on Advances in Abrasive Technology (ISAAT), September 26th–28th, Dearborn, MI, USA 7. Denkena, B., de Leon, L., Wang, B., 2008, Grinding of Microstructured Functional Surfaces – A Novel Strategy for Dressing of Microprofiles. Production Engineering. Research and Development Online, DOI-10.1007/s11740-008-0134-0 8. Schäfer, F., 1975, Entgraten, Krausskopf Verlag, Mainz 9. Warnecke, G., Dollmeier, R., Barth, C., 2002, Gratbildung beim Schleifen von gehärtetem Stahl mit konventionellen und hochharten Schleifscheiben, Industrie Diamanten Rundschau, Band 36, Vol. 3, pp. 202–208 10. Aurich, J.-C., Sudermann, H., Bil, H., 2005, Characterisation of burr formation in grinding and prospects for modelling, CIRP Annals, Vol. 54, No. 1, pp. 313–316 11. Aurich, J.-C., Sudermann, H., Braun, O., 2006, Experimental investigation of burr formation in the surface grinding of tool steel. Journal of Engineering Manufacture, Band 220, Vol. 4, pp. 489–497 12. Barth, C., Dollmeier, R., Warnecke, G., 2001, Burr Formation in Grinding of Hardened Steel with Conventional and Superabrasive Wheels, Transactions of NAMRI/SME, Vol. 29, No. 1, pp. 273–278 13. Tönshoff, H.-K., Denkena, B., 2004, Spanen, Springer Verlag, 2. Auflage, Berlin
Burr Measurement
Burr Measurement System for Drilled Hole at Inclined Exit Surface H.P. Hoang and S.L. Ko
Abstract Burr is an undesirable projection as a result of plastic deformation in metal cutting and blanking operation. Burr minimization and effective deburring process are required strongly to improve the efficiency of machining and the quality of product. The necessity of a suitable measurement device for inspecting drilled hole from micro to macro burrs is required. The burr measurement system uses conoscopic holography sensor which offers certain advantages over interferometry, triangulation and dynamic focusing, for point-by-point distance mapping. In terms of this work, the improvement of burr measurement system which has ability to measure burr on curved and inclined exit surfaces which causes the out of range signals will be introduced and discussed in this paper. Keywords Burr measurement Conoscopic holographic
·
Inclined
surface
·
1 Introduction Burr is formed as a result of plastic deformation in metal cutting and blanking operation [1], which is undesirable in manufacturing and has great influence on the accuracy of the parts. To remove or prevent these burrs effectively, burr must be measured with accuracy. For the burrs formed in machining are irregular and very sharp in shape, it is usually very difficult to measure burr accurately. If the geometry of burr is measured properly, the proper deburring method or processes can be recommended. Burrs are formed along
H.P. Hoang, S.L. Ko () Department of Mechanical Design and Production Engineering, Konkuk University, 1 Hwayang dong, Gwangjin gu, 143-701, Seoul, Korea e-mail:
[email protected] url: http://premalab.re.kr
the edge periphery, generally non-uniform and usually contain thin and sharp peaks. The height of the burr formed in precise and micro drill may vary from thousands to a few microns while the hole size usually are of ten to one tenth millimeters in diameters. Specially in drilling at inclined exit surface, burr height distribution is different as in Fig. 1. Such burr geometry and large measurement range makes it difficult to find proper methods for burr measurement [2]. Contact stylus is slow and plastic deformation due to the pressure might reduce the real burr height. Optical instruments at submicro or nanometer scale have high resolution, however their vertical and lateral measurement ranges are very small [3]. Optical sensors at larger scale, on the other hand encounter the difficulty of the burr edges with steep angles which may cause insufficient light intensity or direct reflections to the detector [4]. In an experiment [5], three kinds of sensors, interferometry, triangulation and conoscopic holograhy sensors for similar measurement range mounted on a coordinate measuring machine were used to scan radially a standard cylinder specimen having diameter of 0.5 mm. When the beam spot moved away from the centerline of the cylinder, only the conoscopic holography sensor, also referred as the scanning conoscopic probe (SCP), could measure the form accurately while the other signals from two sensors became lost or unstable. Compared with the other two sensors, the SCP has small beam spot size, which is also important because the burr edges are thin. For these reasons, we have selected the SCP for our system. Besides simple case of burr measurement on no exit angle surface, there are many intersecting holes in the drilling process where burrs locate at curved or inclined surface as in Fig. 2. It is difficult for measuring and analyzing these kinds of burrs. Depending on the size of hole and angle of inclined exit surface (Fig. 3), the measuring distance will change along this surface and thus make the measurement signal out of range.
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_17, © Springer-Verlag Berlin Heidelberg 2010
157
158
H.P. Hoang and S.L. Ko
Fig. 1 Burrs formation at inclined exit surface
Fig. 2 Burrs observed at curved and inclined intersecting holes
angle of each ray in a complete solid light cone and analyzes the global behavior of all the rays while a triangulation system, for example, only measures the angle of a single ray. This procedure is more precise and stable. Another advantage of the method is that the incident beam and the reflected cone are coaxial and very narrow, make it possible to measure cavities and high slopes surfaces (up to 85◦ ) [7–9].
Fig. 3 Definition of exit surface angle [6]
2 Description of the System 2.1 Conoscopic Holography Sensor Conoscopic holography, invented by G.Y. Sirat and D. Psaltis in 1985, is a holographic technique based on light propagation effects in uniaxial crystals. A SCP, shown in part of the Fig. 4, consists of a birefringent crystal coupled with two circular polarizers. A single ray which emitted by projecting a laser diode source to the surface, are split by the first polarizer into two rays. These two orthogonally polarized components propagate inside the crystal at different velocities along almost the same geometrical path. The second polarizer (analyzer) recombines them creating interference which is detected by the CCD. The phase delay between the two components is dependent on the angle, which in turn depends on the distance to the point in space. The SCP measures the
Fig. 4 Conoscopic holography diagram (Source: Optimet Co.)
Burr Measurement System for Drilled Hole at Inclined Exit Surface Table 1 ConoProbe technical specification
Static resolution (μm) Lens 16 <0.1 Lens 25 <0.1 Source: Optimet Co.
159 Precision (μm)
Working range (mm)
Laser spot size (μm)
Angle measurement
Data rate (pps)
<1 <2
0.6 1.8
8 22
85◦ 85◦
850 850
The measuring range and precision of the probe can be adjusted by only changing the objective lens. Two lenses with the focal length of 16 and 25 mm, namely lens 16 and lens 25, are used in our system. Their performance characteristics are summarized in Table 1.
2.2 Measurement System The system is composed of both hardware and software components as in Figs. 5 and 6. The hardware part contains the SCP and a 3 axis translation stage to form a 3D measurement. The software part contains two programs developed in Visual C++ and OpenGL, one for controlling the machine and one for analysis and visualization of burr data.
2.3 Softwares The software is programmed with threading and messagedriven methods for handling events with the hardware. To
Fig. 5 Layout of system controllers
monitor the stage controller’s feedback, the software issues threads instead of the traditional use of endless loop, which almost freezes the computer. As the stage finishes one scan line, the thread function post a message that notices the software sending command to EC1000 to stop capturing data and transfer them from the controller’s buffer to memory. Hence, the user can freely interact with the online graphics and functions during scanning. The stage movements and scan parameters can be recorded as macro commands for measuring multiple samples. The software for system control and data acquisition is developed in a similar manner of commercial computer control softwares. All functions for managing the stage movements and adjusting the SCP’s parameters are laid out in the interface as in Fig. 7. By choosing proper setting values of sensor according to the property of surface and deciding the scanning parameter such as step length, line interval, speed, direction and area of scan, the control software can simply and quickly be used for both flat and inclined surfaces. Online 3D graphics and statistics information of the scan provide quick and intuitive feedback to the user.
160
H.P. Hoang and S.L. Ko
In former measurement system, the measuring range of the probe can be adjusted only by changing the objective lens. To automatically adjust the distance in Z axis and adapt for the different geometry boundaries of the measuring surface, curve following functions are developed in burr measurement control software. This function is applied to adjust the vertical distance between the sensor and measuring surface. By using this method, the sensor can follow curved and inclined surfaces, which cause many difficulties in scanning, and be capable to cover the wide range of Z axis. In analysis software, digitized data are plotted in 3D and 2D sectional graphs that display both surfaces. After analyzed, the 3D image is mapped with color according to the height of the burr, thus providing a clear visualization of the calculation result which is shown in Fig. 8. Parameters of burrs are calculated in the region where they exist to avoid other peaks that might occur in the scan. This rectangular area enveloping the hole or edge is detected automatically. Other interested regions can also be specified by simple dragging the computer mouse over the 3D image. The average and maximum height of burr and width of burr are to be calculated. The roughness of selected surface can be obtained. Output data can be exported to database format or to a text file that can be read and analyzed in other metrology software.
Fig. 6 Burr measurement system
Fig. 7 The control software
Burr Measurement System for Drilled Hole at Inclined Exit Surface
161
Fig. 8 The analysis software
3 Analysis of Measurement Data 3.1 Error Elimination To detect noises, we firstly threshold the Signal to noise ratio (Snr) value tagged with each measured point. As previously described, the points with nonzero of Snr lower than 45% are regarded errors. Most of them are out-of- range signals occurring as the probe scans across the hole and therefore are set to an equal height to visualize the depth of this hole. Others errors reside sparsely due to optical noises and appear as local valleys on the surface. Since the waviness filtering algorithm that we use greatly depends on the peaks and valleys, these fake valleys should be removed prior to that step. The filtering algorithm is simple. If an error is surrounded with high Snr points, it will be replaced with the ordinate of the least square plane passing through these points. Otherwise, it is inside the hole and remains unchanged as in Fig. 9. In rare cases, error appears in large area due to poor light setting or surface defects. The automatic function fails to remove fully these kinds of errors. However, we can remove them manually in a fast way. The software provides two tools for doing this. The first tool, as in Fig. 10a, is using an “eraser” – circle the error area with the computer mouse, all the defects inside will be replaced by the ordinates of the least square plane passing through the boundaries of that area,. The second tool as in Fig. 10b, using a “clipping – plane”, defines a plane to chop all nonsense or meaningless
Fig. 9 The automatic error elimination function
values. Though these manual functions are tedious, they are still necessary to guarantee the sample after being processed is free from errors.
3.2 Characterize Burr Geometry The profile after being filtered is composed of three components: the roughness, the burrs and the hole depth, set as z = z depth. To distinguish between burrs, regarded as protrusion peaks, and the roughness, a straightforward method is plotting the Abbott-Firestone bearing curve, a well-known tool in tribology engineering to describe the surface texture of an object. In physical term, the bearing area curve is generated by plotting a curve at different depths, thus each point on the bearing area curve has the physical significance
162
H.P. Hoang and S.L. Ko
Fig. 10 The manual error elimination function – using an “eraser”
of showing what linear fraction of a profile lies above a certain height. The burrs appear as non linear region on the left side and are determined by non zero values of the second derivative of the curve. This method, however, requires sorting the profile data, which we already tried to avoid since it is very timeconsuming. Therefore, we use a simpler way of taking advantage of the coefficients δk of the robust Gaussian regression filtering (RGRF) [10]. The zero value of δk combined with positive value of shorter wave length component (roughness and burr) at profile points could serve as an automatic threshold to determine the burrs quickly. Hence a point is considered as a burr when its height is approximately higher 4.5 times than the median of the base surface’s roughness. This criteria works well for most of the case of our experiments depends on whether the user pays more attention to micro or large burrs. In addition to burrs quantities, computation for the roughness is also essential since a deburing process not only change the burrs around the edge but the whole workpiece’s surface. A deburring method is considered effective if only it removes the burrs and does not degrade the surface quality.
Fig. 11 Burrs on inclined exit surface
Therefore, the system should measure the burrs and the surrounding roughness at the same time in order to evaluate the effectiveness of a deburring process. Calculated burr parameters are average burr height Ba , peak burr height Bp , average burr width Bw , volume of burr Bv . Calculated roughness parameters of the base surface are Ra , Rq , Rt , Rp , Rv and Rz . Note that the burrs are already excluded from the calculation of these roughness parameters.
4 Experiment Result Figures 11 and 12 display the 3D photo-realistic images of different burr formations that are measured by our system. The primary profiles (P-profiles) on left figures are first errorremoved raw data and the right figures are filtered by applying the RGRF with cutoff 0.8 mm to obtain the burr geometry. Burr shapes are reserved throughout the filtrations. The example in Fig. 13 demonstrates an application of burr measurement to inspect the result of a deburring process. Burrs are formed at two intersecting holes 1 and 2 of an
Burr Measurement System for Drilled Hole at Inclined Exit Surface
163
Fig. 12 Burrs at intersecting drilled holes
Fig. 13 Burr formed at crossing drilled holes (lens 25, area = 4 mm × 4 mm, x = y = 10 μm, scan time = 13 min, λc = 0.8 mm. In Fig. 13b, Ba = 41 μm, in Fig. 13c, Ba = 14 μm)
automotive engine block, which is difficult to be measured and deburred efficiently. The cylinder form and the waviness in the primary profile are filtered by applying the RGRF of 2nd degree, with cutoff value of 0.8 mm to get the result in Fig. 13b. Figure 13c displays the filtered profile of this sample after deburring by using a deburr tool along hole 2’s axis. As can be seen, burrs still remain, which comes to an initial conclusion that this tool cannot cut fully at the lowest positions. In the case of concave and convex surfaces, which are difficult to be measured, we choose the scanning area for flat and inclined surfaces. By following the main line along the
Fig. 14 Measurement and analysis results of convex surface
right side of the hole, it is measured successfully (Figs. 14 and 15). As a disadvantage of this measurement system, we need to notice about the quality and smoothness of the surface around this main line. If there is any defect or scratch, these kinds of defect will affect and reduce the quality of measurement signals and cause some errors on the surface of measured data. In another experiment, we test the ability of the system to measure the drilled holes at inclined exit surface at different angles from 15 to 45◦ . The diameter of holes is 9 mm, so scanning area is set at value of 12 by 12 mm. For the same quality of surface, different values of setting power must be
164
H.P. Hoang and S.L. Ko
Fig. 15 Measurement and analysis results of concave surface Fig. 16 Measurement results of drilled holes at 15, 30, and 45◦ of inclined exit surfaces
chosen when using conoscopic sensor to preserve enough energy reflected back to sensor due to the angle of these surfaces. Figure 16 shows the successful measurement of the burr on inclined surface up to 45◦ .
5 Conclusions In this work, we present a system for burr measurement on curved and inclined surfaces using the Conoscopic probe. The Conoscopic holography is proved to be the proper techniques for measuring the burrs since burrs contain steep and thin edges. The system is composed of both hardware and software components to digitize and analyze burrs in many cases, from millimeter to micrometer scale, from burrs on planar to curved and inclined surfaces. An advantage of the analysis software is that it separates and computes both the
burrs and the base surface roughness parameters. Results are displayed with color-coded 3D graphics, thus provides an intuitive verification of the analysis accuracy. In case of burr on curved and inclined surfaces, this system performs a good ability to measure successfully these difficult surfaces due to the limitation of sensor using Z axis adaptation. The conoscopic sensor can not cover all the measured ranges such as drilled hole at crossing intersection or deep inclined surface which are out of measurement range. Furthermore, in the next step of development, we will build up a master burr specimens which helpful for evaluating the accuracy of this measurement system. The burr width and height measured by our system can serve as decisive information for deburring processes. They are also useful for building up an expert system for burr minimization or burrs on complex surface. Acknowledgments This work is supported by Korean Research Foundation from 2007 to 2009.
Burr Measurement System for Drilled Hole at Inclined Exit Surface
165
References 6. 1. Ko, S. L., Dornfeld, D. A., 1991, A study on burr formation mechanisms, Transactions of the AMSE, Journal of Engineering Materials and Technology, Vol. 113, No.1, 75–87 2. Koelsch, J., 2001, Divining edge quality by reading the burrs, Quality Magazine, 24–28 3. Schwenke, H., Neuschaefer Rube, U., 2002, Optical methods for dimensional metrology in production engineering, Keynote paper of CIRP Annals, Vol. 51, 685–699 4. Gillespie, L., 1998, Inspecting for burrs, Manufacturing Engineering, Vol. 120, 70–76 5. Ko, S. L., Park, S. W., 2004, Development of an effective measurement method for burr geometry, Proceedings of 7th International
7. 8. 9. 10.
Conference on Precision Surface Finishing and Deburring Technology (PSFDT), USA Min, S. K., Dornfeld, D. A., Nakao, Y., 2003, Influence of exit surface angle on drilling burr formation, Journal of Manufacturing Science and Engineering, Vol. 125, No. 4, 637−644 Sirar, G., Psaltis, D., 1985, Conoscopic holography, Optics Letter, Vol. 10, 4 Malet Y., Sirat G. Y., 1998, Conoscopic holography application: multipurpose rangefinders, Journal of Optics, Vol. 29, 183–187. Optimet Co. Ltd, Conoprobe 1000 specification. To, M. H., Ko, S. L., 2006, Application of advanced filtering algorithm for measuring microburr, International e-conference on Computer Science (IeCCS)
Burr Measurement: A Round Robin Test Comparing Different Methods V. Franke, L. Leitz, and J.C. Aurich
Abstract Remaining burrs after machining pose a severe risk for components life, if the burrs get loose. To reduce or eliminate burrs several deburring technologies are applied. To choose a deburring system and to reveal the results of deburring it is necessary to be able to measure burrs. The results of a round robin test conducted within the working group of burrs within the CIRP (International Academy for Production Engineering) to compare different burr measurement systems are presented. Keywords Burr measurement · Milling · Drilling
1 Introduction Secure detection of remaining burrs in parts is an essential goal of production engineering investigations. Furthermore, measuring of burr geometry is necessary for any research with the aim to minimize or avoid burr formation. Currently, there is a large number of burr measuring and detecting methods available. The choice of an appropriate system depends on application conditions, requested measuring accuracy and burr values to be measured like burr height, burr thickness, burr volume or burr hardness. Burr height and thickness are the most frequently and easily measured burr values [1]. A German study named “SpanSauber”, [2], identified a certain lack of standardized measurement systems for burrs. Over 71% of the companies interviewed in the survey study SpanSauber still use – among other measuring methods – the fingernail test for burr detection. The results of the fingernail tests depend on subjective evaluation of the measuring personnel.
V. Franke (), L. Leitz, J.C. Aurich Institute for Manufacturing Technology and Production Systems, University of Kaiserslautern, P.O. Box 3049, 67653 Kaiserslautern, Germany e-mail:
[email protected] url: www.fbk-kl.de
2 Burr Values In most cases, burrs are defined as undesirable or unwanted projections of the material formed as a result of the plastic flow from cutting and shearing operations. A comprehensive definition can be found in [3]. A burr is a body created on a workpiece surface during the manufacturing process of a workpiece, which extends over the intend and actual workpiece surface and has a marginal volume compared to the workpiece, undesired, but to some extended, unavoidable. Schäfer [4] used a random cross section for describing basic burr parameters. He stated that each burr can be characterized by its longitudinal and cross sectional profile and defined the following burr descriptions and measurement categories.
• The burr root thickness bf is the thickness of the burr root area measured in the cross section. • The burr height h0 is defined by the distance between the ideal edge of the workpiece and the highest point in the cross sectional area. • The burr root radius rf as shown in Fig. 1 is determined by positioning a circle to the burr root. • The burr thickness bg describes the thickness parallel to the burr root area at a distance of rf , as measured in the cross section [4].
In most cases the longitudinal profile of a burr is not very informative, and therefore, it is rarely used to describe burrs. The length of the burr is of more interest because it describes how much of the total edge length exhibits a burr. To compare different measurement methods the participants of the Round Robin were asked to measure the mean burr height, the maximum and minimum burr height as well as the burr height course along the measured distance. Burr root thickness, burr root radius and burr thickness are difficult to measure or can not be measured applying non-destructive measurement methods.
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_18, © Springer-Verlag Berlin Heidelberg 2010
167
168
V. Franke et al. h0
vf br
rf
bg bf: burr root thickness rf: burr root radius bg: burr thickness h0: burr height u: effective exit surface angle
burr profile in longitudinal direction
thermore, measuring of burr geometry is necessary for any research with the aim to minimize or avoid burr formation. Currently, there is a large number of burr measuring and detecting methods available. The choice of an appropriate system depends on application conditions, requested measuring accuracy and burr values to be measured like burr height, burr thickness, burr volume or burr hardness. Burr height and thickness are the most frequently and easily measured burr values [1]. The large number of detection and measuring methods can be structured according to various criteria: • one-, two- or three-dimensional, • destructive or non-destructive, • with or without contact [1]. Furthermore, Leopold [1] divided measuring methods in two groups: in-process and out of process, Fig. 2.
burr profile in cross direction
3.1 Optical Systems
Fig. 1 Measurement values of a burr [4]
3 Burr Measurement Systems Secure detection of remaining burrs in parts is an essential goal of investigations in production engineering. Fur-
Various optical systems to detect or measure burrs are available. Camera systems, microscopes, laser and interferometer are among the most important optical systems. Camera systems and microscopes deliver images of the workpiece surface. Measured values are analyzed with
Methods of Burr Detection and Burr Measurement
Out of process
With contact
In-process
Contactless
Proces monitoring Force, moment Sound emission analysis
Stylus methods Metallographical profiles Electro-mechanical
Optical
Optical microscope Borescope/endoscope Scanning electron microscope
Fig. 2 Methods of burr detection and measuring [1]
Light-slit methods Laser triangulation Fring pattern projection Autofocus methods Confocale microscopy
Eddy-current sensor Inductive sensor Computer tomography
Burr Measurement: A Round Robin Test Comparing Different Methods
169
Fig. 5 Cross section of a burr Fig. 3 Digital fringe projection
the help of specific measurement software. Laser systems, interferometer or fringe projection (Fig. 3) interpret interference patterns gained by scanning the workpiece with a light source.
3.2 Tactile Systems Stylus methods are suitable to measure burr heights only. Figure 4 illustrates a scanned drilling burr profile. The real profile of the burr will be falsified because of the conical shape of the tracer point. To avoid this effect advanced calculation is necessary. Furthermore, to characterize non-uniform burrs a single measurement is not sufficient. Measurement methods with workpiece contact are always limited in their application range due to the workpiece material stiffness.
Burrs can be destroyed or deformed because of the contact forces.
3.3 Destructive Measurement For accurate burr analysis it is necessary to prepare metallographic cross sections of the workpiece. Using metallographic cross sections, Fig. 5, allows measuring overall burr values as defined by Schäfer [4]. In the cross sections burr hardness and structural changes in the material which result from the cutting process can be measured as well. Furthermore, it is the only method to measure burr length and burr thickness for rolled back and spiral burrs. However the preparation of metallographic cross sections is very time consuming and allows only the measurement at one specific workpiece position.
4 Round Robin
Fig. 4 Burr analysis applying stylus method
In 2006 within the CIRP a Working Group on Burrs was established. Its goal was to collect available research results about burrs and to foster research into the area of burrs, with a focus on burr analysis and control as well as on cleanability and burr removal. Discussions within the working group revealed a large interest in burr measurement systems. The question emerged if results of burr measurements are comparable. To answer this question the Round Robin on Burr Measurement was started. Seven different institutions participated. The following paragraphs describe the investigated workpieces and the applied measurement systems.
170
V. Franke et al.
4.1 Design of Workpieces
4.2 Applied Measurement Methods
Two workpieces have been prepared to measure burrs at specific positions. The workpiece material was AISI 4140H steel. The investigated burrs occurring during drilling and milling operations are further described below. To measure drilling burrs holes with two different diameters (6 and 12 mm) were drilled with a carbide twist drill. The applied coolant was 9% emulsion. Holes 1, 3, 5 and 7 (Fig. 6) have been drilled with external coolant supply with a cutting speed of 80 m/min and feed of 0.11 m/rev. Holes 2, 4, 6 and 8 have been drilled with internal coolant supply with a cutting speed of 100 m/min and feed of 0.15 m/rev. Further two exit angles were measured. The exit angles were 90◦ (holes 1, 2, 7 and 8) and 60◦ (holes 3, 4, 5 and 6). The cutting parameters for the drilled workpiece are summarized in Table 1. Different diameters and exit angles were chosen to enable a comparison of measurement results at different workpiece geometries. Figure 6 shows the different measurement positions. Another workpiece with burrs at tool entry and exit was milled. Workpiece material is AISI 4140H steel. A carbide milling cutter with a cutting speed of 90 m/min, feed per tooth of 0.04 mm/rev and a cutting depth of 2 mm was applied. The cutting parameters are summarized in Table 2. The exact positions for burr measurement can be seen at Fig. 7.
The participants of the Round Robin were free to choose any non-destructive measurement method. After all measurements were conducted, additionally cross sections were prepared for comparison. Six measurement methods were applied. All but one were optical systems, as a mechanical system the stylus method was applied. The different measurement systems will be described shortly in the following paragraphs. 4.2.1 Digital Fringe Projection Measurements with a digital fringe projection system are based on a triangulation process. A cos2 -shaped luminosity distribution is projected. The system enables height information not only in stripe position but also in grey tone. An equidistant stripe pattern is projected onto a surface, the measurement area or area of interest. The workpiece is observed under a certain angle (triangulation angle). Figure 8 illustrates the operating mode of the digital fringe projection. 4.2.2 Confocal Measurement System Depth of Focus (DoF) The system Depth of Focus is a confocal measurement system developed by GFE Schmalkhalden (Fig. 9). Virtual
Fig. 6 Prepared workpiece with burrs from drilling Table 1 Cutting parameters workpiece drilling
Table 2 Cutting parameters for milling
Material: Tool: Diameter: Coolant: Concentration:
AISI 4140H Carbide twist drill 6 mm, 12 mm Emulsion 9%
Material: Tool: Diameter: Coolant: Concentration:
Position Cutting speed: Feed: Coolant concept: Exit angle:
AISI 4140H Carbide milling cutter 14 mm Emulsion 9%
2, 4, 6, 8 100 m/min 0.15 mm/rev internal 2, 8 (90◦ ) 4, 6 (60◦ )
1, 3, 5, 7 80 m/min 0.11 mm/rev External 1, 7 (90◦ ) 3, 5 (60◦ )
Cutting speed: Feed per tooth: Depth of cut: Coolant concept:
90 mm/min 0.04 mm/rev 2 mm external
Burr Measurement: A Round Robin Test Comparing Different Methods Fig. 7 Prepared workpiece with burrs from milling
Fig. 8 Digital fringe projection (Source: GFM)
Fig. 9 Confocal measurement systems DoF developed by GFE Source: GFE Schmalkhalden
171
172
Fig. 10 Measurement assembly for burr measurement with confocal white light interferometer Source: ISF, Dortmund
V. Franke et al.
Fig. 11 Measurement assembly for burr measurement with non contact 3d measurement system Source: LMAS, Berkeley
equidistant defined images of the workpiece edge are created. The dimension of the measuring field is 740–740 μm.
4.2.3 Confocal White Light Interferometer The confocal white light interferometer enables a non contact 3D-charactarization of surfaces (Fig. 10). The measuring head moves in vertical direction and detects the confocal signal for each point. A stack of vertical sections is calculated and a 3-dimensional image of the height figures is generated.
4.2.4 Non-Contact 3D Measurement System The operating mode of the non contact 3d measurement system (Fig. 11) is similar to digital fringe projection. Burr height is identified when measuring the distance between zero point level and current defined level.
4.2.5 Stylus Profilometer The stylus profilometer (Fig. 12) was applied to measure burr height. The workpiece surface was scanned with a stylus. The stylus displacement provides information about height deviation along the measured distance. Fig. 12 Stylus profilometer
4.2.6 Metallographic Cross Sections The preparation of metallographic cross sections is a destructive burr measurement method. It has to be taken into account that burr measurement is only possible at specific positions and the preparation of the cross sections has a major
influence on measurement results. It is not possible to gain general information about the whole burr. Measured values depend on chosen measuring position. Figures 13 and 14 illustrate exemplarily the positions for milling and drilling burr measurement.
Burr Measurement: A Round Robin Test Comparing Different Methods
173
Fig. 13 Example of cross section of drilled hole
Fig. 14 Example of cross section of milling burr
5 Results of the Round Robin The following measured values serve as a base to compare different measurement systems:
M-5, and M-8 are all conducted with a digital fringe projection system from GFM. The application of one measurement system at different institutions enables a statement on comparableness of results measured with the same system.
• maximum burr height, • mean burr height.
5.1 Burrs from Drilling Operations
These data were analyzed for burrs from drilling and milling operations. The results are presented in the following paragraphs. The burr heights were detected with nine non destructive (five different systems) and one destructive measurement methods. The results of the cross sections are summarized in separate figures, as the cross sections only regard specific positions. These positions might not be the positions with maximum burr formation. The investigations M-2, M-3,
Figure 15 illustrates the maximum burr height measured at holes with an diameter of 12 mm (hole 1 and 2) and of 6 mm (hole 7 and 8), see Fig. 6, at an exit angle of 90◦ . Similar results for maximum burr height were detected for all measured holes with an exit angle of 90◦ . Mean burr height has not been measured at all institutions. If the mean burr height is declared, about 20–80 positions have been measured. Figure 16 illustrates the results of
Maximum burr height hole 1, 2, 7, 8 450 400 M-1
Burr height [µm]
350
M-2
300
M-3
250
M-4 M-5
200
M-6
150
M-7 M-8
100
M-9 50 0 Hole 1
Hole 2
Hole 7
Hole 8
Fig. 15 Maximum burr height of holes with a diameter of 12 mm (hole 1 and 2) and 6 mm (hole 7 and 8) at a 90◦ exit angle
174
V. Franke et al. Mean burr height hole 1, 2, 7, 8 160 140
M-1
Burr height [µm]
120
M-2 M-3
100
M-4
80
M-5 M-6
60
M-7
40
M-8 M-9
20 0 Hole 1
Hole 2
Hole 7
Hole 8
Fig. 16 Mean burr height of holes with a diameter of 12 mm (hole 1 and 2) and 6 mm (hole 7 and 8) at a 90◦ exit angle. Median of each measured hole illustrated with a black line
the mean burr height measured at holes with an diameter of 12 mm (hole 1 and 2) and of 6 mm (hole 7 and 8) at an exit angle of 90◦ . For better comparison the median for each hole of all non destructive measurement systems is marked with a black line. Three out of four (five) results reveal similar values. Figures 17 and 18 display the maximum und mean burr height for holes drilled at an exit angle of 60◦ . Hole 3 and 4 have a diameter of 12 mm, the diameter of hole 5 and 6 is 6 mm. For better comparison the median for each hole of all non destructive measurement systems is marked with a black line. The results reveal a larger variation in values in comparison to burrs at 90◦ exit holes. A feasible explanation for this deviation might be difficulties when measuring angular exits. The light source might have been adjusted differently when measuring the burr height.
Figure 19 illustrates different measurement positions of a light source. Figure 20 contains the measurement results of the cross sections of the drilled holes at position a and b. The location of the different positions is illustrated in Fig. 13. To enable a comparison between the results of non destructive and destructive measurements the median for each hole of all non destructive measurement systems is marked with a black line. A large variance between burr heights measured with non destructive measurement methods and with the help of cross sections is revealed. It has to be taken into account that cross sections only examine burrs at specific positions of a hole. Cross sections do not allow conclusions on the burr size along the examined area. The advantage of preparing cross sections is the possibility to measure several burr values and to take the micro structure of the workpiece into account. The micro structure has not been investigated here.
Maximum burr height hole 3, 4, 5, 6 450 400 M-1
Burr height [µm]
350
M-2
300
M-3
250
M-4 M-5
200
M-6
150
M-7
100
M-8 M-9
50 0 Hole 3
Hole 4
Hole 5
Hole 6
Fig. 17 Maximum burr height of holes with a diameter of 12 mm (hole 1 and 2) and 6 mm (hole 7 and 8) at a 60◦ exit angle
Burr Measurement: A Round Robin Test Comparing Different Methods
175
Mean burr height hole 3, 4, 5, 6 160 140
M-1
Burr height [µm]
120
M-2 M-3
100
M-4
80
M-5 M-6
60
M-7 40
M-8 M-9
20 0 Hole 3
Hole 4
Hole 5
Hole 6
Fig. 18 Mean burr height of holes with a diameter of 12 mm (hole 1 and 2) and 6 mm (hole 7 and 8) at a 60◦ exit angle. Median of each measured hole illustrated with a black line
surements the median for each position of all non destructive measurement systems is marked with a black line (Fig. 25). A large variance between burr heights measured with non destructive measurement methods and with the help of cross sections is revealed. It has to be taken into account that cross sections only examine burrs at specific positions and not the whole measuring distance.
6 Summary and Conclusions Fig. 19 Difficulties when measuring angular exits
5.2 Burrs from Milling Operations Figures 21 and 22 present the maximum and mean burr height of entry burrs from milling. Nearly all measurement methods detected the maximum burr height at position d, as expected due to the boundary conditions of the machining process. Measurements at position a, b and c provide a large variation in values. The small burr height impedes the measurement and analysis. The maximum and mean burr height of the exit burrs are displayed in Figs. 23 and 24. Due to the machining conditions the maximum burr was expected to be at position c. Eight out of nine measurements confirmed this. Again the measurement of small burrs impedes the results. A large variation of values can be detected. Cross sections of the different positions have been prepared of the milling burrs as well. To enable a comparison between the results of non destructive and destructive mea-
In these investigations a comparison of burr measurement at different institutions and with various measurement methods is presented. Burrs from drilling at a 90◦ exit angel reveal comparatively little variation in their results of measured burr height. The largest variation in burr height measurement occurred when measuring drilling burrs at an exit angle of 60◦ . A possible reason for this could be the workpiece positioning to the light source or the stylus. Small burrs cause difficulties when measuring them. The values show large variation. The burr height measured in cross sections differed from the results of non-destructive measurement methods. It has to be taken into account that cross sections only regard one specific burr position and not the whole burr. The results of the round robin test stress the necessity of standards on workpiece positioning for the measurements. Further there should be a standardized analysis algorithm for measured data to enable comparison. Very high caution is necessary if data of burr measurement is compared. The workpiece position has a large impact on the results of burr measurement. Further work is necessary to develop approaches for standardization burr measurements to enable comparison.
176
V. Franke et al. Burr height 450 400
Burr height [µm]
350 300 250
Position a
200
Position b
150 100 50 0 Hole 1
Hole 2
Hole 3
Hole 4
Hole 5
Hole 6
Hole 7
Hole 8
Fig. 20 Burr height of all measured holes (cross sections) at position a and b. Median of non destructive measurement of each hole illustrated with a black line
Maximum burr height entry burr 180 160 M-1
Burr height [µm]
140
M-2
120
M-3
100
M-4 M-5
80
M-6
60
M-7
40
M-8 M-9
20 0 Position a
Position b
Position c
Position d
Fig. 21 Maximum burr height of entry burrs of milled workpiece
Mean burr height at entry burr 140
Burr height [µm]
120
M-1 M-2
100
M-3 M-4
80
M-5 60
M-6 M-7
40
M-8 M-9
20 0 Position a
Position b
Position c
Position d
Fig. 22 Mean burr height of entry burrs of milled workpiece. Medial of each measured position illustrated with a black line
Burr Measurement: A Round Robin Test Comparing Different Methods
177
Maximum burr height at exit burr 180 160 M-1
Burr height [µm]
140
M-2
120
M-3
100
M-4 M-5
80
M-6
60
M-7
40
M-8 M-9
20 0 Position a
Position b
Position c
Position d
Fig. 23 Maximum burr height of exit burrs of milled workpiece
Mean burr height at exit burr 140
Burr height [µm]
120
M-1 M-2
100
M-3 M-4
80
M-5 60
M-6 M-7
40
M-8 M-9
20 0 Position a
Position b
Position c
Position d
Fig. 24 Mean burr height of exit burrs of milled workpiece. Median of each measured position illustrated with a black line
Burr height 140
Burr height [µm]
120 100 80 60 40 20 0 Groove 1-a Groove 1-b Groove 1-c Groove 1-d Groove 2-a Groove 2-b Groove 2-c Groove 2-d
Fig. 25 Burr height of all measured positions (cross sections) of milled workpiece. Medial of non destructive measurement of each position illustrated with a black line
178 Acknowledgements We would like to thank all the participants of the round robin: F. Barthelmä (GFE Schmalkhalden, Germany), D. Biermann (ISF Dortmund, Germany), E. Brinksmeier (IWT Bremen, Germany), C. Harzbecker (IWU Chemnitz, Germany), D. Dornfeld (LMAS Berkeley, USA), S.-L. Ko (Konkuk University, Korea).
References 1. Leopold, J., Schmidt, G., 2004, Methods of Burr Measurement and Burr Detection, VDI-Berichte, 1860:223–229 2. Aurich, J. C., 2006, SpanSauber, Untersuchung zur Beherrschung der Sauberkeit von zerspanend hergestellten Bauteilen, Ergebnis-
V. Franke et al. workshop, Lehrstuhl für Fertigungstechnik und Betriebsorganisation, Technische Universität Kaiserslautern 3. Beier, H. M., 1999, Handbuch Entgrattechnik: Wegweiser zur Gratminimierung und Gratbeseitigung für Konstruktion und Fertigung, Hanser Verlag 4. Schäfer, F., 1975, Entgraten, Krausskopfverlag, Mainz
Deburring Processes – Fundamentals
Deburring with CO2 Snow Blasting E. Uhlmann, M. Kretzschmar, F. Elbing, and V. Mihotovic
Abstract Injection molding is one of the most common production processes for polymer products. The geometries of the moulded parts are more and more complex, which leads to multipart casts and therefore to more burrs. Deburring plastics by using blasting with solid carbon dioxide (CO2 ) was investigated in this paper. Solid CO2 has a low hardness and therefore it is suitable for machining sensitive surfaces. The investigations have shown that deburring plastic parts by CO2 snow blasting is an alternative to common deburring methods. The achieved results were satisfying and proved to have a high potential for industrial use. Further development takes places in a public funded research project. The goal is to provide a market ready system and to establish this system in the next three to four years. Keywords Deburring · Thermosets · Dry ice · CO2 snow blasting
materials. When deburring for example plastics they reach a limit due to the material properties and the geometry of the parts. For mechanical deburring, plastic parts are often not hard enough or too small to be treated with conventional tools. Plastic components substitute more and more metallic components, for example in the transmissions. Injection molding is, as one of the most common production processes in the plastics industry, applied for an efficient production. For the bulk of these parts it is necessary that they are free of burrs, especially for parts which are installed in transmissions. To prevent the formation of burrs during the manufacturing process a disproportional effort is required. Hence, deburring after manufacturing is the most common procedure. The goals are to keep the accuracy of shape and size, the avoidance of harmful dust and vapors and also to maintain the cycle time of the manufacturing process. To achieve these goals an alternative deburring method such as blasting with solid carbon dioxide can be used.
1 Introduction An efficient production is characterized by high labor productivity and low production costs. To reach these goals the process chains will need to be increasingly automated. The generation of burrs cannot be avoided in automated production, especially for light metals and plastic parts [1]. Investigations on the mechanisms of burr formation, characterizing burrs and also technical options to prevent burr formation were conducted in recent years. By now deburring methods for most burr types have been established. These deburring methods can be automated for most of the burrs and many
2 Fundamentals of Blasting with Solid Carbon Dioxide Blasting with solid carbon dioxide is gaining in importance as a machining technology for industrial purposes. With solid carbon dioxide, an innovative blasting agent is used which possess outstanding properties. The resulting advantages are unique compared to the conventional blasting agents.
2.1 Properties of Carbon Dioxide E. Uhlmann, M. Kretzschmar (), V. Mihotovic Institute for Machine Tools and Factory Management (IWF), Technical University Berlin, Office PTZ 1, Pascalstr. 8-9, 10587 Berlin, Germany e-mail:
[email protected] url: www.iwf.tu-berlin.de F. Elbing CryoSnow GmbH, Zitadellenweg 20e, 13599 Berlin, Germany
Carbon dioxide (CO2 ) is an odorless, colorless and nontoxic gas. Furthermore it is electrically non conductive, chemically inert, non-flammable and bacteriostatic [2]. The specific characteristics of CO2 become apparent by an examination of the phase diagram, Fig. 1 [3]. Carbon dioxide can only be in
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_19, © Springer-Verlag Berlin Heidelberg 2010
181
182
E. Uhlmann et al.
Fig. 1 Phase diagram of carbon dioxide [3]
the gaseous or the solid state at ambient pressure, where the solid state is referred to as CO2 snow or dry ice. The phase change directly from the solid to the gaseous state is termed as sublimation. This property is the major advantage of these blasting processes, because only the removed material has to be disposed of. The carbon dioxide pressed to dry ice pellets has a hardness of 1.5 Mohs, which corresponds to gypsum [4]. The hardness of CO2 snow is even lower. Blasting with solid carbon dioxide is therefore suitable for machining sensitive surfaces without causing damage [5, 6]. On the one hand carbon dioxide is nontoxic, but on the other hand it supersedes oxygen bearing the risk of asphyxiation particularly in enclosed areas. At a CO2 concentration of 3–5 vol.% the heart rate and breathing can be affected and it may cause dizziness and headache. At a concentration of more than 7 vol.% it may cause unconsciousness and with rising concentration it can leads to death [3]. The carbon dioxide, which is used in industry and also for dry ice blasting and CO2 snowblasting, originates in Germany with one-third produced from natural sources. The remaining two-thirds are a by-product of chemical processes. Producing carbon dioxide exclusively by burning fossil fuels was a common practice until the 1970s. The carbon dioxide for technical applications is not especially produced for this purpose, but rather already existing CO2 is used, meaning the process is environmentally neutral. So the use of the CO2 as a blasting agent is a beneficial step between production and release of this CO2 and hence it has no additional effect on
global warming [7]. Liquid carbon dioxide is stored either in low pressure tanks (LPT) at 20 bar and –20◦ C or in high pressure tanks at 57 bar and 20◦ C. The storage time of liquid carbon dioxide is almost unlimited, whereas for dry ice storage is limited to only a few days.
2.2 Working Mechanism CO2 snow blasting is an air blasting process. According to this, the blasting agent is pneumatically accelerated on the surface of the part to be machined. Compressed air blasting processes use a mechanical effect to treat the surfaces. This effect is caused by the mechanical impact of the blasting agent on the surface of the parts. CO2 snow blasting uses two additional effects, the thermal and the sublimation effect. The thermal effect is a result of the low temperature of solid carbon dioxide of approx. –78.5◦ C. The surface of the substrate is cooled down locally and immediately. Therefore the adhesive forces in the material are reduced and the contamination or burr will separate. The CO2 snow sublimates while impacting on the surface. This leads to a rise in the mechanical effect due to the pressure impulse as a result of the volume expansion (600 times) during the sublimation. After the sublimation the carbon dioxide is in the gaseous state and released into the atmosphere. Therefore only the removed material has to be disposed of. Thus disposal or
Deburring with CO2 Snow Blasting
Fig. 2 Mechanisms of blasting with solid carbon dioxide, from left: thermal, mechanical and sublimation effect
recycling costs, which occur for example when using sandblasting, can be avoided. Another effect, which is presently still a hypothesis, is described as the solvent effect. In the supercritical state CO2 has properties of a solvent. So it is assumed that during the impact on the surface the necessary temperature and pressure conditions can be achieved [8–12]. Figure 2 shows the three effects, the thermal, the mechanical and the sublimation effect, during a cleaning or decoating process.
2.3 Process Variants Blasting with carbon dioxide can be basically divided into two process variants: blasting with dry ice supply and blasting with liquid CO2 supply. In the first case preformed dry ice pellets, which are produced in a separate process, are used as blasting media. The dry ice pellets are added in blasting devices to the compressed air stream and accelerated within nozzles on the surface to be machined. By blasting with the
Fig. 3 Process variants of CO2 snow blasting
183
liquid CO2 supply, the CO2 is feed in liquid condition to the blasting system. Due to the expansion of the pressurized CO2 it cools down, as a result of the Joule-ThomsonEffect, so that a certain fraction goes into the solid state, which is called CO2 snow. The CO2 snow is then accelerated by compressed air within the blasting nozzle. The expansion and also the generation of the CO2 snow particles can be divided into two variants, expansion into ambient pressure and expansion into blasting pressure. For expansion into ambient pressure a two-component concentric nozzle is used, Fig. 3a. The liquid carbon dioxide expands at the outlet of the nozzle. CO2 snow crystals are formed and accelerated by the compressed air jacketed jet. The expansion of liquid CO2 into blasting pressure takes place in a agglomeration chamber in front of the real blasting nozzle, Fig. 3b. Larger particles can be produced, which have a higher abrasiveness compared to the method mentioned before. A further variant is a blasting device, which was developed at the IWF of the Technical University Berlin. This blasting nozzle has a adjustable agglomeration chamber, Fig. 3c. The expansion takes place inside the blasting device, but separated from the blasting pressure. The generated particles are transported by a concentric air jet to the blasting nozzle. The particles are accelerated inside the nozzle. Beside these process variants, there are some further devices for blasting with solid carbon dioxide. An important device of those variants uses also an agglomeration chamber. Inside this chamber the CO2 snow particles will be strengthen. The CO2 snow particles are accelerated by compressed air within the blasting nozzle on the surface to be treated, Fig. 3d. This blasting device was used for the investigation for this paper.
184
E. Uhlmann et al.
3 Motivation The material properties of plastic components are the limiting factor for the applicable deburring method. For deburring elastic materials a significant technical effort is necessary. Often only a poor cutting quality can be achieved, because elastic parts deform under the pressure of the deburring tools. For this reason cryogenic media, e.g. liquid nitrogen, is used for an embrittlement of the burrs. The embrittled burr is then either removed by a process similar to the vibratory grinding or by a wheel blasting process with durable blasting media [13]. On the one hand it is not possible to cool only the burr rather the whole part, which potentially leads to damage the parts. On the other hand the deburred parts have to be cleaned afterwards. Furthermore some plastics have a glass transition temperature at ambient temperature, so that a further embrittlement by using cryogenic media is not possible. Another deburring method for plastic parts is deburring with infrared radiant heaters, but this process is also limited by the material properties. Deburring by using infrared heaters can only be used for materials which can be melted without decomposing itself [14]. For these reasons deburring plastic parts by blasting with solid carbon dioxide, especially CO2 snow blasting, should be qualified. The advantages of the CO2 snow blasting can be traced back to properties of the blasting agent. Due to the working mechanism, mentioned above, the range of materials for deburring can be enlarged. The mechanical effect works independent from the material properties. The thermal effect is contributory especially in deburring thermoplastics. In contrast to sandblasting, where the blasted parts have to be cleaned and the sand has to be separated from the burrs, the CO2 particles sublimate and no blasting media remains. Due to the sublimation the treated parts don’t have to be cleaned or dried.
4 Experimental Analyses The investigations in this paper were conducted on plastic components made of thermosets. These components are parts of an automatic transmission and therefore they have to be free of burrs to avoid damage to the gearbox. The parts were manufactured by injection molding. The manufacturer of these parts is the Christophery Kunststofftechnik GmbH, Brilon, Germany as a supplier of the ZF Friedrichshafen AG, Friedrichshafen, Germany. The parts are axis symmetric and show two burrs on the lateral area due to the mold parting line, Fig. 4. The tests were conducted at the Institute for Machine Tools and Factory Management (IWF) of the Technical University Berlin. A prototypical test station was set up to rotate
Fig. 4 Part with burrs
the parts. As blasting device a CryoSnow JP 25 blasting pistol was chosen, the function is described in Sect. 2.3. The blasting nozzle was orientated concentric and orthogonal to the rotation axis of the parts to be deburred. For the investigations the following parameters were chosen: • • • •
rotational speed, n = 1, 2 s–1 number of treatment per burr = 10, 20 blasting pressure, pb = 10, 13 bar mass flow of CO2 , m ˙ CO2 = 90 kg/h
For both blasting pressures the burrs were treated 10 and 20 times at different rotational speeds. This leads to a certain blasting time depending on the rotational speed. So the influence of the number of passes trough the blasting jet and the relative speed between the CO2 snow jet and the burr can be determined. These values are important for an efficient deburring in industrial applications. To determine the surface quality the three dimensional measuring system MicroProf by Fries Research & Technology GmbH (FRT), Bergisch Gladbach, Germany was used. This method allows measuring a surface without any contact and scanning the topography with a high resolution. Besides the topography also the roughness and the contour can be measured. In z-direction an accuracy of up to a few nanometers can be achieved. The height of the burr (hb ), the burr root thickness of the burr (tb ) and the arithmetical mean deviation of the assessed profile Ra was chosen as measured values. The Ra value was measured to determine a possible damage of the surface, because the whole part is exposed to the CO2 snow jet during the deburring process. The measuring range, Fig. 5, was scanned with the MicroProf before and after the treatment with CO2 snow blasting. The resolution of the scan was 300 × 300 points within the measuring range. The height and
Deburring with CO2 Snow Blasting
185
Fig. 7 Topography before treatment Fig. 5 Measuring range
the thickness of the burrs were measured at five positions of the burr and then averaged. Also the roughness was measured at five positions besides the burrs and averaged, too. For each parameter three parts were machined and measured.
5 Results In Fig. 6, a part after the deburring process is displayed. It can be seen that the burr is substantially reduced. In comparison to Fig. 4 the thin burr in the flute is removed completely. The topography of the burrs is exemplary shown as 3D pictures in Figs. 7 and 8. Figure 7 show the burr before treatment. In Fig. 8 the burr treated 20 times with a blasting pressure of 13 bar at a rotational speed of 1 s–1 , is displayed. It can be seen that the height of the burr is reduced significantly,
Fig. 6 Machined part
Fig. 8 Topography after treatment
but the burr root thickness of the burr is almost the same. The burr root thickness varies, before as well as after treatment, in a small range of about 10 μm. This variation is a result of the manufacturing process. The burr root thickness remains almost constant as long as the burr is not completely removed. The results of the analysis of all topographies concerning the height of the burrs are shown in Fig. 9. It can be seen, when blasting with 10 bar blasting pressure the rotational speed has no effect on the height of the burr. As expected, the height for 20 times treatment is decreased. When blasting with a pressure of 13 bar, the rotational speed has to be taken into account. At a rotational speed of 2 s–1 , the height of the burrs is considerably higher than at a rotational speed of 1 s–1 . This is caused by the higher relative speed between the blasting jet and the burr. Nevertheless the height is reduced from 99.4 to 43.7 μm at 13 bar blasting pressure, treated 20 times at a rotational speed of 1 s–1 .
186
E. Uhlmann et al. 120 10 x treated
20 x treated
10 bar
μm 10 s
5s
20 s
10 s
13 bar Parameter:
80 burr height hb
untreated
mco2 = 90 kg/h αB = 90°
60
Blasting equipment : CryoSnow JP 25
40
20
0 1
2 1 rotational speed n in s–1
2
Fig. 9 Burrs heights depending on the rotational speed and number of treatment 120 10 x treated
20 x treated
μm
mean roughness value Ra
80
untreated 10 bar
10 s
5s
20 s
13 bar
10 s
Parameter: mco2 = 90 kg/h
60
αB = 90° Blasting equipment :
40
CryoSnow JP 25
20
0 1
2 1 rotational speed n in s–1
2
Fig. 10 Mean roughness value depending on the rotational speed and number of treatment
The mean roughness value Ra of the machined surfaces is displayed in Fig. 10. When blasting with a pressure of 10 bar no substantial difference compared to the untreated roughness can be seen. No significant difference is distinguishable when blasting with 13 bar and a 10 times treatment. With a treatment 20 times and 13 bar blasting pressure, a noticeable rise in roughness is visible. This is in correlation with the height of the burrs. Due to the higher abrasiveness at 13 bar, the height of the burr can be reduced significantly but the roughness rises noticeably.
The achieved mean roughness value is within the requested tolerance.
6 Conclusion and Outlook The investigations in this paper have shown that deburring thermosets with CO2 snow blasting is feasible. Therefore it is an alternative to conventional deburring methods. The height of the burrs was reduced significantly, showing the
Deburring with CO2 Snow Blasting
high potential for CO2 snow blasting as a deburring method. After machining the mean roughness value is higher, but only for some parameters. The achieved mean roughness values were always within the specified tolerance. In a prospective research project CO2 snow blasting will be developed into an alternative deburring system. Therefore the process parameters have to be optimized. This includes for example the blasting pressure, positioning of the nozzle to the work piece, the rotational speed of the parts, CO2 mass flow etc.. Furthermore an appropriate handling system will be designed, for bulk goods and also for single parts. For process monitoring and quality control an inline monitoring system will be developed. The machine development is focussed on the reduction of noise emission and designing a concept for discharging air. This research project has the goal to provide a market ready system which will be established on the market in approx. three to four years.
References 1. Beier, H. M.; 1999, Handbuch der Entgrattechnik. Wegweiser zur Gratminimierung und Gratbeseitigung für Konstruktion und Fertigung. Carl Hanser Verlag, München, Germany 2. Uhlmann, E.; Stahl, U.; Veit, R.; Kristan, G.; 2006, Kalt gestellt – Keimfreie Oberflächen durch Strahlen mit Trockeneis. Pharma + Food, 9:52–54
187 3. N.N.; 2004, Eigenschaften der Kohlensäure. Fachverband Kohlensäure-Industrie e.V. 4. Krieg, M.; 2008, Analyse der Effekte beim Trockeneisstrahlen, Dissertation, Fraunhofer IRB Verlag 5. Uhlmann, E.; El Mernissi, A.; Hollan, R.; 2006, Adhäsive Oberflächeneigenschaften – Was bringt Trockeneisstrahlen? 2:44–47 6. Elbing, F.; Uhlmann, E.; Hühns, T.; Nolze, G., 2003, Verbesserung der Oberflächeneigenschaften von Metallen mit Trockeneisstrahlen. Galvanotechnik, 94(1):75–82. 7. Krieg, M.; 2007, Industrielle Nutzung von Kohlendioxid, Futur – Vision Innovation Realisierung, 2:14–15 8. Krieg, M.; 2005, Trockeneisstrahlen – mit Schnee oder mit Pellets?, JOT – Journal für Oberflächentechnik, 45(6):50, 52–55 9. Redeker, C.; 2008, Abtragen mit dem Trockeneisstrahl., Dissertation, Fortschr.-Ber. VDI Reihe 2, Nr. 639, VDI Verlag Düsseldorf 10. Haberland, J.; 1999, Reinigen und Entschichten mit Trockeneisstrahlen – Grundlegende Untersuchung des CO2 -Strahlwerkzeuges und der Verfahrenweise. Dissertation, Fortschr.-Ber. VDI Reihe 2, Nr. 502, VDI Verlag Düsseldorf 11. Uhlmann, E.; Spur, G.; Elbing, F.; 2002, Mehr als nur eine Reinigungsverfahren - Möglichkeiten des Trockeneisstrahlens. Metalloberfläche 56(4):14–18 12. Uhlmann, E.; Elbing, F.; 2002, Trockeneisstrahlen von Aluminiumoberflächen – Zur Festigkeitssteigerung von Klebverbindungen. Zeitschrift für wirtschaftlichen Fabrikbetrieb ZWF 97(3): 102–107 13. Plömacher, E.; 1990, Cryogenes Entgraten – die Ergänzung zur gratarmen Formteilherstellung. Das Elastomerwerkzeug. VDI-Verlag, pp. 117–127 14. Bopp, M.-L.; 2006, Kontaktfrei thermisch entgraten. In: Kunststoffberater, 51/7/8:31–33
A Study on Deburring Inconel 718 Using Water Jet Technology F. Boud, J. Folkes, N. Tantra, S. Kannan, and I.W. Wright
Abstract One of the most significant problems encountered in machining is that of burr formation. All edges must be completely defect-free, with generous radius, for operational, safety and aesthetic reasons. The machining operation or CNC grinding process that shapes the component produces burrs and sharp edges. The deburring process is intended to remove these imperfections and produce specific edge profiles. Disadvantages of manual deburring methods include the cost of labor, time taken to deburr the component, quality and health and safety issues. A promising method for deburring is the use of water jet technology. Water jet machining is now providing a simple and economical approach to selectively removing burrs from areas that are difficult to access with consistent results and absolute uniformity. This paper is aimed at evaluating the process of water jet deburring of Inconel 718. Trials conducted to study the effect of water jet deburring process proved to be effective. Keywords Deburring · Water jet technology · Inconel 718 · Burr control
1 Introduction Increasingly, manufacturers are expected to deliver burr-free parts to the point of use. Large quantities of material removal usually require high depths of cut which are undertaken using powerful machine tools. Once a part has been machined, a finishing operation is usually required to perform small material removal to bring the part into tolerance of the specification. Often, as a result of machining, burrs are generated
on the edges of machined surfaces. These burrs need to be removed using a finishing technique termed as deburring. In the machining industry manual methods are commonly employed for burr removal. Fully automating the deburring operation is challenging. Further, removal of internal burrs of various sizes and shapes sometimes becomes an extremely difficult task. In such situations water jet deburring offers a potential solution to the problem. Water jet machining, one of the most recent nontraditional methods, uses a jet of high pressure and velocity water which can be impinged on materials resulting in material removal. The water jet process provides many unique capabilities and advantages that can prove to be very cost effective. Beyond cost cutting, the water jet process is very versatile and can be used in high production applications. High-pressure water deburring has a number of advantages over other deburring processes, the first and foremost being that the part is clean and free of machining media remnants (such as grit from grit blasting) after deburring. Manual deburring results in inconsistent quality, the work is often labor intensive, and internal features are very often difficult to reach. Even when deburred, the part still needs to be cleaned. With robotic deburring using grit blasting, again, internal features cannot always be reached, very small chips cannot be removed with total certainty, and, again, parts still need to be post processed (i.e., checked for the presence of grit and processed for residual grit removal). Water jet does solve many of the problems where burrs are either difficult to reach or amenable to automation. It can deburr miniature parts as well as large parts, and does not result in the generation of heat affected zones.
2 Water Jet and Burrs F. Boud (), J. Folkes, N. Tantra Department of Mechanical Materials and Manufacturing Engineering, University of Nottingham, Nottingham, NG7 2RD, UK e-mail:
[email protected] S. Kannan, I.W. Wright Rolls-Royce plc, Derby, DE24 8BJ, UK
It is generally accepted that burrs refer to material projections located along the edges of a component. Burrs can pose serious consequences as the appendages, when loosened from the component at a later stage, can cause major damage to the device or system.
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_20, © Springer-Verlag Berlin Heidelberg 2010
189
190
According to Lee and Ko [1] burrs are unnecessary byproducts produced by plastic deformation after the cutting phase in the machining process. A burr is formed on component edges as a result of machining, which produces undesirable edge geometry [2]. Burrs are encountered in practically all manufacturing processes where the material removal is by shearing or with a heat source. In shearing processes the sharpness of the cutting tool is a major factor in the formation of burrs. Other parameters such as the cutting speed, feedrate and work materials also influence the prevalence of burrs. The specifications for components are becoming more stringent with respect to tolerances and surface finish. Burrs reduce the accuracy of the parts and subsequent assembly processes. Typically deburring accounts for up to 25% of the total production cost [3]. To reduce or even eliminate the deburring effort, the burr size must be reduced [4]. The deburring of machined components is a major bottleneck in most manufacturing organizations. Burrs cannot be eliminated completely at the machining stage itself but can only be minimized by controlling various machining parameters. Hence deburring becomes an essential secondary operation. Deburring includes both the removal of burrs and maintenance of the edge quality. Most of the deburring techniques ensure the removal of the burr but do not ensure the controlled edge quality. Moreover, these techniques may cause a change in the dimension of components, distortion of the components and other side effects. Controlled and consistent edge quality can improve product performance and life. In recent times, non-conventional techniques have been gaining increasing acceptability for deburring applications [5, 6]. Deburring of external burrs is comparatively easier, owing to accessibility and the fact that the quality of the exter-
Fig. 1 Experimental set up
F. Boud et al.
nal deburring can easily be ensured by visual inspection. However, deburring of inaccessible areas is challenging due to the difficulty of the deburring tool in approaching the deburring area, lack of space for positioning the deburring tool and non-visibility of the burr area. Deburring of inaccessible areas via conventional methods does not ensure burr removal and edge conditioning; deburring via nonconventional techniques provides a better solution. Electrical discharge machining (EDM), electrochemical machining (ECM) and abrasive jet machining (AJM) are some of the commonly employed non-conventional machining techniques [7]. The water jet process provides many unique capabilities and advantages that can prove very effective in the cost battle. Water jets are used in high production applications across the globe. They compliment other technologies such as milling, laser, EDM, plasma and routers. Water jets do not generate noxious gases or liquids and do not create hazardous materials or vapors. No heat effected zones or mechanical stresses are left on a water jet cut surface. It is truly a versatile, productive, cold cutting process.
3 Experimental Set-Up The experiment was conducted on a five axis water jet machine supplied by Ormond LLC equipped with a KMT Streamline SL-V 100S Plus pump and a Roctec 100 Nozzle with a length of 76.2 mm, and an inside diameter of 1 mm, the orifice diameter was 0.35 mm. The experimental set up is illustrated in Fig. 1, where the different features of the table are shown: the clamps on the
A Study on Deburring Inconel 718 Using Water Jet Technology
191
Fig. 2 Workpiece
workpiece, the angle of the nozzle is shown at 30◦ with the jet stream of water. Inconel 718 rectangular test blocks, of equal sizes were used with a burr produced at 0.25 and 0.5 mm depth of cut for each one. The workpiece with its dimensions can be seen in Fig. 2. The marks made by the water jet are shown by the dark areas in the figure.
4 Experimental Procedure The aim of the experiment is to deburr a workpiece with a burr produced using a 0.25 and 0.5 mm depth of cut using Water jet. Various parameters were used in initial tests in order to set the parameters that will be chosen for the final tests, and different burr sizes were used in order to establish the possibility of removing burrs by water jet. An example of some of the parameters used can be seen in Table 1. It was concluded to use the following parameters: pressure, standoff, speed and angles. The conditions for all the deburring tests with burrs produced at 0.5 and 0.25 mm can be seen in Table 2. All the
Table 2 Deburring Experiment conditions using burrs produced at 0.5 and 0.25 mm depth of cut 0.5 mm depth of cut 0.25 mm depth of cut Test No.
Speed (mm/min)
Angle (deg)
Test No.
Speed (mm/min)
Angle (deg)
23 24 25 26 27 28 29 30 31 32 33 34
1000 1000 500 250 150 150 150 75 75 75 150 500
30 30 30 30 30 30 30 30 30 30 30 30
35 36 37 38 39 40 41 42 43 44 45 46
500 250 250 500 1000 750 500 250 500 500 250 750
30 30 30 30 30 30 90 90 90 90 90 30
experiments were performed using water only at a pressure of 350 MPa (50,000 psi), standoff of 50 mm with either 30◦ or 90◦ angles and varied speed between 75 and 1000 mm/min.
5 Results and Discussion Table 1 Deburring experiment conditions using varied parameters and depth of cut Test Speed Angle Test Speed Angle No. (mm/min) (deg) No. (mm/min) (deg) 1 2 3 4 5 6 7 8 9 10 11
500 1000 2500 5000 500 500 1000 2500 1000 2500 500
90 90 90 90 30 30 30 30 30 30 30
12 13 14 15 16 17 18 19 20 21 22
1000 500 1000 500 1000 500 100 50 500 500 500
30 30 90 90 30 30 30 30 30 30 30
Feasible parameters that were considered most relevant for the experiments were identified as is usual to do so [8]. The photo and the laser gauge trace of the burrs (before deburring) are shown in Fig. 3 where the burr is produced from 0.5 mm depth of cut. Figure 4 shows the photo and the laser gauge trace of the burrs (before deburring) where the burr is produced from 0.25 mm depth of cut. For the burr produced at 0.5 mm, it can be seen from the photo in Fig. 3 that the height of the burr is approximately 475 μm. The laser gauge trace, as shown in Fig. 3, shows the height of the burr as approximately 500 μm. While the size of the burr as shown in Fig. 4 for the burr produced from
192
Fig. 3 Photo and laser gauge trace of Burr from 0.5 mm depth of cut
Fig. 4 Photo and laser gauge trace of Burr from 0.25 mm depth of cut
F. Boud et al.
A Study on Deburring Inconel 718 Using Water Jet Technology
0.25 mm is approximately 285 μm, and the height of the burr is approximately 300 μm. The surface roughness (Ra and Rv) were measured for the workpieces with the burrs that were produced at 0.5 and 0.25 mm depth of cut. The workpiece with the burr produced at 0.5 mm was found to have Ra of 0.98 μm and Rv of 5.38 μm. For the workpiece with the burr produced at 0.25 mm, the Ra was 0.92 μm and the Rv was 3.41 μm. Balasubramaniam et al. [6] found, in abrasive water jets, that the stand-off distance is the statistically most significant factor on the extent of edge radius generated. As the stand-off distance increases, the size of the edge radius increases. The edge radius is generated due to the erosion caused by reflected particles. The trajectories of the reflected particles change with the changing penetration depth. The photo in Fig. 5 shows a photo and laser gauge trace of the deburred area of 0.5 mm depth of cut, at test 33. The conditions were by using water at pressure of 350 MPa (50,000 psi), standoff of 50 mm, at speed of 150 mm/min and at nozzle angle of 30◦ with an area scan of 0.3 mm overlap. The photo in Fig. 6 shows photo and laser gauge trace of the deburred area of 0.25 mm depth of cut, at test 37. The conditions were by using water at pressure of 350 MPa (50,000 psi), standoff of 50 mm, at speed of 250 mm/min
193
and at nozzle angle of 30◦ with an area scan of 0.3 mm overlap. The photo in Fig. 7 shows photo and laser gauge trace of the deburred area of 0.25 mm depth of cut, at test 42. The conditions were by using water at pressure of 350 MPa (50,000 psi), standoff of 50 mm, at speed of 250 mm/min and at nozzle angle of 90◦ with an area scan of 0.3 mm overlap. From tests 33, 37 and 42, it is demonstrated that the water jet can be used to deburr as verified by the photos and the laser gauge trace of the burrs in Figs. 5, 6 and 7. After deburring the surface roughness of the workpieces were measured again and the results for the 3 tests (33, 37 and 42) were recorded, and can be seen in Table 3, all conditions were as previously mentioned for the tests. For test 33 the Ra was 2.39 μm while the Rv was 9.29 μm. Whereas the Ra for test 37 was 1.72 μm and the Rv was 3.91 μm. The Ra for test 42 was 3.65 μm and 9.33 μm for the Rv. The increase in surface area is where the area has been machined. These results seem to be affirmative with Balasubramaniam et al.’s [5] conclusions that abrasive jet deburring has the advantage over manual deburring methods of generating an edge radius automatically which increases the quality of the deburred components.
Fig. 5 Photo and laser gauge trace of deburred area (0.5 mm depth of cut, at nozzle angle 30◦ at test 33)
194
Fig. 6 Photo and laser gauge trace of deburred area (0.25 mm depth of cut, at nozzle angle 30◦ at test 37)
Fig. 7 Photo and laser gauge trace of deburred area (0.25 mm depth of cut, at nozzle angle 90◦ at test 42)
F. Boud et al.
A Study on Deburring Inconel 718 Using Water Jet Technology Table 3 Surface roughness results (Workpiece with burr from 0.5 mm depth of cut and workpiece with burr from 0.25 mm depth of cut) Workpiece with 0.5 mm depth of cut Test No. Pre-water 31 32 jet Ra (μm) 0.98 2.76 3.14 Rv (μm) 5.38 7.97 8.09
33
34
2.39 9.29
1.58 5.52
Workpiece with 0.25 mm depth of cut Test No. Pre-water 37 38 39 40 42 43 44 45 jet Ra (μm) 0.92 1.72 1.37 0.96 0.92 3.65 2.23 2.73 3.64 Rv (μm) 3.41 3.91 4.93 3.58 3.11 9.33 6.79 7.66 10.3
6 Conclusion and Outlook • Photographs and trace laser gauge of the surface confirmed successful deburring. • Water jet can be used successfully for deburring burrs that are produced by different machining conditions. • It is possible to remove burrs with water jet without too much damage to the substrate. • Parameters should be altered according to burr sizes. The minimization or eradication of burrs that is produced during the water jet process is therefore of the utmost importance. It is recommended that more work needs to be carried out in this area in order to find out more about the procedure under different conditions.
195 Acknowledgements The authors would like to thank Professor Philip Shipway and Barry Holdsworth of the Department of Mechanical, Materials and Manufacturing Engineering at the University of Nottingham and Dr Steven T. Halliday of Rolls-Royce plc for their advice and support.
References 1. Lee, K. U., Ko, S. L., 2008. Development of deburring tool for burrs at intersecting holes, Journal of Materials Processing Technology, 201: 454–59 2. Ko, S-L., Dornfeld, D. A., 1996. Analysis of fracture in burr formation at the exit stage of metal cutting, Journal of Materials Processing Technology, 58: 189–200 3. Bone G. M., Elbestawi M. A., Lingarkar R., Liu L., 1991. Force control for robotic deburring, ASME Journal of Dynamic System Measurements of Control, 113: 395–400 4. Chang, S. S. F., Bone, G. M., 2005. Burr size reduction in drilling by ultrasonic assistance, Robotics and Computer-Integrated Manufacturing, 21: 442–50 5. Balasubramaniam, R., Krishnan, J., Ramakrishnan, N., 1998. Investigation of AJM for deburring, Journal of Materials Processing Technology, 79: 52–58 6. Balasubramaniam, R., Krishnan, J., Ramakrishnan, N., 2000. An empirical study on the generation of an edge radius in abrasive jet external deburring (AJED), Journal of Materials Processing Technology, 99: 49–53 7. Balasubramaniam, R., Krishnan, J., Ramakrishnan, N., 1999. An experimental study on the abrasive jet deburring of crossdrilled holes, Journal of Materials Processing Technology, 91: 178–82 8. Duflou, J. R., Kruth, J.-P., Bohez, E. L., 2001. Contour cutting of pre-formed parts with abrasive waterjet using 3-axis nozzle control, Journal of Materials Processing Technology, 115: 38–43.
Ice Blasting – An Innovative Concept for the Problem-Oriented Deburring of Workpieces B. Karpuschewski and M. Petzel
Abstract The paper describes the development of an operational alternative for particle blasting with deep frozen ice, an in this form solid and hard blast medium. Primarily the practical realization of this processing step for the above described case is presented, especially in less accessible locations on complex components. The procedure investigated in this article is essentially a blasting procedure using a solid blast medium. The innovative idea lies in the use of ordinary ice as a blast medium. The advantage of ice is its property not to leave any solid residue behind. Consequently it is applicable to the blasting treatment of complex component geometries. The use of conventional blast media for purposes of blast deburring entails the subsequent removal of blast medium residue. In the case of complex component geometries, complete removal of said residue is not feasible. The objective is to create the potential to implement the above-described procedure of blasting by use of deep-frozen ice. This requires a new form of equipment which allows for specialized fabrication of suitable ice. The ice blasting procedure examined here must not be confused with dry ice blasting, which uses frozen carbon dioxide CO2 as a blast medium. Keywords Ice blasting · Deburring · Deep frozen water-ice
efforts are inevitable. As a result, even in times of continuous automation components must often be deburred manually. In order to counteract this trend and keep up with heightened requirements of production, implementation of a new and innovative deburring procedure, which will be presented here, needs to be promoted. The procedure is referred to as ice blasting or ice deburring. The procedure investigated here is essentially a blasting method using a solid blast medium. The innovative idea at the foundation of this endeavour lies in the use of ordinary ice as a blast medium. The advantage of ice is its property to not leave any solid residue behind and in that it is consequently applicable to the blasting treatment of complex component geometries, as shown in Fig. 1. In this current case the diameter of hole is 5 mm and the position of burr in hole is 30 mm distant from access. The use of conventional blast media for purposes of blast deburring entails the subsequent removal of blast medium residue after work. In the case of complex component geometries, complete removal of said residue is not feasible. Even if elaborate removal is carried out afterwards, additional costs are generated that make the applied procedure seem ineffective. By using ice as blast medium, elaborate removing procedures become obsolete which, especially in the case of complex component geometries, marks a vast and decisive advantage.
1 Introduction Deburring of complex components is subject to formidable efforts in terms of time and expenses. In order to ensure quality and due to lack of operational alternatives, these
B. Karpuschewski (), M. Petzel Institute of Manufacturing Technology and Quality Management, Otto-von-Guericke-University of Magdeburg, P.O. Box 4120, 39016 Magdeburg, Germany e-mail:
[email protected] url: www.ifq.ovgu.de
Fig. 1 Example-workpiece with intersection boreholes, valve plate
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_21, © Springer-Verlag Berlin Heidelberg 2010
197
198
Hardness of water-ice 6
Heat-treated steel
+
5 Mohs-hardness
The objective is to create the potential to implement the above-described procedure of blasting by use of deep-frozen ice. This requires a new form of equipment which allows the specialized fabrication of suitable ice. The ice blasting procedure examined here must not be confused with dry ice blasting, which uses frozen carbon dioxide as a blast medium. Dry ice or CO2 blasting does not remove material by mechanical means, but primarily through expansion of the dry ice pellets, which in the process of sublimation abruptly increase their volume. In this case, the resulting pressure wave rips off material already loosened by thermal shock [1].
B. Karpuschewski and M. Petzel
4
x x
3 2
x
Limestone x Sulfur Blackwelder, 1940 + Koch & Wegener, 1930 x
1 0 0
–10 –20 –30 –40 –50 –60 –70 –80 –90 –100 Temperature / °C
2 Theoretical Backgrounds
Fig. 2 Mohs hardness of water-ice [2]
2.1 Ice as an Abrasive Blast Medium
Another aspect should be noted at this point: As a result of its temperature-dependent hardness, ice is the only blastmedium that can be adapted to the conditions and requirements of the process, from hard to very soft.
The hardness of dry ice is about 2 Mohs, which roughly compares to the hardness of gypsum. This is not sufficient for deburring metallic materials. This is different with ice, frozen water (H2 O). At temperatures in the upper freezing range, ice possesses the same properties as dry ice, but these change as the influence of cooling energy increases. The advantages of ice deburring result from the utilization of physical properties of the employed blast medium, ice. Ice constitutes an easily available, comparatively easy to fabricate blast medium, which subsequent to transmitting kinetic energy has the ability to change its physical state in such a way that it will not leave any solid residue after use. Aside from removed material, only liquid residue like water or emulsion is left behind. A second, even more decisive property of ice lies in the temperature dependence of its hardness, as shown in Fig. 2. The hardness of ice is indirectly proportional to its temperature, which means that with decreasing temperature, hardness will increase. Thus, the Mohs hardness of ice is 2 at a temperature of –10 ◦ C, similar to gypsum. In contrast, its Mohs hardness is 6–7 at a temperature of –80 ◦ C, comparable to the hardness of glass or steel. This property of ice is at present neither considered nor utilized in ice blasting procedures, yet providing an innovative application in this field. “An ideal blast medium should have an edged form, has a hardness of at least 6 Mohs and desintegrates into gas at room temperature completely [1].” Even though ice does not completely fulfil all three of these requirements since it will melt to a liquid and not become gaseous, it is very close to the definition of a perfect blast medium according to [1]. In order to fulfil that last requirement, machined workpieces could be dehumidified by simple drying procedures.
2.2 Mechanism of Ice Blasting The primary procedural mechanism is based on the impact of ice particles on the workpiece surface. In the process of hitting the workpiece that needs deburring, kinetic energy is transformed into severing energy through deceleration. Figure 3 demonstrates the mechanism of the severing process. It indicates that by observing a certain blast angle, a chip is severed out of the surface. If there is a burr present, it will be removed preferentially, since it protrudes from the surface [3]. Pressure surges produced by the transmitted kinetic energy of the ice particles result in plastic deformations at the burr and thus it is removed from the surface of the workpiece. To what extent the influence of molten water, which comes
+γ
v
1 Blast particle (tool) 2 Workpiece v Particle speed +γ γ Chip angle
v
1 2
Fig. 3 Behaviour of a solid blast medium at the surface of a workpiece [3]
Ice Blasting – An Innovative Concept for the Problem-Oriented Deburring of Workpieces
as a side effect of the alternating influence of impact energy on the ice particles, adds to the build-up of pressure surges in the liquid, is subject to speculation. This phenomenon can be observed in the procedures of water jet cutting as well as in high pressure water jet deburring. When inner tensions rise up to the point where they exceed the material strength, cracks begin to appear. Spreading and aggregation of these cracks will then lead to the removal of surface material [4]. At the same time multiple pressure waves from the impact of drops accumulate to pressures so high that material gets broken out of the substance matrix. Another secondary mechanism is being caused by the influence of low temperatures on the brittle fracture characteristics of a wide variety of materials, e.g. steels, non-ferrous metals, or plastics. Only non-austenitic steels are the exception. In all other materials, embrittlement caused by cold will lead to a weakening of the atomic grid structure, also known as substance or particle matrix, and as a result thereof to a decrease in material strength. This embrittlementis brought about by the utilization of deep-frozen ice as the blast medium. Extremely low temperatures in the range of up to –100 ◦ C deprive the environment of heat and cooling it down in the process. Naturally, the workpiece is being predominantly affected by this. Depending on the heat conductance capacity of the blasted material the temperature in the edge layers will drop until it reaches
199
the temperature of the shot. Additionally, thermal shock to the material is influenced by rate of feed, nozzle offset and compressor pressure. In table 1 the advantages and disadvantages of the new deburring operation are shown, with short comments to other deburring processes.
3 Process Descriptions 3.1 The Production Process of Blast Ice As it is shown in Fig. 4, the ice is produced in a cold-resistant pressure tank with a heatable cone nozzle which is on the top of the equipment and a hopper mounted on the bottom of the device. A layer of low temperature insulation is attached on the outside. The mode of operation is as follows: Inside the tank, which is cooled by liquid nitrogen LN2 running through a circular pipe in the upper region of the cylinder, water is being dispersed through a water spray nozzle in the lid. It then freezes in the cold atmosphere of about –120 ◦ C. The frozen ice particles with diameters ranging from 0.2 to 0.8 mm gather in a hopper in the lower region of the device, which feeds them to an outlet. The angle of the hopper should be dimensioned so as to ensure easy trickling of the ice particles. After the particles have gathered in the Water H2O
Table 1 Comparison of advantages and disadvantages of the ice blasting deburring method, with comments to other deburring operations Advantages Comments Water ice with low temperature has Dry ice (frozen CO2 ) has not the nearly the hardness of stainless required hardness to remove steel or glass, it works like an hard material from the surface hard abrasive medium No laborious post-cleaning Expensive cleaning operations are required required by using ECM, TEM, AFM, ..., and cutting processes High flexibility In case of ECM for each part a new device will be needed No limited depth for deburring of By using brushes or other intersection boreholes deburring tools the working depth is limited Cold brittleness of the burr All other shape-cutting operations decrease the required cutting have a defined, constant cutting force force Good automatable –
10°C p0 = 1 p1 > p0
Nitrogen LN2 –190 °C
Nitrogen LN2 –190°C
–120 °C
Disadvantages
Comments
No defined edge radius producible
With cutting tools a defined edge –100 °C radius is producible As competitive process to hand Ice particle deburring or water jet deburring – Fig. 4 Production process of blast ice
Special deburring technology Corrosion without drying
200
B. Karpuschewski and M. Petzel Water
Nitrogen p0 = 1
p1 > p0
Cryo-tank
Workpiece
first place, they had to be prepared beforehand using wirecut EDM. This was necessary in order to make the areas exhibiting the burrs accessible for blast processing. Additionally, optical analysis of the burr, prior to as well as after processing, would have been impossible without this procedure. Prevalent parameters of the experiment and its environment were:
Blast medium (Ice)
Compressed air Nozzle Blast medium (Ice)
Fig. 5 Procedural description of ice blasting process
lower part of the device, they remain there for a short while and, in doing so, are cooled down further until they reach the blast temperature.
3.2 The Ice Blasting Process When the particles have reached the desired final temperature of –120 ◦ C maximum, the inherent overpressure will help conveying them through a tube to the blast unit, an injector nozzle. This process is shown schematically in Fig. 5.
4 Practical Investigations of the Removal Capacity of Deep Frozen Ice Until now, no equipment for the fabrication of ice has been available to allow a verification of the active principles of ice deburring. Nevertheless, initial sample experiments should be conducted in order to verify these principles and to determine input parameters for the construction of said equipment. It thus appears legitimate to use crash ice as raw material for the fabrication of suitable ice particles. The ice was stirred in a container specially prepared for this purpose, a calorimeter filled with liquid nitrogen, first in a crudely chopped form, then reduced to smaller pieces and sifted. The in this way fabricated ice proved to be fine-grained and trickled easily. The dimensions were in a range of 0.5–2 mm diameter and the structure was the same like broken glass. Its motion properties were the same as those of dry sand. On the basis of the conducted investigation an assessment of the removal capacity of deep-frozen ice particles used as blast medium has been made. For this purpose, referee components featuring burrs were subjected to conditioning with this blast medium. To be able to process these parts in the
• • • •
temperature of blast ice: –120 ◦ C ambient temperature: 21 ◦ C working pressure/blast pressure (constant): 9 bars negative pressure conveyance through injector nozzle
Through realization of practical experiments, the afore stated removal capacity could also be proven practically. Definite removal capacity regarding the referee burrs on processed components has been demonstrated. This becomes obvious when comparing the original and processed state, as is shown in Figs. 6 and 7. The hardware utilized in the practical investigation only allowed an operation at a low performance. In terms of particle velocity, it should be possible to increase performance up to 300%. Furthermore, the above-described fabrication process of the used ice could have extremely negative effects on material structure and homogeneity of the blast medium. The actual hardness characteristics of ice at temperatures of –120 ◦ C also remain unknown. Because of this, there are also resources found in these respects in order to increase removal performance. Exact investigation as to the improvement of material properties resulting from altered production and cooling methods can not take place until realization of a suitable
1000 μm
Fig. 6 Bore hole before conditioning with deep frozen blast ice
Ice Blasting – An Innovative Concept for the Problem-Oriented Deburring of Workpieces
Fig. 7 Bore hole after blast deburring processing with deep frozen blast ice
fabrication device for the production of deep-frozen blast ice. The equipment is currently under construction. Building of this experimental installation will presumably be completed by the beginning of 2009.
201
also holds true for the reliable detection of burrs on said components. All of these developments are made in an attempt to optimize and advance methods for the removal of burrs and prevention of their formation. A novel procedural variant for blast deburring processing of workpieces has been presented. It includes the use of deepfrozen ice as blast medium for the removal of burrs on complex component geometries. In this context, temperaturedependent hardness and removal capacity of ice have been confirmed. A possibility for the fabrication of ice according to the needs of production has been pointed out and described. The utilized procedural principle is based on quick-freezing of dispersed water. Defined water mist is created in a very cold atmosphere, in which it freezes to ice particles. The cold atmosphere is being artificially generated in a steel cylinder. This is done by using liquid nitrogen. Ice is a blast medium with a linear increase in hardness as temperature decreases continually below –0 ◦ C. Depending on the chosen temperature of the ice particles, these particles can be used with the same blast equipment for a variety of processing modes and different materials. In this way, soft metals and plastics may be processed just like steels and other hard materials.
References 5 Summary and Outlook Removal of adherent burrs on complex components and workpieces constitutes a problem in production engineering. Especially the automated implementation into the process of production is difficult to achieve. Reasons for this are the complexity of components, the insufficiently foreseeable shape in which burrs manifest and the imprecise means to predict their areas of origin on the component. Huge research efforts are undertaken to successfully forecast the formation of burrs using computer-based methods of simulation. This
1. Haberland, J.: Reinigen und Entschichten mit Trockeneisstrahlen – Grundlegende Untersuchungen des CO2-Strahlwerkzeuges und der Verfahrensweise; Dissertation Universität Bremen; FortschrittBerichte VDI, Reihe 2, Nr. 502; Düsseldorf, VDI Verlag, 1999 2. Juhnke, M.; Weichert, R.: Erzeugung von Nanopartikeln durch Feinstzerkleinerung bei hohen Reinheitsanforderungen; Vortrag zur GVC/DECHEMA-Jahrestagung, Wiesbaden 06.-08.09.2005; Chemie Ingenieur Technik, 2005 3. Beier, H.-M.: Handbuch Entgrattechnik – Wegweiser zur Gratminimierung und Gratbeseitigung für Konstruktion und Fertigung; München, Hanser Verlag, 1999 4. Rieger, H.: Über die Zerstörung von Metallen beim Aufprall schneller Wassertropfen, Zeitschrift Metallkunde, 57, S.693–699, 1966
Deburring Processes – Applications
Study of Internal Deburring of Capillary Tubes with Multiple Laser-machined Slits H. Yamaguchi and J. Kang
Abstract Laser micromachining processes cut material by melting the workpiece with thermal energy. The solidified material adheres to the cut edges and forms burrs, which cause many serious problems in manufacturing. Using existing conventional techniques, it becomes difficult to remove the heat-affected hard burrs that project inside capillary tubes. The aim of this research is to develop a deburring process for burrs that project into areas hard to reach by conventional techniques. This research applies the magnetic abrasive finishing process to capillary tube deburring. Through the manipulation of the magnetic field at the finishing area, the process controls the motion of the magnetic abrasive introduced into the tubes. This study includes a discussion of the principle for the internal deburring process of tubes with multiple slits and the refinement of the finishing machine. Experiments demonstrate the deburring characteristics of multiple laser-machining burrs projected inside 0.0165 in (419 μm) inner diameter tubing. Keywords Laser-machined burr · Burr removal · Magnetic abrasive finishing · Capillary tube
1 Introduction Laser micromachining processes are widely used for cutting, drilling, marking, texturing, and structuring for medical, electrical, optical, automobile, and aircraft applications [1]. The processes cut material by melting the workpiece with thermal energy and are capable of making small holes or slits (e.g., features found in coronary stents) from outside capillary tubes [2]. For example, the multiple slits in the tube wall increase the flexibility of the tube, as is desired
H. Yamaguchi (), J. Kang Department of Mechanical and Aerospace Engineering, University of Florida, P.O. Box 32611, Gainesville, FL 32611-6300, USA e-mail:
[email protected] url: www.mae.ufl.edu/hitomi/
in components such as stents. The applications, however, are not limited to short components such as stents, and the length of such flexible capillary tubes can exceed 1 m. However, the solidified material can adhere to the cut edges and form burrs. Moreover, some molten material splashes inside the tube, solidifies, and adheres to the inner surface. This causes serious problems in manufacturing. Conventional deburring tools cannot be easily introduced into flexible capillary tubes that have inner diameters smaller than 0.5 mm and incorporate multiple slits with, and their motion cannot be adequately controlled. Using existing conventional techniques, including chemical processes, it becomes difficult to remove the burrs that project inside the flexible capillary tubes. Moreover, disposal of the chemical reagents often contributes to the destruction of the environment. Recent developments in nonconventional machining process technology have enabled automated micro deburring and surface finishing for precision parts. They include electrochemical machining [3], ultrasonic vibratory finishing [4], ultrasonic cavitation deburring [5], thermal energy deburring [6, 7], abrasive flow machining [8], laser deburring [9], magnetic barreling [10], and magnetic abrasive finishing (MAF) [11, 12]. For finishing slender passages, abrasive flow machining and MAF are typically the only potentially viable processes. Abrasive flow machining performs deburring of the small holes of diesel injection nozzles made by drilling, electrodischarge machining, or laser machining. The pressurized media (a mixture of a polymer carrier, lubricant, and abrasive), pass through the holes and finish the inner surfaces, as well as remove the burrs. As long as the abrasive behavior is controlled by media pressure at the end of the hole, the pressure drop along the length of the hole will impede the expansion of the process application. This results in the difficulties in applying abrasive flow machining to internal deburring of slender capillary tubes. In internal finishing by MAF, the magnetic abrasive consisting of iron and aluminum oxide grains is introduced into a tube, and, in the presence of a magnetic field, it is pushed against the inner surface of the tube by magnetic force. The
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_22, © Springer-Verlag Berlin Heidelberg 2010
205
206
relative motion of the magnetic abrasive against the inner surface of the tube results in the surface finishing. Manipulating the magnetic field along the tube axis drags the magnetic abrasive and achieves the internal finishing of long tubes. The process has been successfully applied to the internal finishing of 400 μm inner diameter tubing [13]. The aim of this research is to develop a deburring process for burrs that project into areas hard to reach by conventional techniques, and this research applies MAF to internal deburring of flexible capillary tubes incorporating multiple slits. This study includes a discussion of the process principle for the internal deburring of tubes with multiple slits and the refinement of the finishing machine. Experiments demonstrate the deburring characteristics of multiple lasermachined spiral burrs inside 0.0165 in (419 μm) inner diameter (ID) tubing and reveals the conditions required for successful capillary tube deburring.
2 Processing Principle Internal finishing by MAF utilizes one of two configurations. One is the rotating-tube-stationary-pole system [13], and the other is the rotating-pole-stationary-tube system [14]. These systems are chosen to suit the workpiece geometry. The former system is for short workpieces that are rotatable at high speeds. The latter system was developed for nonrotatable workpieces, which have long, large-sized or non-rotationsymmetrical geometry, such as elbows, bent tubes, and slender tubes. Since the workpieces provided for this study were relatively short (55 mm long) and rotatable at high speeds, the application of the rotating-tube-stationary-pole system was considered. Figure 1 shows a schematic of the internal finishing process using a rotating-tube-stationary-pole system for stain-
H. Yamaguchi and J. Kang
less steel capillary tubes. Permanent magnets generate the magnetic field needed for attracting the magnetic abrasive to the finishing area, pressing it against the inner surface of the workpiece. In a nonuniform magnetic field, the magnetic force Facts on the magnetic abrasive and is calculated using the following equation [13]: F = Vχ H · gradH
(1)
where V is the volume of the magnetic abrasive, χ is the susceptibility, and H and grad H are the intensity and gradient of the magnetic field, respectively. If the tangential component of the magnetic force acting on the magnetic abrasive is larger than the friction force between the magnetic abrasive and the inner surface of the workpiece, the magnetic abrasive shows smooth relative motion against the inner workpiece surface when the workpiece is rotated at high speed. Manipulating the poles along the workpiece axis causes the magnetic abrasive to move in the axial direction following the pole motion, effectively finishing the inner surface and removing the burrs. A previous study reported that MAF was applied to remove burrs created by drilling 2 mm diameter holes through 5056 aluminum alloy tube (Ø38 × Ø35 × 150 mm) [15]. During processing, the drilled holes were covered by tape (0.2 mm thickness) to prevent the magnetic abrasive and lubricant from leaking out. The tube was treated practically as a rigid standard tube for processing. If the tape thickness is smaller than the clearance between the tube and pole, which is determined by magnetic field analysis, this method is simple and effective. For the capillary tube deburring, the pole-tube clearance should be set at around 0.1 mm. In the present research, sealing the multiple spiral slits of the tube by taping while maintaining an approximate 0.1 mm clearance between the tube and pole tip is not practical. In this study, therefore, it was proposed to rotate the tube in the direction in which the motion itself acts to narrow the spiral slit width. For this reason, the rotating-tube-stationary-pole system was chosen for this study.
3 Experimental Setup and Conditions
Fig. 1 Schematic of processing principle using a rotating-tubestationary-pole system
The previous research reported that the key to achieve finishing of capillary tubes was the combination of the methods that (a) control the magnetic field at the finishing area, (b) determine the appropriate amount of mixed-type magnetic abrasive, and (c) support the capillary at three points [13]. The finishing setup and conditions were refined to satisfy the abovementioned requirements. The effects on the experimental setup and conditions of the geometrical differences
Study of Internal Deburring of Capillary Tubes with Multiple Laser-machined Slits
207 Table 1 Experimental condition
Fig. 2 Photograph of finishing equipment
between the capillary tubes with and without spiral slits will be discussed below.
3.1 Experimental Setup Figure 2 shows an external view of the experimental setup developed in the laboratory. The stainless steel capillary tube with spiral slits is held by the chuck. While the capillary tube is rotated at high speeds, centrifugal force acts on the tube. Depending on the direction of the tube rotation, the tube either elongates or shortens by changing the slit widths. To diminish the run-out of the capillary tube beyond the finishing area, the other end of tube is held by a flexible jig, which can react to the changes in the tube length. Neodymium-iron-boron permanent magnets (10 × 12 × 18 mm, residual flux density 1284 mT, coercive force 1440 kA/m) generate the field needed for attracting the magnetic abrasive to the finishing area. The magnets can be oscillated in the axial direction (by a crank mechanism connected to the motor) and also fed in the axial direction of the tube with or without oscillation. The pole tip geometry shown in Table 1 was determined based on the previous study of capillary tube finishing, and the pole tip surface was covered by 0.0046 in (0.1143 mm) thick polytetrafluoroethylene tape. The tube was in contact with the pole tip surface during finishing. This minimized the clearance between the pole tip and finishing area to maintain a strong magnetic field at the finishing area and contributed to diminishing the run-out of the tube during finishing. The tape protected the tube from rubbing against the pole tip.
3.2 Finishing Conditions Table 1 shows the experimental conditions. SUS304 stainless steel tubes (ID 0.0165 in (419 μm), OD 0.023 in (584 μm),
and length 55 mm) with spiral slits were provided as workpieces for this study. The undeformed slit width was measured at about 12 μm. The tube was chucked 12.5 mm from one end. Since laser-machined burrs have various shapes and heat affected zones, sharp and hard cutting edges combined with strong magnetic force are required to remove the burrs. Conventionally used for 304 stainless steel tube finishing, mixed-type magnetic abrasive – a mixture of relatively large-sized iron powders and magnetic abrasive (mean diameter 80 μm and comprising iron particles and Al2 O3 abrasive grains under 10 μm in diameter) – could not remove the heat-affected hard burrs. Hereafter, this magnetic abrasive will be referred to as WA magnetic abrasive. Diamond abrasive in a paste form was, thus, added to the mixed-type magnetic abrasive with lubricant. The WA magnetic abrasive has irregular microasperities on the surface produced during the manufacturing process. These microasperities allow the WA magnetic abrasive to hold the nonferrous diamond abrasives through the lubricant at the finishing area. One of the important properties of the lubricant must be the viscosity, which must be high enough to prevent leaking through the slits and to encourage the lubrication between the abrasive cutting edges and target surface but low enough to be introduced into the capillary tube.
208
The necessary amount of the magnetic abrasive can be determined according to the ratio of the volume of the magnetic abrasive to the overall volume inside the tube (in a length corresponding to the width of the pole tip). There exists a threshold of the supplied abrasive amount between 43 and 55 vol.%; a supplied amount less than the threshold is required for efficient finishing [13]. The mixture of 0.1 mg WA magnetic abrasive and 0.4 mg iron particles – which results in 47 vol.% taken up by the iron powder and WA magnetic abrasive – was supplied for the experiments. According to Preston’s equation [16], the material removal rate is a function of the finishing pressure, relative speed between the tube and abrasive, and the Preston coefficient, which is determined by the finishing system (e.g., by the abrasive type and size, and friction coefficient between the abrasive and tube). The magnetic force described in Eq. (1) is the main parameter determining the finishing pressure and is determined by the magnetic field. The relative speed between the tube and abrasive is controlled by the combination of tube rotational speed and pole reciprocation in the direction of the tube axis. The higher the relative speed, the longer the theoretical sliding distance of the abrasive against the tube surface. However, the cutting tools (diamond abrasive with lubricant) are not ferrous, and the centrifugal force caused by the tube rotation at high speeds must spin them out of the finishing area. The tube rotational speed was set at 2500 min–1 . Under the conditions with high speed tube rotation, the mixture of the magnetic abrasive and iron particles shows difficulties to follow the poles feed in the tube axis direction because the friction force becomes much larger than the magnetic force. The burrs also obstruct the mixture to follow the poles feed. Therefore, the feed speed was set at the lowest of the experimental setup, 0.67 mm/s. The crossing angle θ of the cutting marks generated by the diamond abrasive is calculated as 2 × tan–1 (vf /vr ), where vf is the pole feed velocity, and vr is the tube rotational speed. Under the experimental conditions, θ was calculated to be 1.4◦ . The effects of the diamond abrasive size on the processing characteristics were examined in this study. Three sizes of diamond abrasive were prepared: 4–8, 20–40, and 50–70 μm, which are smaller, slightly larger, and much larger than the slit width, respectively. The finishing unit was reciprocated over 4 mm at 0.67 mm/s, finishing an 8-mm length located 3 mm from the end opposite the chuck. The stylus of the surface roughness profilometer, which is the device most commonly used for surface roughness measurement, could not be inserted inside the 419 μm ID capillary because of the small tube diameter. This created difficulty in measuring and tracking the changes in the surface roughness and burr heights with finishing time. The tube was cleaned using ethanol in an ultrasonic cleaner every 6 min, and the changes in the material removal with finishing time
H. Yamaguchi and J. Kang
were tracked only by measuring the weight reduction by the process with a micro-balance (10 μg resolution). Following replenishment of the mixture of the ferrous powders and abrasive every 6 min, the finishing experiments were continued. After the finishing for a certain period, the tube was sectioned along the tube axis, and the surface was evaluated using an optical profilometer and microscope.
4 Finishing Characteristics The time sufficient for finishing was initially determined by examining the changes in the finishing characteristics with finishing time using the 50–70 μm diameter abrasive. The diamond abrasive is larger than the undeformed slit width, measured to be 12 μm. Figure 3 shows the micrographs of the unfinished surface and the surface finished for 30 and 60 min. Although areas damaged by the laser machining process are observed between the slits, the initial surface burrs adjacent to the slits are the target of this study. The initial heights of the burrs generated by laser machining were measured by an optical profiler to be in the range between 8 and 50 μm.
Fig. 3 Micrographs of unfinished surface and surfaces finished for 60 and 30 min with 50–70 μm diamond abrasive
Study of Internal Deburring of Capillary Tubes with Multiple Laser-machined Slits
During the process, the tube was deformed such that the slits were narrowed due to the tube rotation. When the diamond abrasive encounters the slit during the process, the abrasive momentarily loses contact with the surface, but contacts the surface again on the other side. According to the cutting marks, the reengagement of the abrasive must be smooth. The diamond abrasive must gradually remove material from the peaks of the burrs and the microasperities of the tube surface. The conditions for 30 min were found to be insufficient for effective deburring. By increasing the finishing time, however, the edge and surface finishing were both smoothly performed. Figure 4 shows the changes in the material removal with finishing time. The variation of the material removal at the beginning was due to the variation in the initial burr conditions. This indicated that about 0.3 mg of material removal might be necessary for both edge and surface finishing of the tubes. Accordingly, it was shown that MAF is applicable for inner surface finishing and removal of laser-machined spiral burrs projected inside flexible capillary tubes. Next, the effects of the diamond abrasive size on the deburring characteristics were examined with three sizes of diamond abrasives: 4–8, 20–40, and 50–70 μm diameter. Figure 5 shows micrographs of the unfinished surface and the surfaces finished with 4–8 and 20–40 μm diamond abrasive for 60 min. In the case with 20–40 μm diamond abrasive, shown in Fig. 5(a), the burrs and the initial surface unevenness remain due to a lack of material removal. This is because of the smaller cutting edges of the diamond compared to the 50–70 μm diamond abrasive. In the case of the 4–8 μm diamond abrasive, shown in Fig. 5(b), no burrs remain, but deep, irregular undulations are observed on the finished surface. During the process, the magnetic force pushes the iron particles, which push the WA magnetic abrasive, which, in turn, push the diamond abrasive against the inner surface of the tube. Some 4–8 μm diamond abrasive must agglomerate between the WA magnetic abrasives and tube surface. The mass of the diamond abrasive pushed by the WA magnetic
Fig. 4 Changes in material removal with finishing time
209
Fig. 5 Micrographs of surfaces finished with 4–8 and 20–40 μm diamond abrasive
abrasive and ferrous particles must participate in the finishing performance, resulting in the deep scratches. This also led to the removal of the burrs and surface finishing with irregular undulations. Figures 6 and 7 show the three-dimensional shapes (measured by an optical profiler) of the unfinished inner surface and the inner surfaces finished with three sizes of diamond abrasives for 60 min. In Fig. 6, the surface data in the area of a slit is passed through a low-pass filter so that the burr shapes can be examined. The irregularity of the surface brought about measurement difficulties, resulting in the discontinuity of the surface seen in Fig. 6. However, except for the slit, the surface, is, in fact, continuous. The abovementioned burr heights were obtained as peak-to-valley values of lines drawn perpendicular to the slit from several observations. To examine the surface roughness, Fig. 7 shows the unfiltered area between slits. As shown in Figs. 6(a) and 7(a), the case with 50–70 μm diamond abrasive demonstrated the burr removal and surface finishing. The surface was modified from 0.67 μm to 0.12 μm Ra. In the case of the 20–40 μm diamond abrasive, the initial burrs remained after 60 min, as shown in Fig. 6(b). This condition must have merely removed material from the peaks of the surface irregularities and left the 5–7 μm high burrs. Although the three-dimensional surface shape observation barely shows the initial micro-unevenness remaining, the small material removal resulted in slowed surface roughness improvement, from 0.65 μm to 0.21 μm Ra (see Fig. 7(b)). However, the 4–8 μm diamond abrasive removed burrs successfully, as shown in Fig. 6(c). The surface finished by the 4–8 μm diamond abrasive consists of the accumulation of shallow cutting marks on the deep undulations with relatively longer wavelength, which must be caused by the
210
H. Yamaguchi and J. Kang
Fig. 6 Three-dimensional burr shapes measured by optical profiler
agglomerated diamond abrasive. This generated the surface roughness value to 0.35 μm Ra (Fig. 7(c)). The cases with 4–8 and 20–40 μm diamond abrasives show another trend, as well. After finishing, a mixture of lubricant, diamond abrasive, and chips was observed outside the tube, and cutting marks were observed on the outer surface of the tube, as shown in Fig. 8. The heat generated during the process must decrease the viscosity of the lubricant, and the 4–8 μm diamond abrasives, which are smaller than the slits, must have leaked through the slits. The diamond abrasive became sandwiched between the rotating tube and pole tip, and achieved external surface finishing of the tube. The unagglomerated 4–8 μm diamond abrasive must pass easily through the slits, and a large number of abrasive must participate in finishing of the outer surface. On the other hand, the 20–40 μm diamond abrasive could have been larger than the slit widths. In practice, a distribution including smaller-sized diamond abrasive results from the manufacturing process. Alterna-
tively, the diamond abrasive could be crushed during the process, and some abrasive would thus be reduced in size. The diamond abrasive that was able to migrate from the internal finishing area must have caused the scratches on the outer surface and slightly removed the material from outer surface. Figure 9 shows changes in the material removal with finishing time. The material removal in the 20–40 μm diamond abrasive case is lowest of the three conditions. It is noted that the material removal in the 4–8 and 20–40 μm diamond abrasive cases are the results of finishing of both the inner and outer surfaces. In the case with 20–40 μm diamond abrasive, the lack of large cutting edges of the abrasive resulted in the smallest material removal. Regardless of the fact of being the smallest abrasive of the three, the effects of the agglomeration of the 4–8 μm diamond abrasive in the internal finishing and the leakage of the diamond abrasive from the slit for the outer finishing led to the greatest material removal.
Study of Internal Deburring of Capillary Tubes with Multiple Laser-machined Slits
Fig. 7 Three-dimensional surface shapes measured by optical profiler
Fig. 8 Micrographs of outer surface tube
Fig. 9 Changes in material removal with finishing time
211
212
5 Conclusions This research examines the application of MAF to the deburring of capillary tubes with laser-machined burrs projected inside them, and it describes the mechanism and conveys the required conditions to achieve internal deburring based on the finishing experiments. The results of this research can be summarized as follows: 1. The finishing experiments showed the feasibility of MAF for both surface and edge finishing of the capillary tubes with multiple laser-machined spiral slits. 2. Hard, sharp cutting edges, e.g., those of diamond abrasive, are necessary for removing the laser-machined burrs. Maintaining the presence of nonferrous diamond abrasive at the finishing is a key to achieve deburring the inside of flexible capillary tubes with spiral slits. • The rotating-tube-stationary-pole system should be applied to the deburring process, and the direction of the tube rotation must be chosen such that the widths of the slits become narrower. • The diamond abrasive should be larger than the slit width to avoid losing the abrasive cutting edges from the finishing area. • The viscosity of the lubricant should be high enough to avoid leaking through the slits and to encourage the lubrication between the abrasive cutting edges and target surface but low enough to be introduced into the capillary tube. • The ferrous particles must have microasperities on their surfaces to keep the nonferrous abrasive cutting edges at the finishing area. In case of internal deburring of flexible capillary tubes with slits, increasing the tube rotational speed encourages the opening of the slits, which leads and to splashing of the lubricant from inside to outside the tube through the slits, reducing the amount of lubricant in the processing area. Moreover, the burrs are obstacles for the ferrous particles mixed with diamond abrasive when the particles follow the pole movement in the axial direction; this limited the increase of the axial feedrate. As a result, the processing speed can be rather slow. To improve the processing rate per tube, it is desired for the proposed process to have multiple processing areas for a single tube and to process multiple tubes simultaneously. The discussion of the optimization of the process efficiency, including the refinement of the processing machine, will be discussed in future work. Acknowledgments The authors would like to express their thanks to Creganna for their support in providing workpieces and to Gregory
H. Yamaguchi and J. Kang W. Sawyer of the University of Florida for his support in providing surface evaluation equipment.
References 1. Dahotre, N.B. and Harimkar, S.P., 2008, Laser Fabrication and Machining of Materials, Springer, NY. 2. Meijer, J., et al., 2002, Laser Machining by Short and Ultrashort Pulses, State of the Art and New Opportunities in the Age of the Photons, CIRP Annals – Manufacturing Technology, Vol. 51, No. 2, 531–550. 3. De Silva, A.K.M., Altena, H.S.J., and McGeough, J.A., 2003, Influence of Electrolyte Concentration on Copying Accuracy of Precision-ECM, CIRP Annals – Manufacturing Technology, Vol. 52, No. 1, 65–168. 4. Choi, H.Z. et al., 2004, Micro Deburring Technology Using Ultrasonic Vibration with Abrasive, Proceedings of the 7th International Conference on Deburring and Surface Finishing, 231–238. 5. Choi, H.Z. et al., 2003, Micro Deburring Technology Using Ultrasonic Cavitation, Proceedings of International Conference on Leading Edge Manufacturing In 21st Century: LEM21, Vol. 2003, 1013–1018. 6. Thermal Deburring Apparatus and Method, US Patent 4561839. 7. Brockbank, T.K., 2004, Thermal Energy Deburring, Proceedings of the 7th International Conference on Deburring and Surface Finishing, 225–230. 8. DeLo, D.P., Greenslet, J.M., and Munko, G., 2007, Improved Microhole Precision and Performance by MicroFlow Abrasive Flow Machining, Proceedings of the 15th International Symposium on Electromachining, 399–403. 9. Lee, S.H. and Dornfeld, D.A., 2001, Precision Laser Deburring, ASME Journal of Manufacturing Science and Engineering, Vol. 123, 601–608. 10. Sugiura, O., Imahashi, N., and Mizuguchi, M., 1997, Development of a Cylindrical Type Magnetic Barrel Finishing Machine, Journal of the Japan Society for Precision Engineering, Vol. 63, No. 3, 399–403. 11. Ko, S.L., Baron, Y.M., and Park, J.I., 2007, Micro Deburring for Precision Parts using Magnetic Abrasive Finishing Method, Journal of Materials Processing Technology, Vol. 187–188, 19–25. 12. Yin, S. and Shinmura, T., 2004, Vertical Vibration-Assisted Magnetic Abrasive Finishing and Deburring for Magnesium Alloy, International Journal of Machine Tools & Manufacture, Vol. 44, No. 12–13, 1297–1303. 13. Yamaguchi, H., Shinmura, T., and Ikeda, R. 2006, Study of Internal Finishing of Austenitic Stainless Steel Capillary Tubes by Magnetic Abrasive Finishing, ASME Journal of Manufacturing Science and Engineering, Vol. 129, No. 5, 885–893. 14. Yamaguchi, H., Shinmura, T., and Kaneko, T., 1996, Development of a New Internal Finishing Process Applying Magnetic Abrasive Finishing by Use of Pole Rotation System, International Journal of the Japan. Society for Precision Engineering, Vol. 30, No. 4, 317–322. 15. Shinmura, T., Hamano, Y., and Yamaguchi, H., 1998, A New Precision Deburring Process for Inside Tubes by the Application of Magnetic Abrasive Machining, 1st Report, On the Process Principle and a Few Deburring Characteristics (in Japanese), Transaction of the Japan Society of Mechanical Engineers., Vol. 64, No. 620, Series C, 1428–1434. 16 Preston, F.W., 1927, The Theory and Design of Plate Glass Polishing Machines, Journal of the Society of Glass Technology, Vol. 11, 214–256.
Robotic Deburring Based on On-line Burr Measurement L. Liao, F. Xi, and S. Engin
Abstract A new approach is presented in this paper for modeling and control of an automated deburring process that utilizes a hybrid robot and a compliant toolhead. The robot is used for motion control while the toolhead is used for force control. This toolhead has a pneumatic spindle that can be extended and retracted by three pneumatic actuators to provide the tool compliance. The model for deburring process is developed based on the dynamics of the compliant toolhead and the micro-contact model. By integrating a linear encoder, this toolhead can be used for on-line burr measurement through sensing the tool length variation due to burrs. For the deburring control, a closed-loop tool length controller is developed to regulate the tool length through the tool length sensing. This control method has been modeled, simulated, and implemented. Both simulation and experiment results demonstrate the effectiveness of the presented method. Keywords Robotics deburring · On-line burr measurement · Compliant deburring toolhead
1 Introduction Burrs are formed due to the plastic deformation/breakage of the material near the edges during the removal process. Parts must be cleaned before assembly free from burrs in order to ensure safe handling, facilitate good inspection and allow for a proper fit. All metal parts that are cast, formed, or machined require finishing of the edges to remove burrs and break sharp corners. The proper way of burr removal and its cost
L. Liao, F. Xi () Department of Aerospace Engineering, Ryerson University, 350 Victoria Street, Toronto, M5B 2K3, Canada e-mail:
[email protected] url: www.ryerson.ca S. Engin Pratt & Whitney Canada Corp., 1000 Marie-Victorin, Longueuil, Quebec, J4G 1A1, Canada
depend on the part geometry and the burr size. To reduce the deburring cost, the most suitable deburring method should be used and the burr size must be decreased beforehand. This can be achieved if the burr formation mechanism can be described analytically, which makes it possible to control the sizes of burrs and to minimize their appearances through an optimum choice of cutting conditions, tools, and workpiece geometry. Recently, more research and interest has been focused on the problems associated with burrs from machining. Much success has been gained in the study and modeling of burr formation in orthogonal and oblique cutting [1–6]. Deburring is a finishing process that uses bound abrasives or deburring tools to remove the sharp edges and burrs on a workpiece. Traditionally, deburring has largely been a manual operation that is very labor intensive, highly skill dependent, inefficient with long process time, high cost, error prone, and hazardous due to abrasive dust. Automation is a solution to overcome these problems. However, the successful implementation of an automated deburring system requires in-depth studies on the deburring process. In the past, limited research has been carried out to investigate various methods for automated deburring systems [7–12]. The conventional automated deburring systems are either based on robots or computer-numerical control (CNC) machine tools. The existing deburring control methods can be broadly categorized as rigid tools and compliance tools. The rigid tool method uses the actuators of the motion systems to generate the tool force, leading to a coupling between the tool force and the machine motion. This coupling causes the problems including time delays, intolerance for tool/part misalignment, and prone to tool breakage. The compliance tool method overcomes these problems by using a separate force device to generate the tool force, independent of the motion actuators. In practice, tool compliance can be implemented either passively or actively. A passive compliant tool is made by means of a passive device, such as spring. It relies on compliance in the tool itself to maintain a nominal contact force [13]. The contact force applied on the part during the deburring process is actually the compliance force from the tool. The advantage of the
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_23, © Springer-Verlag Berlin Heidelberg 2010
213
214
passive compliant force control is that passive compliance can be set at high stiffness to achieve a fast response rate, and the device is relatively cheap. The disadvantage is that the passive compliance cannot be actively regulated especially when the part geometry varies significantly. An active compliant tool uses built-in actuators to adjust the tool compliance actively. There are several compliant tools on the market. ATI Industrial Automation [14] has developed two basic types of compliant tools, namely, radialcompliant (RC) tools, and axial-compliant (AC) tools. Both active RC and AC tool are realized using pneumatics actuator to actively adjust the tool compliance in respective directions. The other two companies, PushCorp Inc. [15] and Robotic Accessories Leader [16], also provide the compliance tools, similar to ATI’s AC tool. The current automated deburring control method is to maintain a constant deburring force without on-line burr measurement. This method is not practically effective. For example, when the burr surface area is large, the contact stress between the part and the tool is low and the contact force should be increased in order to remove the burr effectively. Otherwise, it would under-cut. When the burr surface area is small, the contact stress is large and the contact force should be decreased to avoid over-cut. In this paper a new method is presented for the automated deburring control based on on-line burr height measurement. This method has been implemented on a system that utilizes a hybrid robot and a compliant toolhead. In what follows, the robotic deburring system is first described, and then the control method is presented along with modeling, simulation, and experiments.
2 Robotic Deburring System Figure 1 shows a complete robotic deburring system that has been developed at Ryerson University. This system consists of a hybrid robot and a compliant toolhead. It is a decoupled system, because the robot is used only to provide the feeding motion of the tool relative to the part, whereas the toolhead is used only to provide the tool force for deburring. The robot has five axes, composed of a three-axis parallel tripod and a two-axis gantry. A full description of the robot is given in [17]. Figure 2 provides a picture zooming on the compliant toolhead that is attached to the moving platform of the robot. This toolhead is an active axial-compliant force device. It is made of a housing and a moving mount. A pneumatic tool spindle is held in the centre of the moving mount and guided axially by a linear bushing inside of the housing, which is fixed to the robot’s moving platform. There are three pneumatic cylinders that are evenly distributed and constrained to extend and retract the spindle in the tool axial direction.
L. Liao et al.
Pressure sensor
Gantry Tripod robot
Part Toolhead Control valve
Fig. 1 The robotic deburring system developed at Ryerson University
Linear encoder Moving platform
Spindle
Housing Pneumatic cylinders
Moving mount Tool
Fig. 2 The compliant toolhead with on-line burr measurement
Furthermore, as shown in Fig. 2, a linear encoder is added to measure the tool length when extending and retracting. Through this sensing, the burr height can be determined in real time by comparing the measured tool length with a set value along a planned tool path. This robotic deburring system provides two ways of tool path planning. If the part has a CAD model, a tool path can be created by an in-house software package called P-CAM. Figure 3 shows a snapshot of this software. If the part does
Diameter or width Length Spindle speed Stroke Simulation stroke Grid number Structure number Sphere
Cone
Square
Cylinder
Force Cutting plane
Refresh
Next step
Machine type Workpiece type
Fig. 3 CAD-based tool path planning
Robotic Deburring Based on On-line Burr Measurement
215
Fig. 4 Reverse engineering-based tool path planning
not have a CAD model, a reverse engineering method is used to measure and curve fit the edge profile, based on which a tool path can be generated, as shown in Fig. 4.
burr, (3) Poisson burr, and (4) tear burr. The roll-over burr is a leftover chip remaining on the edge of the part. The cut-off burr is formed due to material breakage. The Poisson burr is the lateral protrusion during cutting. The tear burr is formed on the edge when the material is torn, such as in punching. The burr location is generally predictable either at the entrance of cutting or at the exit, with the latter being dominant. However, the burr geometry varies. Some experimental research has been done on milling and drilling. Table 1 lists the basic types of burr geometry for milling [18] as well as those for drilling [6]. The experiment results in [18, 19] show that the depth of cut is the primary factor for the burr height; feed rate and cutting speed are secondary for the same tool geometry. Based on the afore-mentioned literature on burr classification, a software package has been developed at Ryerson that can simulate the burr geometry for basic machining operations, such as milling and drilling. As shown in Fig. 5, for a selected machining operation, burr geometry can be generated and used for the subsequent deburring control simulation, as discussed in the subsequent section.
3 Modeling of Deburring Process 3.2 Deburring Modeling 3.1 Burr Classification Burr classification is essential for modeling and simulation of the proposed deburring control method. In the literature [18, 19], the burrs resulting from the metal removal process are categorized into four types: (1) roll-over burr, (2) cut-off
As mentioned before, the objective of the proposed deburring control method is to maintain a constant tool length along a planned tool path. In this section, a dynamic model of the tool system is presented based on the axial compliant tool model, as shown in Fig. 6.
Table 1 Burr geometry
Machining method
Milling [18]
Burr geometry Knife type burr
Wave-type burr
Machined surface
Machined surface Burr height
Drilling
[20]
Uniform burr
Cur-type burr Machined surface Burr height
Crown burr
Edge breakout Machined surface Burr height
Uniform burr with drill cap
Secondary burr Machined surgace Burr height
Irregular burr
Burr height
216
L. Liao et al.
Milling
Drilling
Feed
Turning
Depth of cut
Cutting speed
Knife type Deburring simulation
Curl type Wave type
Machined surface
Edge breakout Secondary
Burr height Burr height before
Burr height after
Fig. 5 Simulation of burr geometry
In deburring, the tool is pressed against the part by three single acting pneumatic cylinders with spring return in the tool axial direction, denoted by z. The tool dynamic model is derived treating the tool system as a spring-damper-mass system subject to the control force and contact force
M¨z + C˙z + Kz = up − F
(1)
where M is the tool mass, C and K are the damping and stiffness of the spring inside the cylinder, respectively, up is the control force, and F is the contact force between the part and the tool. It is noted that the tool axial direction is the tool length direction. In other words, variable z represents the tool length. The contact force in Eq. (1) is derived as follows. Since the deburring tool under study is abrasive, in light of the micro cutting model [21], the average micro depth of cut
considering the multiple grains of the abrasive deburring tool can be expressed as h=
F/N 2π rg HB
where HB is the hardness of the part (N/mm2 ), rg is the average radius of the grain (mm), and N is the number of grains inside the contact area. Assuming that the part is cut by h for each rotation, then the velocity at which the tool cuts into the part is equal to the spindle rotational speed ω multiplied by h as z˙ = ωh =
ωF/N 2π rg HB
(3)
where the tool rotational speed ω is expressed in rps (revolutions per second). From Eq. (3), the contact force can be derived linearly proportional to the tool axial cutting velocity (the same direction as the tool length) as F = Ce z˙
(4)
where Ce is given as
Moving Platform Fc = Pc Ac K
z
(2)
Ce =
2π rg HB N ω
(5)
Cylinder
Substituting Eq. (4) into Eq. (1) yields the dynamic model of the tool system in terms of tool length variable z as Tool
M Ce
M¨z + (C + Ce ) z˙ + Kz = up
F Contact Model y
Fig. 6 Tool system modeling
(6)
Taking Laplace transform on both sides of Eq. (6) yields the tool length response function for the open loop control
Robotic Deburring Based on On-line Burr Measurement (zd) Desired Tool Error(e) Length + −
Input up PID Controller
1/(Ms2 + (C + Ce )s + K)
217
(z) Tool Length
Fig. 7 Closed-loop tool length control block diagram
system as Z(s) =
Ms2
Up (s) + (C + Ce ) s + K
(7)
Based on Eq. (7), a closed-loop tool length control can be formed for the deburring process and Fig. 7 shows its block diagram. This is the control method based on on-line burr measurement, because the tool length signal is fed back to compare with the desired tool length. The control effort is generated based on the tool length error by applying a tuned PID controller. The transfer function for the proposed tool length control is derived as Up (s) = Kzp Ez (s) + Kzi
Ez (s) + Kzd sEz (s) s
(8)
where Ez (s) = zd (s) − z(s)
(9)
Kzp , Kzi , and Kzd are the proportional gain, integration gain and derivative gain, respectively.
3.3 Deburring Simulation Simulation is carried out using Matlab Simulink. Figure 8 shows the Simulink model built for the closed-loop tool
Fig. 8 Simulink deburring model
Fig. 9 Deburring simulation result
length control described in the preceding section. Figure 9 is an example of simulation on a knife-type burr generated using the software shown in Fig. 5. The burr height is 0.5 mm, burr length is 5 mm, and the part length is 200 mm. For the given feed rate of 1 mm/s, the total traveling time is 200 s. As shown in Fig. 9, the tool starts moving in the area without the burr, then encounters the burr at the time of 120 s and leaves the burr at the time of 170 s. The system parameters used for simulation are given in Table 2. These parameters represent the tool system used in the experiment. Through tuning, the control gains Kzp , Kzi , and Kzd are selected at 3500, 200, 400, respectively. As shown in Fig. 8, a zero order hold method is used to discretize the tool deburring motion along the feeding direction. This discretizion allows for the simulation of the tool performing deburring at each discretized point in consideration of the tool response function in Eq. (7). For the given feed rate, the time interval between two adjacent discretized joints is 1 s. Figure 9 shows the simulated deburring
218
L. Liao et al.
Table 2 System parameters Description
Symbol
Value [units]
Tool mass Tool stiffness Tool damping Hardness Radius of the grain Grid number in contact area Tool rotation speed
M K C HB rg N ω
0.1528 [kg] 351.8797 0.1 300 21.1e-6 40 20 [rev/sec]
result after one deburring path, which results in a burr reduction about 0.06 mm. Further simulation results show that the lower the feed rate, the higher the burr reduction. Part
4 Deburring Control Experiment The proposed deburring control based on on-line burr measurement has been implemented on the robotic deburring system described in Sect. 2. Figure 10 shows the block diagram for the entire deburring process including control system and tool/part interaction. The complete tool system also includes a pneumatic valve and three pneumatic cylinders that generate the tool force, i.e. up in Eq. (6). Due to the high response rate requirement, a flow valve is used instead of a pressure valve as done traditionally. Since the flow valve exhibits strong nonlinearity, a linearization method is developed to establish a proportional relation between the control voltage and the control force around a set point. The details are presented in [22]. The first experiment presented here is the deburring of the edge of an aluminum plate. Figure 11 shows the experiment set-up. The control gains Kzp , Kzi , and Kzd for the actual control system are tuned at 0.9, 1, and 0.06, respectively. They are different from the simulation due to uncertainty in the actual control system. Since in this case the nominal tool path is a straight line, the desired tool length is set at a constant value of 4.5 mm. The objective of the proposed control method is to maintain the tool length at 4.5 mm. The experimental result of the deburring control is given in Fig. 12. It can be seen from Fig. 12 that the total deburring period is approximately 60 s. (zd) Desired Tool Length + –
error PID Controller
(Vp) Voltage Input up
Fig. 10 Block diagram for the entire deburring process
Fig. 11 Edge deburring experiment
The tool encounters the burrs in the first half of period as separated by a vertical dashed line in Fig. 12. When hitting the burr, the tool is pushed upward to shorten its length, i.e. below 4.5 mm, as shown in the top figure of Fig. 12. This negative tool length change informs the controller that there is a burr, hence to raise the control voltage to increase the control force, as shown in the bottom figure of Fig. 12. In the second half, the top edge sags, causing the tool to extend, i.e. above 4.5 mm, as shown in the top figure of Fig. 12. Correspondingly, the control voltage reduces to decrease the control force to avoid over-cut, as shown in the bottom figure of Fig. 12. One may notice from Fig. 12 that though the control voltage signal follows the tool length change, the response is not sufficiently sharp. This is due to the long time delay in the pneumatic system. An electrically actuated tool system would provide a better response rate. The second experiment is on the same part to investigate the burr reduction rate for our deburring system. Figure 13
(Pc) (F) Electrical Pressure Pneumatic Force Pneumatic Cylinder Valve
Tool/Part Interaction
(z) Tool Length
Robotic Deburring Based on On-line Burr Measurement
No burr area
219 Table 3 Profile variance and mean value before and after deburring Profile mean (mm) Profile variance Original profile First deburring Second deburring
0.8437 0.7648 0.7337
0.0414 0.0302 0.0209
Burr area
Control force ↑
Control force ↓
Fig. 12 Experimental results of deburring control; top – the time history of the tool length; bottom – the control signal of the pneumatic valve
shows the burr height changes after two paths of deburring. The edge profile is measured using a 2-D laser profile scanner. The resolution of the profile scanner is 10 μm in the z-direction, i.e. in the same direction as the burr height. It can be clearly seen from Fig. 13 that the burrs are effectively removed by applying the proposed closed-loop tool length control method. At the locations without burrs, the part geometry is maintained as no over-cut, due to the regulation of the tool length. To quantify the burr reduction rate, the mean value and variance of the edge profile are obtained based on the measurement data given in Fig. 13. Table 3 lists the calculation results. The mean value represents the average height of the profile, and the variance represents its roughness. It can be seen that the mean value is reduced
Fig. 13 Burr reduction before and after deburring
by 0.08 mm from the first path of deburring and again by 0.03 mm from the second path. Hence the average burr reduction is about 0.05, which is close to 0.06 of the simulation result presented in Sect. 3.3. Furthermore, a small variance indicates a smoother part surface. It can be clearly seen from Table 3 that the profile variance value decreases as the number of deburring paths increases, meaning that the edge profile becomes smoother. The third experiment is on a zigzag path, mimicking the deburring of the fir trees of a turbine disk. The experiment
Fig. 14 Zigzag path deburring mimicking the fir trees of the turbine disk
Fig. 15 Zigzag path fitting
220
Fig. 16 Zigzag tool path planning
set up is shown in Fig. 14. Since this is an in-house prepared test piece, there is no CAD model. Therefore, the reverse engineering-based approach is used to measure and fit the zigzag profile as shown in Fig. 15, based on which the tool path is generated as shown in Fig. 16.
5 Concluding Remarks A full discussion has been provided on the new deburring control method presented in this paper that is based on on-line burr height measurement. This control method is developed using an axial compliant toolhead equipped with a linear encoder that can sense the tool length change. A closed-loop tool length control method is put forward, modeled, simulated and implemented. Both simulation and experiment results have demonstrated the effectiveness of the proposed deburring control method. The main advantage of this method over the existing deburring control methods is that it can not only measure the burr height, but also adjust the control force accordingly, to avoid under or over cut, and possible tool breakage for large burrs. The developed deburring simulation and control system has been proven practically useful; hence the authors are keen on commercialization.
References 1. Nakayama, K., Arai, M., 1987, Burr formation in metal cutting, Annals of the CIRP 36(1):33–36.
L. Liao et al. 2. Ko, S.-L., Dornfeld, D.A., 1991, A study on burr formation mechanism, Journal of Engineering Materials and Technology, 113: 75–87. 3. Chern, G.-L. Dornfeld, D.A., 1996, Burr/breakout model development and experimental verification, Journal of Engineering Materials and Technology, 118:201–206. 4. Hashimura, M., Chang, Y.P., Dornfeld, D.A., 1999, Analysis of burr formation mechanism in orthogonal cutting, Journal of Manufacturing Science and Engineering, 121:1–7. 5. Klocke, F., Hoppe, S., Fritsch, R., 2004, FE-modeling of burr formation in orthogonal cutting, in: Proceedings of the Seventh International Conference on Deburring and Surface Finishing, 7–9 June, University of California, Berkeley:47–56. 6. Min, S., Dornfeld, D.A., Kim, J., Shyu, B., 2001, Finite element modeling of burr formation in metal cutting, Machining Science and Technology, 5(3):307–322. 7. Huissoon, J.P., Ismail, F., Jafari, A., Bedi, S., 2000, Automated polishing of die steel surfaces, International Journal of Advanced Manufacturing Technology, 19:285–290. 8. Tam, H., Lui, O.C., Mok, A.C.K., 1999, Robotic polishing of freeform surfaces using scanning paths, Journal of Materials Processing Technology, 95:191–200. 9. Einav, O., 1995, Large work envelope fully-automated aircraft panel polishing cell, in: Proceedings of the International Robotics & Vision Automation Conference, 9–11 May, Detroit, Michigan. 10. Sasaki, T., Miyoshi, T., Saito, K., 1991, Knowledge acquisition and automation of polishing operation for injection mold, Journal of the Japan Society of Precision Engineering 25(3):193–199. 11. Kawata, K., Sawada, Y., Yamashita, M., 1992, A new method of teaching and path generation for automatic die and mold polishing system, in: Proceedings of Japan/USA Symposium on Flexible Automation:971–974. 12. Proctor, F.M., Murphy, K.N., 1989. Advanced deburring system technology, ASME Winter Annual Meeting, 10–15, December, San Francisco, CA, PED-38. 13. Rasmussen, B., Derby, S., 1988, Design and evaluation of robotic end-effectors for an automated die finishing system, in: Proceedings Symposium on Computer-Aided Design and Manufacture of Dies and Molds (ASME Winter Annual Meeting):61–74. 14. ATI Industrial Automation, http://www.ati-ia.com/index.aspx 15. PushCorp, Inc., http://www.pushcorp.com 16. RAD, The Robotic Accessories Leader, http://www.rad-ra.com/ index.htm 17. Xi, F., Liao, L., Mohamed, R., Liu, K., 2008, A tripod-based polishing/deburring machine, in Smart Devices and Machines for Advanced Manufacturing, eds. Wang, L., Xi, F., Springer: 137–165. 18. Chern, G.-L., 2006, Experimental observation and analysis of burr formation mechanisms in face milling of aluminum alloys, International Journal of Machine Tools & Manufacture, 46(12–13): 1517–1525. 19. Oliver, O, Barrow, G., 1996, An experimental study of burr formation in square shoulder face milling, International Journal of Machinery Tools and Manufacture, 36(9):1005–1020. 20. Kim, J., Min, S., Dornfeld, D.A., 2001, Optimization and control of drilling burr formation of AISI 304L and AISI 4118 based on drilling burr control charts, International Journal of Machine Tools & Manufacture, 41(7):923–936. 21. Xi, F., Zhou, X., 2005, Modeling surface roughness for the stone polishing process, International Journal of Machine Tools & Manufacture, 45(1):365–372. 22. Liao, L., 2008, Modeling and control of automated polishing/deburring, Ph.D. Thesis, Ryerson University, Toronto.
Deburring Machine for Round Billets – Equipment for Efficient Removal of Burrs from Billets M. Schnabl
Abstract Cutting burrs attached to round metal strands cause serious problems for the further processing in rolling mills and finishing lines. Therefore the removal of the cutting burrs is a necessary pass. State of the art methods for removal of burrs from round billets are manual treatment with pneumatic breakers or flame cutting devices. The introduced deburring machine is a new invention which is designed to automatically remove burrs from circular strand casting billets. Keywords Deburring machine · Cutting burrs · Round billets · Strand cast billets
1 Introduction Fig. 1 Round billet with cutting burr
In steel mills endless steel strands are produced by the continuous casting process. Oxygen cutting installations are used to separate the endless strands into single pieces. During the oxygen cutting process a cutting burr, consisting basically of ferric oxides which pass over into a bond bridge of pure steel steadily, arises at the bottom edge of the separation plane. Figure 1 shows an example of a cutting burr at the end of a flame cut round billet. The existence of these cutting burrs results in many very cost intensive and quality reducing problems for the further processing. A summary of the central problems is given below: • Transportation: When a billet with a cutting burr passes the roller conveyor, the cutting burr causes heavy shocks on the rollers thus leading to reduced life time or worst case to damage of the rollers.
M. Schnabl () framag Industrieanlagenbau GmbH, Neukirchnerstrasse 9, 4873 Frankenburg, Austria e-mail:
[email protected] url: www.framag.com
• Rolling process: When a billet with a cutting burr is drawn into the roll stand, the cutting burr induces a shock that exceeds the big shock which arises anyway due to the drawing in of the billet. The result is an increased mechanical wear of the whole installation and especially of the cylinders. • Quality of the final products: When the billet with the cutting burr is transported along the roller conveyor or drawn into the roll stand, the ferric oxide of the burr can be rolled into the pure parent material. This leads to a final product with added impurities and therefore to final products with insufficient quality. For this reasons it is clear, that the precondition for a safe rolling or finishing process and high quality final products is a preliminary product without cutting burrs.
2 State of the Art There are some established machines for the mechanized deburring of plate slabs and billets with rectangular cross
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_24, © Springer-Verlag Berlin Heidelberg 2010
221
222
sections. In the past some attempts and experimental installations have been made in order to find an automatically solution for deburring of round billets, but up to now there was no reliable solution on the market. So till today the only method to remove the cutting burrs from circular billets was by manual treatment under the use of pneumatic breakers, abrasive cutting machines or flame cutting devices. For this process, the cold billet is taken out of the line by a crane and put onto an appropriate support. Then the billet has to be manually turned to an appropriate working position. After the cutting burr is cut or chiseled off, the billet has to be removed by a crane. This procedure is exceeding costly in terms of labor and with regard to the necessary output for the following process this work has to be operated 24 hours a day. Furthermore a manually treatment is always a source for failure and varying quality.
3 The New Approach Derived from established solutions for the automatically deburring of plate slabs and rectangular billets, the approach is to use planning tools for the deburring of the round billets. Several planning tools are held in a radially displaceable manner in a tool support. The billet has to be positioned in such a way, that the axis of the billet is centric to the planning tools. When the planning tools move to the center axis of the billet, they hit the surface of the billet and then they are pressurized against the surface with a preterminated force. The working stroke for deburring of the cut metal strand is achieved by a relative movement between the planning tools and the metal strand in axial direction to the metal strand. Figure 2 shows the principle of the new approach within a test device.
M. Schnabl
4 Constructional Implementation of the Deburring Machine The target was to design a reliable apparatus for automatically deburring of round billets in such a way, that advantageous constructional conditions are obtained for the removal of the cutting burr and that the machine matches to the rough ambient conditions in a steel mill or a finishing line.
4.1 Basic Strategy Especially advantageous conditions are obtained, when the feeding drive for the removal of the cutting burr is done with a punch that extends in the longitudinal direction of the billet and that moves the billet relative to the planning tools acting on the face side of the end of the billet. The planning tools are assembled on a stationary tool support. Therefore no additional fixing of the billet is required for the deburring process and no feed drive is needed for the tool support. The roller conveyor, which needs to be provided anyway for the feeding and removing of the cut billets can thus advantageously be used as a guide means during the advancing movement by the punch. The tool support comprises at least two sets of annular distributed planning tools which are arranged behind one another in the longitudinal direction and which are staggered against each other set by set. The provision of minimum two sets of mutually staggered planning tools offers a simple possibility of fully deburring round billets of different diameters because a cutting sequence of the individual planning tools which closes over the circumference is obtained through the mutual offset. The precondition is an annular tool support for the planning tools, which encloses the round billet and which planning tools are adjustable themselves radial relative to the fixed support. If the planning tools are distributed over the entire circumference this is furthermore an advantageous precondition for deburring of a round billet, because it is not necessary having knowledge about the circumferential position of the cutting burr. When the punch for feed drive is realized in kind of an actuating cylinder, the punch can also be used as a breaking stop for the cut billets that are supplied via the roller conveyor to the deburring apparatus.
4.2 Basic Requirements to the Deburring Machine
Fig. 2 Test device with radially arranged planning tools
To match the ambient conditions in a steel mill or finishing line the deburring machine has to fulfill several basic requirements, the most important of them are described in the
Deburring Machine for Round Billets – Equipment for Efficient Removal of Burrs from Billets
following. The first fundamental requirement is the achievable output of the deburring machine. The output of a steel mill should only be limited by the achievable casting velocity, subsequent rectification work mustnt´ be a limiter for the output. This is of big importance for the concept of the deburring machine. The cycle time of the deburring process must be in a magnitude, that the deburring machine can work inline with the continuous casting and the necessity to manipulate the cut billets should be as low as possible in order to achieve a good integration into the continuous casting line. The range of workable billet diameters should be as wide as possible, whereas the setup time for different diameters should be kept as low as possible. In case of tool wear the change of the planning tools has to be kept short and simple. The second basic requirement is the reliability of the deburring quality. The machine has to support an adequate good deburring result in a repeatable and reliable manner. The quality of the deburring result is basically affected by the parameters of the planning tools, the material of the billet and the identification of the right process parameters. The third basic requirement is to stop the billets, which move along the roller conveyor with a certain speed in such a way, that they are concentric to the radial arranged planning tools and at zero speed in the start position for the deburring sequence. The necessary workable diameter range and the different lengths of the cut billets determine a wide range of different unit weights and in combination with the variable speed of the roller conveyor a wide range of different kinetic energy of the moving billets, which has to be operated by a damping system accurate and in very short time in order to achieve the specified cycle time. A further basic requirement, that is valid for all kind of installations in the steel mill industry is, that the design of the machine must be extremely robust to meet the heavy duty use and to guarantee highest availability.
Fig. 3 Design model of the deburring machine
cylinder in order to position the axis of the different diameters centric to the radially assembled planning tools. Figure 4 shows the tool support of the deburring machine, the so called “cutting star”. The base body of the cutting star is an annular steel plate which is used as assembly platform
4.3 Description of the Deburring Machine Figure 3 shows a design model of the prototype of the deburring machine. This prototype is able to operate a diameter range from 180 up to 450 mm and a length of the cut billets from 4m up to 12m. The maximum allowable speed of the roller conveyor is up to 30 m/min. The outer base frame of the machine is a massive welded steel construction which is fixed to the foundation. A V-shaped roller is fixed to the outer frame aligned with the roller conveyor of the casting or finishing line. The inner frame of the machine is also a welded steel construction and is vertically guided relative to the outer frame. The inner frame represents the supporting structure for the annular planning tools and can be positioned vertically by a hydraulic
223
Fig. 4 Tool Support of the deburring machine
224
M. Schnabl
Fig. 5 Possible line setup for inline deburring
for the planning tools. The planning tools consist of a housing that constitutes the guidance for the radially displaceable rectangular mandrel. The planning blade is assembled to the tip of the mandrel. The feed motion of the mandrel is performed by a hydraulic cylinder. This hydraulic cylinder also performs the pressure which is necessary to squeeze the planning blade against the surface of the billet during the deburring sequence. On the top of the base plate of the cutting star there is a hook shaped support which allows to hang up the cutting star on the inner frame. For the purpose of fixing the cutting stars in the longitudinal direction, an axial clamping device is provided which clamps the cutting stars against the inner frame. This design allows a tool change in a very simple and fast kind. The change of the whole cutting star requires only opening the axial clamping and then lifting the cutting star by crane out of the machine. To achieve a ring of planning tools which is closed in the circumferential direction for all billets in the specified range from 180 up to 450 mm, the prototype is equipped with three cutting stars whereas the planning tools of the individual cutting stars feature a respective angular offset. The function for breaking the billets, running in from the roller conveyor, and the function for the axial feed drive of the billets for the deburring stroke is integrated in one device called “punch cylinder”. An especially developed hydraulic damping system is able to handle the wide range of kinetic energy of the moving billets and to perform the necessary power for the deburring stroke. A full deburring cycle consists of following steps: • Running in of the billet along the roller conveyor • Breaking the billet to zero speed and positioning to start position for deburring stroke with the punch cylinder
• Feeding the planning tools to the surface of the billet and performing the preconditioned deburring pressure • Deburring stroke with the punch cylinder • Running out of the deburred billet The cycle time for this deburring sequence is about 25 s, this is short enough to enable the use of the deburring machine inline to the continuous casting process. Figure 5 shows one possible production line setup for the inline removal of the cutting burrs from both ends of a cut billet. The cut billet (1) is running in along a roller conveyor. In the next step, the cut billet is transported in transversal direction along a stacking table to the roller table (5). The roller table is assembled in the center line of the two deburring devices (3) and (4). This roller table is necessary to feed the billet first into the left (3) and second into the right (4) deburring device where burrs from the left and right end of the billet are removed. After the removal of burrs the billet is once more transported in transversal direction and is ready for further processing.
5 Deburring Results with the Test Machine A close to production test machine was built in order to check the functions of the mechanical and hydraulic systems on the one hand and to define the ideal process parameters for obtaining the best possible deburring result on the other hand. The design of the close to production test machine is shown in Fig. 6. The most extensive work was to identify the right process parameters to achieve an adequate good and reliable deburring result. In the beginning of the test series the
Deburring Machine for Round Billets – Equipment for Efficient Removal of Burrs from Billets
225
Fig. 8 Optimum deburring result due to optimized process parameters
Fig. 6 Design of the close to production test machine
identification of the right geometry of the planning blade was in the focus. A wrong geometry leads to the result that the planning blades carve deep into the surface of the billet. In this case the surface of the billet features heavy cinch marks after the deburring stroke. Figure 7 shows a deburring attempt with extremely deep cinch marks due to wrong blade geometry.
Fig. 7 Heavy cinch marks on the surface due to wrong blade geometry
Another important thing in this context was to find the right parameters for the prestress of the radial arranged planning tools. Too small prestress can effect that the burr isnt´ removed completely because the blade skips over the burr. Contrariwise an oversized pressure leads to cinches on the surface of the billet and increases the necessary power for the feed drive during the deburring stroke. With some theoretical analysis and some experience it was possible to optimize the process parameters of the prototype in such a way, that the deburring machine performs a deburring result of outstanding quality in a reliable manner. Figure 8 shows a billet after the deburring process with optimized parameters. It is in evidence, that the burr is removed hundred per cent and there are only negligible cinch marks on the surface of the billet. Figure 9 shows the associated removed burr.
Fig. 9 Cutting burr remove from the billet with optimized process parameters
226
6 Conclusion Experimental results of the close to production test machine prove that the approach with the radially arranged planning tools is the right way for deburring round billets. A deburring machine of this kind allows the deburring of round billets
M. Schnabl
with an outstanding quality and superior reliability. The principle is appropriate to a design that meets all requirements in the steel mill industry. The cycle time of the whole process is short enough, that an inline installation is possible in the continuous casting line without extensive additional devices for transport operations.
Removal and Cleanability
Formulation of the Chip Cleanability Mechanics from Fluid Transport S. Garg, D. Dornfeld, and K. Berger
Abstract The presence of solid particle contaminant chips in high performance and complex automotive components like cylinder heads of internal combustion engines is a source of major concern for the automotive industry. Current industrial cleaning technologies, simply relying on the fluid transport energy of high pressure or intermittent high impulse jets discharged at the water jacket inlets of the cylinder head, fail to capture the dynamics of interaction between the chip morphology and the complex workpiece landscape. This work provides a preliminary insight into an experimental investigation of the mechanics of chip transport at play, and how it can be used to build an effective chip optimization model that significantly aids in improving the cleanability of contaminant chips. The objective is to relate the mechanics of chip transport with the chip form parameters as much as possible, which makes the objective and constraints in the optimization model quantifiable. The end objective is of course to transmit this information upstream of the manufacturing pipeline in the form of a Design for Cleanability (DFC) feedback, which highlights the industrial cleaning problem as a design centric issue. Keywords Design for cleanability · Workpiece bottleneck · Chip critical dimension · Fluid lift force · Fluid drag force · Chip orientation
S. Garg (), D. Dornfeld Mechanical Engineering Department, University of California, Berkeley, CA 94720-1740, USA e-mail:
[email protected] url: www.lmas.berkeley.edu K. Berger Daimler AG, Material Technology Department, 70546 Stuttgart, Germany
1 Introduction Chip formation is a very significant aspect of any basic machining process. Unfortunately chip production has conventionally been considered to be only of secondary importance for a production process. The most important emphasis has been given to the finished workpiece and its attributes like surface roughness, ease of machining, dimensional accuracy etc., and in mass production, to achieving the desired production rate and quality control at lowest costs to the facility or the company. The type of chip produced is a good indicator of the machining conditions, the properties of workpiece, tooling material and the quality of machined surface while chip load is an effective measure of cutting forces. But in mass manufacturing, where most of these essential elements of production process are known or fixed at the design stage itself in order to achieve specified values of final properties like hardness, toughness, surface quality etc., and due to cost constraints governing optimum tool life; chip formation is often ignored as a potential area for directing concern. The firsts of the manufacturing concerns related to chip formation dealt with the obstacle provided by the chip generation to the motion of the tool. A simple example is the flute of a drill bit whose function is to allow chips from the workpiece to easily climb up and create way for further drilling. In similar ways, much work has been done in the industrial research to break the chips and remove them from the machining area to facilitate further machining. The challenges due to the chip formation take a major role because of their capacity to act as solid particle contaminants for high performance mechanical components like engine cylinder heads. Contamination of mechanical parts due to process impurities and machining chips that get embedded onto their surface is an important industrial concern as far as the functionality of these parts is concerned because it can lead to their sudden and catastrophic failure during the use phase. Thus, cleaning of mechanical components assumes great significance from an industrial standpoint.
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_25, © Springer-Verlag Berlin Heidelberg 2010
229
230
Contaminant chips can either be the chips from the common machining operations that enter the narrow intricate channels of workpiece geometry, or they may be loose burrs which remain in the workpiece after detachment from the surface. Important work on burr formation and minimization (size, location, shape, etc.) has been done by Dornfeld [1] and co-workers. Although the problem of cleanability of mechanical components has greater complexities compared to deburring such as lack of easy accessibility requiring fluid jets with sufficient pressures and velocities in most internal parts of the workpiece; a complex nature of types of contaminants such as loose chips, sand mold particles, surface grease, dirt etc., it is important to combine the knowledge of both to build a rigorous chip optimization model. The chip morphology plays an important role in influencing the cleanability of chips. The most commonly used industry standard for chip classification based on morphology is the ISO-Classification 3685 [2], as is used in this work. Literature review shows that some amount of research has been done involving the chip size and geometry as applied to machining technology. Reich-Weiser and Dornfeld [3] presented an experimental investigation of the influence of machining parameters on chip geometry for enhanced cleanability. In their work, process parameters such as feedrate, spindle speed, depth of cut, and tool lubrication have been correlated with chip geometry and size for milling and drilling processes. Viharos et al. [4] provided an ANN based chip-form classification model for the turning process. Their work involved building predictive models for cutting chip form based on process parameters, direct measured values of cutting forces and torques, and calculated factors of typical monitoring signals. Sturenberg [5] worked on optical classification of chips generated during the cutting or the machining phase as well as the assembly phase based on chip geometry, and then modifying the tool design for easy removal of such chips from the machined workpiece based on the types of chips specific to that process for which the tool is used. Ávila et al. [6] performed a black box experimental approach of testing the cleanability of different chip geometries using the Impulse cleaning Machine. He found that chip form selectively affects the cleanability of different chip types. The current work focuses on tracing the root of cleanability at the lowest level: which involves local interactions between workpiece and chip geometry and chip-cleaning fluid. Brute force methods for cleaning currently used in industry for increasing water pressures in hope of increasing cleaning force may not be effective and rather may prove more costly in terms of expensive and complicated equipments required for higher pumping power. It is felt that although it is difficult to exactly formulate a theory for such kind of chip-workpiece-fluid interactions, understanding and establishing some kind of a framework through experimental investigation to assess and analyze the impact of the chip size and geometry vis-à-vis the critical bottleneck dimensions of
S. Garg et al.
the workpiece (narrowest channels of engine cylinder head in the current case) can give us the key to understanding the fundamentals of cleanability problem. It would then be meaningful to attack the problem at the grass-root level by focusing on all the stages of manufacturing pipeline starting from the design stage keeping in mind the requirements for generation of only those chip forms that are amenable to cleaning by our existing methods and also keeping a bound on chip sizes during the machining stages so as not to generate chips that are comparable in size to the bottleneck internal dimensions of the workpiece. Thus, integration of cleanliness as an engineering constraint into the early stages of the product development process is critical to the manufacture of mechanical components that meet cleanliness requirements at the lowest possible costs.
2 Experimental Setup The experiments were performed at the production department lab, Daimler AG, Stuttgart. The diesel engine cylinder head OM 646 was used for the purpose. A cross section of the cylinder head passing through the water jacket was first made in Catia CAD software, to ensure two basic criteria are met: firstly, the volumetric flow rate of water passing through the water jacket does not fall below 75–80% of the original amount, so that actual cleaning conditions are maintained as much as possible. Secondly, the section should be made at an appropriate height in the water jacket so that the emerging landscape has characteristic bottlenecks representative of the actual cylinder head. A plexiglass sheet was used to cover the cross section of the head, and was clamped in position with four clamps. Chips of different geometries such as helical, spiral, ribbon etc. and varying sizes were precollected, washed of the lubricant using an ultrasonic vibrator and then separated for use in the experiment. Figure 1 shows the experimental setup used.
Fig. 1 Experimental setup for investigating chip transport through a cut OM 646 cylinder head
Formulation of the Chip Cleanability Mechanics from Fluid Transport
3 Formulation of Cleanability Mechanics: Observations and Inferences As was described above, the purpose of the experiments was to observe and formulate the chip cleanability mechanics and relate it with the chip form parameters as much as possible in order to build an effective chip optimization model that significantly aids in improving the cleanability of contaminant chips. The following observations highlight the mechanics of the chip cleanability at play.
231
ers across its width and a particular feature in the internal landscape is analyzed for its dimension across those layers until a minima is reached somewhere. That establishes the bottleneck dimension for that feature of interest. 2. Using the negative of the cylinder head, i.e. the sand mold: The Al-Si cylinder heads mostly used in automotive industry are manufactured from the sand-casting process. The sand-mold acts as the negative of the cylinder head and thus the dimensions of the mold represent the internal dimensions of the channels and holes of the cylinder head geometry. Methods to estimate the chip critical dimension:
3.1 Chip Critical and Workpiece Bottleneck Dimensions It was observed that as is intuitively expected, the chip critical dimension should be smaller than the workpiece bottleneck dimension, in order for the chips to overcome the bottleneck. A workpiece bottleneck may be a narrow constriction in the width of a channel, or it may be a height obstacle in the cross-sectional landscape which requires the chip to elevate itself above a threshold barrier. However, there are no existing methods or approaches that clearly delineate how to evaluate the chip critical and workpiece bottleneck dimensions. Methods to determine the bottleneck dimensions of the cylinder head: 1. Using a sequence of cut cross sectional layers: The CAD model of the cylinder head is sectioned into a series of lay-
Fig. 2 Critical dimension for the ribbon chip = l
Fig. 3 Basic geometrical measurements in a (a) flat spiral, (b) raised spiral
The specific geometry type of the chips can be exploited to estimate the chip critical dimension (Fig. 2). For ribbon chips, one can estimate the size based on a direct adaptation of the ISO recommendation [7] for estimating particle size, i.e. the size of the particle is determined by its longest dimension, which is the longest distance between any two parallel lines drawn touching the surface of the particle. For the spiral chips, the authors propose that these chips can be modeled on the lines of a regular cylinder, which bounds the smallest volume of space completely enclosing the chip. This is the minimum volume of a known regular geometry that can very accurately model a spiral chip. For the Fig. 3(a) which represents a flat spiral, the easily measurable parameters that are sufficient to determine the critical dimension are the maximum spiral diameter (“D”) and the chip width “h”. For the Fig. 3(b) which represents a raised spiral, the height of the spiral will be
232
S. Garg et al.
width “s”. The chip critical dimension is then expressed by Eq. (2). Conical Chip critical dimension = larger of “D” or (s2 + Dd) Fig. 4 (a) Modeling of relevant chip forms on the lines of a regular cylinder with appropriate parameters, (b) rotation of a rectangle generates a cylindrical surface of revolution
different from the chip width and thus needs to be recorded separately. Figure 4(a) shows a cylindrical geometry with the parameters adapted from the spiral chip. The cylinder can also be envisioned as a surface of revolution of a rectangle around the axis of rotation shown. The critical dimension of the chip is then given by the largest diagonal of the rectangle, as shown in Eq. (1). Spiral chip critical dimension = D2 + h2
(2)
The mass of the spiral chips (flat and conical) can also be calculated from a knowledge of the geometrical parameters of the chips. When these chips are compressed slightly to align the spirals with one another, they can be modeled on the lines of a hollow cylinder and a cone (frustum) respectively, with known mensurational expressions for determination of volume, which along with a knowledge of the density of chip material can give a good geometrical estimation of the mass of these chips.
3.2 Importance of Chip Projected Surface Areas: Role of Fluid Drag and Lift Forces (1)
The critical dimension of other common geometries that can be modeled as an approximate cylinder include helical washer, tubular helical, raised conical spiral and conical helical. Similarly, a conical spiral geometry can also be modeled along the lines of a frustum of a cone, as shown in Fig. 5, where the measurable parameters are the maximum diameter (of the base of the cone) “D”, the minimum diameter (of the top part of the frustum) “d”, and the slant height or the chip
Fig. 5 (a) Conical spiral chip geometry, (b) modeling the geometry as a frustum of a cone, (c) an isosceles trapezium on rotation generates the solid body of revolution shown in (b)
The entrainment of a particle into a turbulently flowing fluid is important in a wide range of natural phenomenon. Entrainment refers to the motion of a particle in the bulk of the fluid rather than rolling along the surface of a bed of particles. Consequently, a particle in motion in a fluid stream must be acted upon by a drag force that acts in the direction of the fluid motion and helps to transport the particle in the direction of the fluid motion; and a fluid lift force that helps in lifting the particle against the net weight force which has a tendency to keep the particle to the surface of the fluid bed. Figure 6 shows these forces acting on a particle on a surface bed that can cause its entrainment into the fluid stream.
Fig. 6 Forces acting on a particle that experiences entrainment into a fluid stream
Formulation of the Chip Cleanability Mechanics from Fluid Transport
233
Fig. 8 Chip orientation with (a) higher projected surface area (b) smaller projected area in the direction of fluid flow (into the plane of paper)
Fig. 7 A schematic diagram of the fluid forces on an entrained chip in a flowing stream
Figure 7 is a corresponding schematic as applied to the transport of a chip in the cleaning fluid medium. The drag force can be expressed as the sum of all pressure forces in the direction of the fluid flow as shown in Eq. (3). D=
p( n. i)dA
(3)
Here “D” is the drag force acting on the chip in the direction of fluid flow, “p” is the fluid pressure, n is the normal to the surface area of the chip, and i is the direction of fluid flow. The above equation can also be expressed in the following way (Eqs. (4) and (5)) by taking the projection of the chip surface area in the direction of the fluid flow. D = p((dA n). i) (4) D = p(dAproj )i (5) Similarly, the lift force can also be seen to be proportional to the chip projected surface area in the z direction as shown in Eq. (6). L
p(dAproj )z
chip width “s” as 4 mm. The workpiece bottleneck dimension was found greater then 9 mm. Conical Chip critical dimension = larger of “D” or (s2 + Dd) = 8 mm In that respect, there is no size constraint for the passage of chip through the bottleneck. With all conditions remaining the same and assuming pressure variation across the section of the chip as constant, a good estimate of the ratio of drag forces experienced by the two chip orientations can be obtained by considering the ratio of their projected surface areas as shown in Fig. 9. Projected surface area in figure 2 1 = (D + d) (s2 − D−d = 27.11 mm2 2 2 Projected surface area in figure = π R2 − r2 = 21.99 mm2 This tells us that the increase in projected surface area in the first case is around 23%, which is one of the reasons why the drag force is higher. The other important reason is shown in Fig. 10. Because the cleaning fluid (water) under the experimental flow conditions used here, has an expected parabolic laminar profile which has the maximum fluid momentum at
(6)
The effect of projected surface areas on the cleanability of a chip can be seen in the Fig. 8 below. For the same flow conditions, the cleaning fluid was able to transport the chip out of the workpiece bottleneck for the chip orientation as shown in Fig. 8(a) while it could not in the case of Fig. 8(b). Upon measurement of the basic chip dimensions, the maximum diameter “D” (base) was found to be roughly 8 mm, minimum diameter “d” (bottom) as 6 mm and the maximum
Fig. 9 Projected surface area estimation for two different orientations of a conical spiral chip
234
S. Garg et al.
cross-section by forward water flow transport, out of it into other channels; or to allow a different orientation for more favorable chip cleanability in the same channel.
Fig. 10 A laminar fluid flow profile through the chip orientation shown in Fig. 8(b) and 9(b)
the center of the profile, the second chip orientation with a hollow concentric ring shaped projected area loses a potentially large amount of fluid drag force.
3.3 Chip Orientation Effects The orientation of chips is also critical for their effective transport and cleanability. The projected surface area of the chips changes with a change in their orientation, as noted before. In that respect, chips that have higher projected surface areas for more stable orientations will have a better chance of getting cleaned effectively. Experiments were performed to note the effect of different orientations on the overall transport of different chip geometries through a bottleneck workpiece dimension. It must be noted however that we cannot control or predict the orientation that a chip will take while passing through a bottleneck dimension. Thus geometries that inherently have a higher projected surface area for their more stable orientations will have a higher probability of getting transported.
3.4 Effect of Back Pressure on the Chip Transport The effect of back pressure was observed as an interesting effect in the experiments that has a potential for application as an important phenomenon in the redesign of next generation cleaning machines. If a closed system cleaning for a workpiece like a cylinder head is followed, intermittent cleaning by disrupting the flow of normal tap water stream through the water jacket inlet can suddenly cause a pressure drop to exist inside the workpiece across a trapped chip. Consequently there is a back pressure on a trapped chip equivalent to the difference in atmospheric pressure and the vacuum pressure inside the cylinder head. This pressure difference can be large enough to drive air into the workpiece from outside the water jacket outlet. This effect can be used to re-channelize chips that have been trapped in a critical
4 Chip Optimization Model The resulting model for chip optimization for enhanced cleanability is an outcome of incorporating and assembling the various pieces of chip cleanability mechanics. It involves addressing the three main aspects of chip attributes: mass, geometry and size. The size constraint at all times should be to keep the chip critical dimension less than the workpiece bottleneck dimension. Within this larger umbrella, a reduction in the mass of the chip (or thickness, for the same cross sectional area) will aid in a better chip transport. This is because firstly the impulse fracture ability (as noted through some prior research experiments at Daimler [5]) shows an improvement with lighter and more fragile chips. Secondly, lighter chips experience lesser net weight force and hence a consequent higher lift force, which again aids in their transport in the fluid stream. For a given mass, the next endeavor should be to optimize the chip geometry. This involves preferring those chip geometries that have a higher projected surface area in the direction of the drag force or the cleaning force; geometries that have higher projected areas inherent in their more stable orientations; and controlling the surface roughness of the chips produced in the manufacturing operations for a lowered or reduced adhesion with the workpiece surface. The model is summed up in a diagrammatic flowchart representation as shown in Fig. 11.
5 Conclusion This paper discusses the importance of following a bottomup approach in trying to address the problem of cleanability of contaminant chips in the automotive industry. Understanding and formulating the mechanics of chip transport in a fluid medium for a given workpiece landscape is essential for building a cleanability driven overall chip optimization model as well as to use the science behind the mechanics to develop the technology for a new generation of cleaning machines that are more effective and economical. The requirements for the optimal chips from a cleanability perspective, is thus an information that in turn should travel up
Formulation of the Chip Cleanability Mechanics from Fluid Transport
235
Fig. 11 A flowchart representation of a chip optimization model based on chip cleanability mechanics
the manufacturing pipeline to induce design related changes (Design for Cleanability “DFC” feedback) at both process and systems level, which highlights the industrial cleaning problem as a design centric issue. This approach should also be closely tied with a Design for Environment (DFE) feedback with an aim to minimize the environmental impact of the cleaning processes and raw materials and reducing the energy use consumption.
Acknowledgments This work is supported by National Science Foundation (NSF) Research Grant DMI-20062085 and Pan-Pacific MIRAI (Manufacturing Institute for Research on Advanced Initiatives). The authors would also like to thank Daimler AG for providing access to laboratory facilities and Mr. Dezsoe Schilling for his insights and contributions to this work. To learn more about the research activities of the LMAS please visit http://lmas.berkeley.edu. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.
References 1. Ávila, M., Gardner, J., Reich-Weiser, C., Vijayaraghavan, A., Tripathi, S., Dornfeld, D., 2005, Burr minimization strategies and cleanability in the aerospace and automotive industry, SAE Transactions, Paper 2005-01-3327 2. Tool-life testing with single-point turning tools, ISO 3685-1977(E) 3. Reich-Weiser, C., and Dornfeld, D., 2006, Experimental investigation of the influence of machining parameters on chip geometry for enhanced cleanability, Laboratory for Manufacturing and Sustainability, Consortium on Deburring and Edge Finishing 4. Viharos, Z. J., Markos, S., Szekeres, C., 2003, ANN-based chipform classification in turning, XVII IMEKO World Congress – Metrology in the 3rd millennium, Dubrovnik, Croatia, 1469–1473 5. Sturenberg, B., 2008, Systematic investigation of the chips caused by residual dirt behavior, Doctoral Thesis, Daimler AG, Stuttgart, Germany 6. Ávila, M., Dornfeld, D., Berger, K., 2006, Analysis of residual chip geometries during impulse machine cleaning tests, Summer Research, DaimlerChrysler, Stuttgart, Germany 7. Particle sizing and counting by microscopic analysis, ISO 16232-7, 2007
Burr Minimization and Removal by Micro Milling Strategies or Micro Peening Processes A. Kienzler, M. Deuchert, and V. Schulze
Abstract Micro milled mold inserts made from hardened and tempered steel can have burrs in the size order of 50 μm at the edges. These burrs often prevent the easy demolding of green bodies from micro powder injection molds. Further surface treatments on the molds are necessary to improve surface quality thus facilitating demolding processes. In the present study, three different processes: micro milling, abrasive micro peening and ultrasonic wet peening, have been investigated for their suitability as viable solutions for reducing or eliminating burrs. The tool used for the micro milling process is capable of removing existing burrs but creates new burrs which are inherent to the tool movement over the machined surface. While abrasive micro peening leads to a reduction of burrs on the mold surface, the material below the impacted surface is plastically deformed. Ultrasonic wet peening showed the best effectiveness at removal of burrs of a wide variety of complex geometries in a short processing time and without plastic deformation of the edge zones. Keywords Burr removal · Micro milling · Micro mold inserts · Micro peening
1 Introduction Micro powder injection molding (μPIM) [1] is a widely used technique for an economic large-scale production of micro parts. Micro milling is proved to be a very cost effective machining technology for the micro structuring of μPIM
A. Kienzler (), V. Schulze Universität Karlsruhe (TH), Institut für Werkstoffkunde I, 76131 Karlsruhe, Germany e-mail:
[email protected] url: www.iwkl.uni-karlsruhe.de
mold inserts made not only from unalloyed steels but also from tool steels with high hardness values (above 700 HV) [2–4]. Depending on the diameter of the milling tool fine structural details in the range of a few 10 μm can successfully be realized [5]. The surfaces of micro milled mold inserts usually show striations in their topography as well as burrs which can cause undesirable structural defects during demolding of the green bodies in μPIM processing [6]. K. Lee and D. A. Dornfeld [7] showed that with appropriate milling parameters it was possible to decrease the burr height but a prevention of burrs was not possible. For this reason a direct use of micro milled mold inserts is not possible thus requiring a finishing process prior to μPIM. Mechanical grinding or polishing, as it is usually applied to macro mold inserts, often lead to damaging of the fine details in the cavities. Deburring processes using laser technique were investigated by S. Hwan Lee and D. A. Dornfeld [8], but only burrs on the outer edges of the workpiece could be removed. Applying the laser removal method to micro mold inserts will result in beam damages on the surfaces adjacent to the burr region. These limitations have led to the development and characterization of alternative processes for the smooth finishing of micro mold inserts. In the Collaborative Research Center 499 “Development, production and quality assurance of molded micro-components made of metallic and ceramic materials” [9] wet and dry peening processes [10] have been selected as viable options for the smooth finishing of micro mold inserts. Peening processes are different from milling processes as they require no form tools that can only be used for one mold insert and also precision problems associated with micro-tool manufacturing are avoided. This paper examines and compares three different processes: micro milling, abrasive micro peening and ultrasonic wet peening, as methods for the minimization or removal of burrs at the edges of micro milled cavities.
M. Deuchert, V. Schulze Universität Karlsruhe (TH), Institut für Produktionstechnik, 76131 Karlsruhe, Germany
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_26, © Springer-Verlag Berlin Heidelberg 2010
237
238
2 Material, Procedure and Experimental Set-up 2.1 Material All tests were carried out with specimens made from the lowalloyed tool steel 30CrMo6 (Toolox 44, 0.3% C, 1.35% Cr, by SSAB Oxelösund, Sweden), which was hardened and tempered at 590◦ C to a hardness of 450 HV1 by the manufacturer. This steel shows a high purity and homogeneity, comparable to electro slag re-melted steels. The microstructure of 30CrMo6 is characterized by a matrix of tempered martensite with homogeneously distributed carbides.
2.2 Micro Milling Together with a cost effective advantage for the micro structuring of μPIM mold inserts, micro milling has a high material removal rate and yields to good surface quality [11]. Achievable surface roughness ranges between Rz = 0.1– 0.3 μm depending on the workpiece material. Cutting tools made from cemented carbide are frequently used for cutting steel materials. Such tools are commercially available in diameters down to 20 μm whereas in the field of research diameters down to 5 μm have been manufactured. Figure 1 shows examples of cemented carbide cutters with diameters of 20 and 7 μm. Typically, coated tools are used for the machining of steel materials. This can be problematic because the coating increases the cutting edge radius and thus leads to higher process forces and an increased ploughing effect [12]. Process related surface patterning is inherent to cutting processes and can be optically visible on cut surfaces. A detailed description of the experimental method is given in [2, 4]. The mechanical structuring of the steel was performed on a Kugler Micromaster 2 micro milling machine tool. It operates with hydrostatic axes, linear actuation in x and y direc-
Fig. 1 Cemented carbide tools, (a) diameter 20 μm (source: Hitachi Tools) and (b) diameter 7 μm. Source: Institut für Produktionstechnik, PMT.
A. Kienzler et al.
tion and a high-precision linear measuring system allowing for an axis accuracy better than 0.5 μm. Because of tolerance-relevant influences resulting from the spindle, the clamping system, the cutter geometry deviation and the cutter deflection, the accuracies of the finished part are within a range of ± 3 μm. Electrical continuity measurement, with an accuracy that is estimated to be less than 1 μm, was used to ensure proper contacting of the component and, thus, the determination of the height level. The tool used for steel machining was a coated ultra fine grain carbide cutter. This cutting material has grain sizes between 0.4 and 0.7 μm. Based on these tool characteristics, micro milled grooves in the steel workpieces were reworked in order to reduce burrs. The burr reduction proceeded by running the micro milling tool in a direction parallel to the micro milled groove at a height displacement of 5 μm below the upper surface and a lateral overlap of 50% to the groove. A second step followed by again running the micro milling tool through the groove.
2.3 Micro Peening A major attribute of an abrasive micro peening process is that the peening agent, usually shots or beads of micrometer size, is accelerated at high velocities towards a workpiece and upon impact material is removed from the surface of the target. Figure 2 shows a typical situation sketch of the nozzle compared to the workpiece. The removal rate depends on the shot velocity, the angle of impact, the shape, the size and the material of the shot. At high angles up to 90◦ the dominance of plastic deformation increases and according to this large edge rounding occurs. With low angles the gain from impact is minimized since the shot deflects from the surface. Therefore at impact angles between 40◦ and 75◦ an optimum for the removal rate for metallic workpieces is supposed. For deburring experiments the shot direction should be vertical to the burr longitudinal direction otherwise burr
Fig. 2 Orientation relationship between the nozzle and the workpiece. The nozzle direction of motion is along the protruding edge (A) in the figure; i.e. along an axis normal to the plane of present image view
Burr Minimization and Removal by Micro Milling Strategies or Micro Peening Processes
239
removal can not be achieved. The removal rate is higher for sharp-edged shots than for smooth, spherical shots. At constant velocity but an increasing shot size, the removal rate increases because of a higher kinetic energy. A blasting machine of the type IEPCO Peenmatic 770 equipped with an additional peening device Micropeen 200 was used for the abrasive micro peening of the machined specimens [10]. The shot used to modify the micro milled cavities and the burr at the edges was Al2 O3 with a size between 20 and 30 μm. For the peening a jet nozzle diameter of 0.8 mm was used at a pressure of 2 bar, nozzle to workpiece distance of 10 mm, nozzle to workpiece surface angle of 45◦ and a nozzle velocity of 1 mm/s. For an effective deburring result, the nozzle was made to advance along the edge vertical to the burr longitudinal direction. Fig. 3 Main processes in ultrasonic wet peening [14]
2.4 Ultrasonic Wet Peening Ultrasonic wet peening is based on cavitation of bubbles and the resulting micro jets of accelerated abrasive media. The specimen is placed in an aqueous suspension of an abrasive medium in which cavitation is generated by an ultrasonic wave. These abrasive particles in the fluid act as nuclei for the formation of cavitation bubbles. The bubbles rapidly become instable and collapse by forming a micro jet which accelerates the abrasive particles towards the workpiece surface (“indirect cavitation”). By the impact of these particles a material removal at the workpiece surface is made possible. Without abrasive particles in the fluid “direct cavitation” is the main process which results in deep round pits at the surface [13]. Figure 3 illustrates the main processes in ultrasonic wet peening [14]. For this process the ultrasonic horn of the processor type “Hielscher UIP-500” with a maximum acoustic power of 500 W and a frequency of 20 kHz was placed in a tank which was filled with an aqueous fluid suspension of an abrasive medium. The fluid consisted of distilled water with 5 wt.% Al2 O3 (particle diameter between 20 and 30 μm). A continuous exchange of the fluid between the ultrasonic horn and the workpiece (distance ∼ 1 mm) was enabled by a magnetic stirrer and an ultrasonic pulse duration of 0.8 s followed by a break with no ultrasound of 0.2 s. The workpiece to ultrasonic horn distance was set by using an adjustable table. Via a water filled cooling coil immersed in the tank, the temperature of the fluid was kept constant at 25◦ C. Figure 4 shows a schematic view of the experimental set-up. The height of the burr was measured every 1 min over a 10 min lapse and then every 3 min for the following 12–21 min.
Fig. 4 Schematic view of the experimental set-up for ultrasonic wet peening
2.5 Surface Characterization Methods A nondestructive analysis of the burrs before and after the finishing processes was performed with a confocal white light microscope of the type “Nanofocus μSurf”. It has to be mentioned here that with this optical method burrs can only be characterized by their width and the height because the incident and reflected light are vertical to the sample surface. Using a scanning electron microscope the true shape of the burr is visible; however an adequate quantitative characterization of the burr shape was only possible by preparation of metallographic cross sections which requires destroying the initial sample state.
3 Sample Preparation 3.1 Cavities An ultra fine grain carbide cutter with a diameter of about 500 μm was used to manufacture grooves of length 5 mm and depth 100 μm using micro end-milling technique. The
240
A. Kienzler et al.
Fig. 6 Cross sectional profile of a burr
Fig. 5 Detail of a groove with burrs at the edges
true width of the grooves was between 500 and 600 μm. At a cutter rotational speed of 30,000 rpm and a feed rate of 10 μm/tooth, it was possible to achieve a total depth of 100 μm by 10 in-feeds at 10 μm/in-feed. By this a distinctive burr was observed to form at the rear portion of the regions worked over by the cutter (see Fig. 5). Micro endmilling creates burrs which have different sizes at the upcut and downcut side of the tool motion. This size difference of the burr is a result of the alternate motion of the rotating tool.
3.2 Burrs Figure 6 shows a cross sectional profile of a burr with the parameters describing the burr. This burr was generated by micro milling where only one single groove with a width of 500 μm and a depth of 100 μm was machined. The burr has a height of about 45 μm, the width at the base is approximately 28 μm. The width is observed to be decreasing with increasing height. A radius of about 3 μm was determined for the coiling base. Beier [15] concluded from publications and own experiments within the last 10 years that burr height, width and radius are not significant values for the description of burrs and that only the width at the base is a reasonable value for burr formation. However, the shape of burrs plays an important role in burr removal processes. For example the specimen depicted in Fig. 6 exhibits a buckled burr requiring that the direction
of the shot accelerated towards the burr is considered during the peening process.
4 Results of Deburring 4.1 Micro Milling It can be seen from Fig. 7a, d that the micro milling process creates burrs at the edges of the grooves. After micro milling along both sides of the groove with a displacement of 5 μm to the upper surface the burrs at the edges are removed (see Fig. 7b). However this process is accompanied by the generation of new burrs which are observed to be oriented parallel to the worked surface with a bending over the cavity (see Fig. 7e). After the second micro milling step, these newly created burrs are removed however with a resulting formation of newer burrs (see Fig. 7c, f).
4.2 Abrasive Micro Peening Figure 8 shows the cross section of one micro milled groove before and after abrasive micro peening with the Al2 O3 shot. Prior to abrasive micro peening the burr height is about 35 μm with a width of 32 μm. After the finishing process the height is reduced to about 10 μm and the width is increased to about 80 μm. Material removal at the base of the burr is not observed. In Fig. 9a SEM view of the abrasively peened cavity is presented. It can be seen that the burr at the edge is not removed but rather is plastically deformed with a bending to the surface.
Burr Minimization and Removal by Micro Milling Strategies or Micro Peening Processes
241
Fig. 7 Detailed view of (a), (d) a micro milled groove, (b), (e) a micro milled groove after removal of burrs by the milling tool, (c), (f) a micro milled groove after the second burr removal step of the micro milling tool
Fig. 8 Cross section of a burr measured with a confocal white light microscope before and after abrasive micro peening with Al2 O3 shots
peening are indicated in Fig. 10. The burr at position 1 has an average height of 17.5 μm and an average width of 37 μm while for position 2 the burr average height of 21.1 μm and average width of 54 μm were determined. In Fig. 11 the burr heights during ultrasonic wet peening are shown. The height of the burr at position 1 is observed to be decreasing up to a machining time of 5 min. At 6 min a short increase is observed followed by a drastic drop in the burr height down to 1 μm. Further ultrasonic wet peening treatment on the specimen only leads to a constant material removal at the top surface. The burr height remains constant at a value of about 2.1 μm which is approximately the surface roughness (Rz).
Fig. 9 Detailed view of the abrasively peened cavity
4.3 Ultrasonic Wet Peening For ultrasonic wet peening two burrs with two different heights and widths at the edge of one micro milled groove were investigated. Figure 10 shows a two-dimensional view of a part of the groove. Two positions (Pos 1 and Pos 2) where the height of the burrs was analyzed during ultrasonic wet
Fig. 10 Two-dimensional view of the groove with the 2 positions of the measured cross section profiles
242
Fig. 11 Burr height vs. time during ultrasonic wet peening
The burr height at position 2 appears to be relatively constant at 17 μm up to 12 min of ultrasonic wet peening and then suddenly decreases down to about 3 μm at 15 min. After 15 min the burr is removed and a residual burr with a height of about 2.8 μm remains even during the further peening time of 5 min. This height value corresponds to the surface roughness (Rz) making it difficult to distinguish between surface and burr. In Fig. 12 the surface quality of a micro milled groove before and after 21 min of ultrasonic wet peening is presented. It can be seen that the burrs created after micro milling have a complex shape. Upon ultrasonic wet peening via Al2 O3 particles burrs vanish with no visible indication of plastic deformation at the edge zones. The ultrasonic wet peening process is noted to also modify the topography of the specimen surface. Striations at the bottom of the cavity vanish after 21 min.
4.4 Discussion The use of the same micro milling tool for burr removal after machining of cavities results in burr reduction or removal however the onset of a cutting process in materials will always create new burrs. Burr sizes at the edge
A. Kienzler et al.
zones are influenced by the type and the ductility of the used material [6]. Abrasive micro peening is potentially suitable for burr removal although experiments have shown that at a peening angle of 45◦ to the cavity surface normal peening distorts the burr profile towards the surface rather than removal of the burr. The burr is only plastically deformed and not removed. Due to the high hardness of the shot (Al2 O3 , > 700 HV) and the relatively ductile material, at a peening pressure of 2 bar the edge rounding effect is severe. To overcome this effect, a steel type of higher hardness will be suitable. An experiment described in [16] using a less ductile tool steel of hardness about 550 HV revealed that the burrs were completely removed and that the radius at the edge was below 5 μm. The result reported in [16] also showed that together with burr removal, striations at the base of the cavity vanish creating a new surface topography of better quality. The impacting shots also produce large compressive residual stresses at the near surface layers which can be in the order of –800 MPa [17]. Ultrasonic wet peening is capable of removing burrs within a few minutes irrespective of the shape of the burr. This efficiency is facilitated by the fact that particles in the fluid are accelerated towards the workpiece surface from all directions. A proposed scenario is that the height of the burr is reduced in a constant manner by the impinging particles followed by a complete collapse after several minutes.
5 Conclusion and Outlook In this study three different processes for burr removal have been investigated. The application of the micro milling tool is a viable machining option of minimizing burrs but nevertheless small burrs always remain and have to be removed in a second finishing step. Abrasive micro peening could be adapted for tool steels with high hardness to avoid large edge rounding. Ultrasonic wet peening showed the best results concerning burr removal. The prospect of applying a combination of these processes may open-up a new possibility of achieving a higher surface quality and therefore improving the molding process in micro powder injection molding and the quality of the molded micro components. Acknowledgements The Deutsche Forschungsgemeinschaft is gratefully acknowledged for financially supporting the investigations performed within the framework of the Collaborative Research Center 499 “Development, production and quality assurance of molded microcomponents made of metallic and ceramic materials”.
References Fig. 12 Detailed SEM view of a micro milled groove (a) before and (b) after 21 min ultrasonic wet peening
1. Piotter, V., Benzler, T., Gietzelt, T., Ruprecht, R., Haußelt, J., 2000, Micro powder injection molding, Advanced Engineering Materials, 2(10): 639–642.
Burr Minimization and Removal by Micro Milling Strategies or Micro Peening Processes 2. Fleischer, J., Schmidt, J., Haupt, S., Halvadjiysky, G., Kotschenreuther, J., 2005, Mikroformeinsätze in gehärtetem Stahl: Ein Vergleich der Verfahren Mikrofräsen, Mikrofunkenerosion und Mikrolaserablation, wt werkstattstechnik online, 95: 887–891. 3. Denkena, B., Hoffmeister, H.-W., Reichstein, M., Illenseer, S., Hlavac, M., 2006, Micro-machining processes for microsystem technology, Microsystem Technology, 12:659–664. 4. Schmidt, J., Kotschenreuther, J., 2005, Micro end milling in hardened steel, In: Löhe, D., Haußelt, J. (eds), Advanced Micro and Nanosystems Vol. III: Micro-Engineering in Metals and Ceramics, Wiley-VCH, Weinheim, 107–130. 5. Fleischer, J., Deuchert, M., Ruhs, C., Kühlewein, C., Halvadjiysky, G., Schmidt, C., 2008, Design and manufacturing of micro milling tools, Microsystem Technologies, 14:1771–1775. 6. Horsch, C., Schulze, V., Löhe, D., 2005, Material states and surface conditioning for mold inserts, In: Löhe, D., Haußelt, J. (eds), Advanced Micro and Nanosystems Vol. III: Micro-Engineering in Metals and Ceramics, Wiley-VCH, Weinheim, 221–249. 7. Lee, K., Dornfeld, D. A., 2005, Micro-burr formation and minimization through process control, Precision Engineering, 29: 246–252. 8. Hwan Lee, S., Dornfeld, D. A., 2001, Precision laser deburring, Journal of Manufacturing Science and Engineering, 123:601–608. 9. Kraft, O., Haußelt, J., Ruprecht, R., Emmerich, B., 2007, Prozessketten des Urformens für metallische und keramische Mikrobauteile – der Sonderforschungsbereich 499, In: Kraft, O., Emmerich, B. (eds), Kolloquium Mikroproduktion, ISBN 978-3923704-61-3, 7–11.
243
10. Horsch, C., Schulze, V., Löhe, D., 2006, Deburring and surface conditioning of micro milled structures by micro peening and ultrasonic wet peening, Microsystem Technologies, 12: 691–696. 11. Fleischer, J., Löhe, D., Kotschenreuther, J., Schulze, V., Deuchert, M., Halvadjiysky, G., Haupt, S., Kienzler, A., 2007, Fertigungsverfahren in der Mikrotechnik, wt werkstattstechnik online, 97(11/12): 847–851. 12. Albrecht, P., 1960, New developments in the theory of the metalcutting process: Part I – the ploughing process in metal cutting, Transactions of the ASME, Journal of Engineering for Industry, 82:348–358. 13. Kienzler, A., Schulze, V., Löhe, D., 2008, Surface conditioning by ultrasonic wet peening, Proceedings of ICSP10th, 15–18 September, Tokyo, Japan:205–210. 14. Ichida, Y., Sato, R., Morimoto, Y., Kobayashi, K., 2004, Material removal mechanisms in non-contact ultrasonic abrasive machining, Wear, 258:107–114. 15. Beier, H.-M., 2001, Gratentstehung – ein umformtechnischer Ansatz, wt werkstattstechnik online, 91:12:765–772. 16. Kienzler, A., Horsch, C., Schulze, V., Löhe, D., 2007, Application of micro peening for deburring and surface conditioning of micro milled moulds, Proceedings of 7th Euspen International Conference, Vol. II, 20–24 May, Bremen, Germany:372–375. 17. Kienzler, A., Okolo, B., Schulze, V., Wanner, A., Löhe, D., 2008, A reliable tool for the improvement of micro powder injection moulds made of steel, Advanced Engineering Materials, 10(7):661–665.
Assessment of Deburring Costs in Industrial Case Studies P.J. Arrazola
Abstract Owing to the fact that deburring is considered a non-added value operation, little attention is paid to estimating its real cost. However, deburring expenses in manufacturing processes can be significant depending on the workpiece or sector considered. They can vary from 2 to 3% for mass production of relatively common parts in automotive sector to 9–10% in the case of aeronautics for small batches series or mass production of complicated parts. This figure can reach even higher values in other sectors such as medicine. Seven case studies from different manufacturing sectors (aeronautics, automotive, heavy transport, oil piping transport. . .) will be outlined in this paper in an attempt to assess real deburring costs and other issues dealing with the reasons why deburring is carried out and the methods used for burr removal. Keywords Burr removal · Deburring · Cost
1 Introduction Deburring is the removal of edges and burrs from manufactured parts. Depending on the burr size, workpiece dimensions, location and series of parts to be manufactured, different approaches are proposed to carry out the deburring operation. For instance, manually by sand papering, hand scratching, hand filing, using motorized tools with mounted stones. . . or automatically by countersinking, thermal energy, high pressure water jet, electrochemical machining. . . [1–4]. Due to the general view that deburring does not have a significant influence on the total part manufacturing cost breakdown, information about its real portion is difficult to
obtain. It is claimed, in the case of the manufacturing of automotive components, that the total cost share is 14% while in aeronautics this figure can reach the value of 30%. Deburring represents a significant fraction of machining costs for manufacturers causing several industrial problems: • During inspection and assembly (interferences, jamming and misalignment of parts). • Safety hazard to personnel due to the burr sharpness. • Damage of components while they are working. • Reduction of fatigue life of components. The increase in the quality of machined workpieces in addition to reducing the cost per piece, are two of the major concerns in today’ manufacturing industry which aims maintaining the current competitiveness. Thus, deburring has become one of the issues that will be necessary to tackle over the next years. In an attempt to assess a better approximation of real deburring costs seven industrial case studies, belonging to sectors like automotive, heavy transport, oil pipe lines and aeronautics will be presented in order to give a broader view on the following: 1. 2. 3. 4.
When and how burrs are produced Reasons why deburring is needed Methodology employed for deburring Estimation of total costs and investments.
For confidential reasons, some information is not presented in the paper.
2 Case Study 1: Disc Brakes P.J. Arrazola () Faculty of Engineering, Manufacturing Department, Mondragon University, 20500 Mondragon, Spain e-mail:
[email protected] url: www.eps.mondragon.edu/investigacion/procmeca
Disk brakes made of cast iron GG-25, are manufactured in quantities higher than 10 million a year (see Fig. 1a). Burrs are produced in several machining operations when the tool is exiting: turning, drilling . . . (see Fig. 1b). Dur-
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, DOI 10.1007/978-3-642-00568-8_27, © Springer-Verlag Berlin Heidelberg 2010
245
246
P.J. Arrazola
a)
b)
a)
c)
d)
c)
b)
d)
Fig. 2 (a) Steering Knuckle; (b) Example of burr produced in milling; (c) Countersinking tool; (d) Holes after countersinking
e) Fig. 1 (a) Disk brake; (b) Example of burr in drilling; (c) Example of burr that could be accepted by a car manufacturer; (d) Countersinking tools; (e) Holes’ edges after countersinking
ing drilling operation, and depending on the working conditions (tool geometry and tool sharpness, cutting speed, feed rate, . . .) a burr or a material breakage is produced at the hole exit edge (average of 6 holes per part). Depending on the car manufacturer, a chamfer or simply an edge with no burrs may be required at the holes’ exit edges. Thus in many cases, an additional countersinking operation should be planned in the process (see Fig. 1c, d, e). This means an additional working post whose total investment is close to C30,000–40,000 per production line. Furthermore, end drill life is defined by the size of the material breakage or burr produced at the edge of the exit hole. The total cost of this additional operation is close to 2–3% compared with the total part cost. In the case of turning, part design plays a key role in avoiding material breakage due to the material brittleness, above all in areas close to the edges where tool exiting can create this type of problem.
countersinking and milling operations (see Fig. 2b). As in the previous case study, the total cost of this additional operation is close to 3% compared with the total part cost. However, costly investments are required (robots, shared with other manipulation operations) in order to maintain demanded part costs (see Figs 2c, d). The investment for this additional drilling post is approximately C50,000 per cell unit for the manufacturing of 250,000 steering knuckles a year over 5 years. As in the previous case study, and only in some cases, burrs or material breakage has been avoided by solutions implemented in the part design.
4 Case Study 3: Flange for Oil Pipes Deburring is carried out mainly to avoid the risk of injuries to workers that will manipulate the workpiece afterwards. Workpiece material is carbon steel A-105 and A-350 LF2. Burrs are produced after the drilling (4–24 holes per part) (see Fig. 3a, b). As more than 1.5 million flanges are produced a year, automatic deburring by countersinking is planned in the production line using rather common cutting tools (see Fig. 3c). The deburring costs are estimated to be approximately 2–3% of the total part cost. The total investment costs to carry out the deburring operation, is 9% of the production line (estimated at C100,000).
3 Case Study 2: Steering Knuckles 5 Case Study 4: Gear Steering knuckles made of GGG-40 (see Fig. 2a), are manufactured in series of more than 3 million a year. Burrs appear in the milling surfaces (4 per part) and drilling operations (8 per part). This requires additional operations of
Deburring is carried out due to two main reasons: functional and safety (risk of injuries). Workpiece material is hardened steel 15CrNi6.
Assessment of Deburring Costs in Industrial Case Studies
247
The steps for the deburring operation are as follows:
a)
b)
1. Eliminate most of the burrs’ material by hand filing (see Fig. 4c). 2. Finishing of the gear teeth exit with hand motorized tool with mounted stones in order to round the edges (see Fig. 4d).
6 Case Study 5: Flange
c)
d)
Fig. 3 (a) Flange; (b) Example of burr produced in drilling; (c) Countersinking tool; (d) Hole after countersinking
Burrs are produced after the gear shaping operation (see Fig. 4a, b). Due to manufacturing in small batches (approximately 100 parts a year), manual deburring is the method used for this case study. Skilled workers carry out the deburring operation in the assembly areas, so as the part does not become scratched. The deburring operation takes approximately 10 min per part and the cost is estimated to be roughly 2–3% of the total part cost. No special investments are needed.
a)
b)
c)
d)
Deburring is carried out owing to two reasons: functional and safety hazard (risk of injuries). Workpiece material is 25CrMo4 (see Fig. 5a). Due to the complicated shape and manufacturing in small batches (approximately 100 parts a year), manual deburring (hand filing, sand papering. . .) is the basic method used for this case study. As in the previous case, skilled workers carry out the deburring operation in the assembly area, so as the part does not become damaged. Burrs are produced after the drilling operation (see Fig. 5b). The deburring operation takes approximately 45 min and the cost is estimated to be roughly 7% of the total part cost. No special investments are needed. The steps for the deburring operation are as follows: 1. Removal of most of the burrs from the holes using a hand motorized tool with mounted stone (see Fig. 5c). 2. Eliminate remaining burrs by manual scraping (see Fig. 5d). 3. Sandpapering of the hole’ edges to obtain a honed edges (see Fig. 5e). 4. Hand filing (see Fig. 5f) and sand papering (see Fig. 5g) of the milled slot. 5. Eliminate burrs of the small orifices manually by countersinking with a drill (see Fig. 5h). 6. Finishing of the whole part with Scotch-BriteTM .
7 Case Study 6: Disc
e) Fig. 4 (a) Gear; (b) Example of burr produced in gear shaping; (c) Hand filing; (d) Edge finishing using a motorized tool with mounted stone; (e) Gear after deburring operation
The reason for which deburring is carried out is due to the risk of the detachment of burrs which can cause serious problems when the workpiece is up and running. Workpiece material is GG-25 (see Fig. 6a). Burrs are produced in the drilling operation in holes that are intersecting (see Fig. 6a). Due to the complicated shape and manufacturing in small batches (roughly 300 parts a year), manual deburring is used. The operation is carried out in the assembly area by skilled
248
P.J. Arrazola
a)
b)
a)
b)
c)
d)
c)
d)
Fig. 6 (a) Example of burr produced in the intersection of two holes; (b) Observing the burr with an endoscope; (c) Hand shaving with the help of the endoscope; (d) Air pressure
8 Case Study 7: Compressor Casing e)
g)
f)
h)
i) Fig. 5 (a) Flange; (b) Example of burr produced in drilling; (c) Edge finishing using a motorized tool with mounted stone; (d) Manual scraping; (e) Sandpapering. (f) Hand filing; (g) Sandpapering; (h) Countersinking by drilling
workers that will employ hand scraping tools and sand paper to remove, with the help of an endoscope, the large burrs (see Fig. 6a, b, c). High pressure air is utilized to extract the loosened burr (Fig. 6d). The deburring operation takes approximately between 30– 45 min and the cost is estimated to be roughly 7% of the total part cost. No special investments are needed.
The workpiece presented in this case study is an aeroengine compressor casing made of stainless steel (Jethete) (see Fig. 7). Burrs in aeronautical parts represent a fairly essential aspect because they can affect its life. Due to the complicated shape and manufacturing in small batches (roughly 60–70 parts a year), manual deburring is the method used for this case study. Burrs are produced in drilling and milling operations. Deburring operation is carried out in the assembly area by skilled workers in order not to scratch the part, using different methods depending on the burr dimensions, norm to be accomplished. . .: sand papering, hand filing, ScotchBrite (TM) , hand-stones, motorized tools (pneumatic or electrical) with mounted stones, cuts, abrasive filament brushes R or even Flex-Hone tools. Other deburring methods that are employed depending on the workpiece considered are: deburring carried out by robots or centrifugal deburring. . .. The deburring operation takes approximately 40 hours per part. Costs can represent 9–10% of the total part manufacturing cost breakdown. No special investments are needed.
9 Conclusions Table 1 shows the summary of the case studies presented in this paper. The following points are the main conclusions of the case studies presented in this work:
Assessment of Deburring Costs in Industrial Case Studies
249 16 18 17
26
25
14
1
20 2 19 22 3 21
15
27
24 4
28
29
23
29
a)
47
46 51
44 48 45
b)
50
47
49
Fig. 7 (a) Features where burrs appear in the LPT case; (b) Features where burrs appear in the compressor casing of low pressure turbines
1. The main reasons for deburring are: a) Functional issues: risk of detachment when the workpiece is up and running and part fatigue life. b) Risk of injuries when manipulating the part 2. Burrs do not seem to be a major issue when considering the whole part manufacturing process; it is assumed that they will always appear during the machining operation. Nevertheless, deburring has a significant influence, depending on the case study on costs, investment, production time, part quality. . . 3. In high series production, burr size is one of the end tool life criteria.
4. In the case of brittle materials, material breakage can occur instead of a burr. 5. Removal of burrs can be made by several methods and using different means: automatically by countersinking or manually by hand filing, sand papering, hand scraping, hand-stones, motorized tools with mounted stones. . . 6. Deburring costs can vary from 2–3% to 9–10% for workpieces employed in sectors such as automotive, heavy transport, oil pipe transport or aeronautics. Nevertheless, and after some more in depth surveys, it can be estimated that these values can be even higher. In fact, there is a clear unawareness of the deburring real costs until the question is raised. . .
<30s 3% of the total part cost
C 50,000
<30s
2–3% of the total part cost
C30,000–40,000
Countersinking
Countersinking milling
Time dedicated to deburring Cost/part (compared with total cost including material, labour cost, writing off the investment. . .) Total investment needed for the deburring Other issues to be taken into acount
>3 Million a year Risk of injuries functional
>10 million a year Risk of injuries functional
Serie Reason why deburing is carried out Solution adopted for deburring
Stering Knuckles GGG-40 Milling–Drilling
Automotive
Disc brakes GG-25 drilling
Study case
Table 1 Summary of case studies Industrial sector Automotive
C 9,000
2% of the total part cost
<30s
Countersinking
>1.5 Million a year Risk of injuries
Carbon steel A-350 LF2 drilling
Oil pipe lines
None
Skilled personnel required; high risk of scrapping
Skilled personnel required; high risk of scrapping
7% of the total part cost
45 min
Skilled personnel required; high risk of scrapping
None
7% of the total part cost
30–45 min
Hand scraping with the help of an endoscope
≈300 a year Risk of injuries functional
≈100 a year Risk of injuries functional Hand filling, Hand scraping, Hand filling, Sand papering, Scotch–Brite TM
Disc(GG-25);
Heavy transport
Flange (hardened steel 15CrNi6);
Heavy transport
None
2–3% of the total part cost
10 min
Hand filling, Hand motorized tool with mounted stone
Gear cutting; hardening steel 15CrNi6 gear hobbing ≈100 a year Risk of injuries functional
Heavy transport
Aeronautics
Skilled personnel required; high risk of scrapping
None
9–10% of the total part cost
≈60–70 a year Functional (compulsory): part life Hand filling, hand–stones, motorized tools (pneumatic or electrical) with mounted stones, cuts, abrasive filament brushes or even R Flex–Hone tools. 40 hours
Compressor body of stainless steel (Jethethe)
250 P.J. Arrazola
Assessment of Deburring Costs in Industrial Case Studies
7. Non negligible investments are needed in some cases where the parts are manufactured in high series. 8. For small series and especially when burr removal is relatively complex, the operation remains still manual, which means the need of skilled workers, risk of unrepeatability, difficulties to estimating real costs. . . Acknowledgments We would like to thank the Basque and Spanish Governments for the financial support given to the projects: MARGUNE (code IE03-107), TAF (IE05-148) and SIMMECAN (CIT020000-2008-20) and to the companies Fagor Ederlan S.Coop, Ulma Forja, Sapa Placencia and ITP for the technical support given to this work.
251
References 1. Gillespie, L., 2003, Hand Deburring, ISBN: 0-87263-642-9, SME 2. Gillespie, L., 1999, Deburring and Edge Finishing Handbook, Society of Manufacturing Engineers. ISBN: 0872635015, SME. 3. LMA Research reports 1999–2000. 4. Tolinski, M., Deburring Processes and Challenges: Issues Intermesh When Choosing the Right High-Volume Deburring Method. Manufacturing Engineering. SME- October 2006, Vol. 137, No. 4.
Author Index
A Arrazola, P.J., 245 Aurich, J.C., 63, 99, 129, 167 B Berger, K., 229 Biermann, D., 13 Boud, F., 189 Brinksmeier, E., 31 D de Leon, L., 55, 147 Denkena, B., 55, 147 Deuchert, M., 237 Dix, M., 117 Dornfeld, D., 3, 229 E Elbing, F., 181 Engin, S., 213 Engmann, J., 129 Essig, C., 107
K Kang, J., 205 Kannan, S., 189 Karpuschewski, B., 197 Kästner, J., 55 Kienzler, A., 237 Klocke, F., 107 Ko, S.L., 157 Kretzschmar, M., 73, 181 L Leitz, L., 99, 167 Leopold, J., 47, 79 Liao, L., 213 Lung, D., 107 M Martinsen, K., 21 Marx, T., 129 Matsumura, T., 47 Melkote, S.N., 89 Mihotovic, V., 73, 181 Min, S., 3 Morehouse, J.B., 89
F Fangmann, S., 31 Folkes, J., 189 Franke, V., 99, 167
N Neugebauer, R., 117 Newton, T.R., 89
G Garg, S., 229
P Petzel, M., 197
H Haberland, R., 129 Heilmann, M., 13 Heisel, U., 139 Hellstern, C., 89 Hoang, H.P., 157
R Reichenbach, I.G., 63 Ringen, G., 21 S Schaal, M., 139 Schmidt, G., 117
J.C. Aurich, D. Dornfeld (eds.), Burrs – Analysis, Control and Removal, c Springer-Verlag Berlin Heidelberg 2010 DOI 10.1007/978-3-642-00568-8,
253
254
Author Index
Schnabl, M., 221 Schueler, G.M., 129 Schulze, V., 237 Sudermann, H., 63 Szulczynski, H., 73
W Wang, B., 147 Wohlgemuth, R., 79 Wolf, G., 139 Wright, I.W., 189
T Tantra, N., 189 Turner, S., 89
X Xi, F., 213
U Uhlmann, E., 73, 181
Y Yamaguchi, H., 205