ISSN 1018-5593 * -* * * * European Commission technical steel research Properties and in-service performance Assessment and analysis of current cyclic property data for steels STEEL RESEARCH European Commission technical steel research Properties and in-service performance Assessment and analysis of current cyclic property data for steels W. Cook British Steel Swinden Technology Centre Moorgate Rotherham S60 3AR United Kingdom Contract No 7210-KA/820 1 July 1991 to 30 June 1994 Final report Directorate-General Science, Research and Development 1998 EUR 17856 EN LEGAL NOTICE Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use which might be made of the following information. A great deal of additional information on the European Union is available on the Internet. It can be accessed through the Europa server (http://europa.eu.int). Cataloguing data can be found at the end of this publication. Luxembourg: Office for Official Publications of the European Communities, 1998 ISBN 92-828-1693-1 © European Communities, 1998 Reproduction is authorised provided the source is acknowledged. Printed in Luxembourg PRINTED ON WHITE CHLORINE-FREE PAPER PAGE CONTENTS 7 1. INTRODUCTION 2. 8 DATABASE 8 2.1 Database Design Considerations 9 2.2 Assessment of Available Data EXPERIMENTAL WORK TO ASSESS THE REPRODUCIBILITY 13 3. OF STRAIN CONTROLLED LOW CYCLE FATIGUE TESTING 3.1 Material for Investigation 13 3.2 Evaluation Procedure 14 3.3 Experimental Results 14 DISCUSSION 18 4.1 Database 19 4.2 Reproducibility of LCF Parameters 20 4.3 Test Direction 21 4.4t Piece Condition 21 4.5 Strain Measurement/Control 21 22 CONCLUSIONS REFERENCES 23 TABLES 25 FIGURES 31 APPENDIX 1 DESCRIPTION OF DATABASE 47 X 2 RELATIONSHIP BETWEEN LCF FATIGUE 59 EXPONENTS Β AND C APPENDIX 3 TEST PIECES DIMENSIONS USED FOR 61 LCF TESTING LIST OF TABLES 1. Laboratories Supplying Low Cycle Fatigue Data, or Contacted for Discussion on Design of the Database 2. Variation in LCF Parameters for a 0.4% C 1.0% CrMo Engineering Steel 3. Longitudinal Monotonic Property Data (42CrMo4) from Heat Treated Plate 4. Transverse Monotonic Property Data (42CrMo4) from Heat Treated Plate 5. Variation in Life from Repeat Testing at Constant Strain Amplitudes (Longitudinal Samples) 6. Effect of Orientation on LCF Properties (a) Transverse Samples (b) Longitudinals 7. Low Cycle Fatigue Results for Ground and Polished and Ground Only Longitudinal Specimens 8. Effect of Including True Strain at Fracture in Calculation on LCF Parameters 9. Variability in LCF Parameters Obtained by Selecting Extremes from Total Longitudinal Sample Population 10. Typical Variability in LCF Parameters as a Function of Number of Tests 11. Fatigue Parameters Obtained from LCF Results and Estimated from Monotonic Tests (Transverse Samples) LIST OF APPENDICES 1. Description of Database 2. Relationship Between LCF Fatigue Exponents b and c 3. Test Pieces Dimensions used for LCF Testing LIST OF FIGURES 1. Difference in Fatigue Coefficients Resulting from Expression of Parameter in Terms of Reversals or Cycles (Illustrated using Fatigue Ductility Coefficients) 2. Strain Distribution Across a Strip (From Ref. 5) 3. Dimensions of LCF Test Piece used for Work on 4 2C r Mo4 4. Variation in Mechanical Properties Across Heat Treated Plate 5. Distribution of Fatigue Lives after Testing at Constant Strain Amplitude (Reversals to Failure) 6. Effect of Test Piece Hardness on Fatigue Life 7. Comparison of Transverse and Longitudinal LCF Properties 42CrMo4 8. Effect of Orientation on Strain/Life Curves for 42 C r Mo4 9.t of Surface Finish on LCF Properties (Ground ν Ground and Polished) 10. Fractographic Examination of Sample Showing Fracture Path Associated with Knife Edge 11. Strain Life Curves for 42CrMo4 (Longitudinal) Determined using all Test Results 12. Variability in LEF Parameters Obtained by taking Selected Results (see Table 9) ASSESSMENT AND ANALYSIS OF CURRENT CYCLIC PROPERTY DATA FOR STEELS British Steel pic ECSC Agreement No. 7210.KA/820 TONAL TECHNICAL REPORT 1. INTRODUCTION In contrast to the constant stress loading traditionally used for the assessment of fatigue properties, steel components, in service, are subjected to a wide and variable range of stress and strain. For many , including those in automotive applications, fatigue damage can accumulate from stresses giving rise to localised plastic strains, for example around local discontinuities, which can significantly reduce component life. Methods of assessing the range and frequency of strains imposed on components are under continuous development. At the same time methods of predicting component life, from a knowledge of these data, are also under development and are being linked into the design process. To enable effective use of these tools property data are required which assess the performance of materials under cyclic conditions imposing small and measured plastic strain cycles. This is obtained by carrying out strain controlled fatigue tests rather than the older stress controlled tests. Methods for determining the relevant fatigue parameters have been under development for some years and they are being refined with experience. Together these methods have the potential for reducing the time involved in the design and validation of components and the overall development costs. Consequently, the interest in cyclic properties is no longer 'of academic interest only'. Laboratories throughout the world are investigating the strain controlled properties of steels (and other metals), to establish the properties attributable to materials in a variety of product forms and conditions. However, these tests have been carried out on a largely uncoordinated basis often for specific purposes and customers. These data could be of considerable value in promoting the use of steels. Consequently, the initial aim of this work was to collate and assess the information presently available in European laboratories, via the production of a database. However, whilst a database has been developed, and data available from some European laboratories have been added, at this stage it is probably of limited value. The main reason for this is the small number of tests carried out where results could be made available. European laboratories approached for information included those involved in ECSC contract work and some others which had carried out LCF testing. Information was supplied only by those laboratories involved in ECSC contracts. Some of the other laboratories had carried out LCF testing either for 'in-house' purpose or on a contract basis, but the results were considered private and could not be made available. However, the indication was that the amount of this 'private' information would have made relatively little difference to the number of datasets available. As the amount of available data was much smaller than anticipated, the project was extended to obtain other information relevant to LCF testing. For example,whilst collating results for the database it was noted that, where similar steels were compared, there would be large differences in the values of the fatigue parameters, whilst all the data appeared to relate to longitudinal properties. The significance of these differences was unknown; therefore work was carried out to provide a first evaluation of the repeatability of LCF testing and the effect of test direction. The work was extended to obtain an initial assessment of surface finish. From these results and assessment of some pointers from the literature, comment has been made on areas requiring further consideration to facilitate future LCF testing. 2. DATABASE The initial aim of this project was to compile a computer database containing the cyclic and LCF (low cycle fatigue) properties of steels. At the commencement of the project it was envisaged that the data for including in the database would be obtained principally by review of the work undertaken at the various laboratories carrying out work funded by the ECSC and these organisations were contacted with a view to participation. The laboratories were requested either to provide data and/or offer opinion on the database format. Details of thes contacted for views on the format and depth of the database are shown in Table 1. Where appropriate these organisations also supplied much of the data included in the database. Approaches were also made to a limited number of other laboratories thought to be concerned with LCF testing. However, the laboratories carrying out relevant testing did not wish to contribute data, considering this to be of a private or proprietary nature. Property data were provided by the participating laboratories in the form of reports, from which the relevant information was extracted. In total approximately 40 datasets were extracted for adding to the database. In view of the limited data available, further results were extracted from literature, in an attempt to widen the scope of the information available, although the amount of additional information was small. 2.1 Database Design Considerations Following discussions with personnel at various European laboratories, it was agreed that ideally the database should be computerised and suitable for use by: mechanical/design engineers, who may require output of property data for particular steels and product types. materials development engineers/metallurgists, who may be interested in the effect of processing variables giving rise to the particular properties. Consequently, in addition to containing property data, generated from monotonie and cyclic testing, it was agreed that the files should contain information on the significant processing parameters controlling the properties. Furthermore, it was agreed that the database should be confined to basic ambient temperature material properties and should not attempt to include, for example, properties of weldments, which would further increase the complexity. To incorporate processing data, the initial concept was simply to produce a list of parameters encompassing the range of likely processing histories from steelmaking to the final test piece. However, this would have required the introduction of an extensive range of parameters, of which many would have been redundant for any particular product form. For example, the properties of heat treated bar are dominated by the heat treatment and prior processing is largely irrelevant. Consequently, a relational design was adopted and this is described in more detail in Appendix 1. In summary, 13 separate datafiles have been created, to accommodate a description of the steel, processing and properties. The fields contained in these files are shown in Appendix 1, Figs. Al.1(a) to Al.l(m), and prompt the scope of inputs considered desirable. The size of each entry is depicted by X for a character field and O for a numeric field. Thus, for example, the field Test Lab', Fig. Al. 1(a), is a character field of up to 40 characters. The field C (carbon) Fig. Al. 1(b), is a numeric field of up to 6 digits long including the decimal point. Field shown as 'memo', allow the user to input free text of unrestricted length and occur mainly in the screens relating to processing information. Memo fields have been used because, from examination of the information received, it was clear that processing data could cover a wide spectrum of variability and complexity, whilst comprehensive information was rarely available. The use of a memo field allows the user to describe relevant processing information, when, and this methodology reduces the number of prompts required in each screen. It is therefore at the discretion of the user to decide on the input required.