Fatigue analysis of two wheel-‐mounted brake disc designs PDF
Preview Fatigue analysis of two wheel-‐mounted brake disc designs
Fatigue analysis of two wheel-‐mounted brake disc designs Master degree project DURANTON Coralie Stockholm 2015 Division of Rail Vehicles School of Engineering Sciences Acknowledgements This master thesis was carried out at the Rolling Stock Engineering Centre (CIM) of the French National Railway Company (SNCF), in Le Mans, France, for the Division of Rail Vehicle at the Royal Institute of Technology (KTH) in Stockholm, Sweden. The project was conducted under the supervision of Sébastien SMITH and Pascal TOMASIN, to whom I want to thank for the appreciated support and useful advice. I am really grateful to have been able to exchange with them and with all the engineers of the ‘Bogie’ department. I also want to show my gratitude to my KTH supervisor and examiner, Sebastian STICHEL and Mats BERG, for their help and guidance, throughout this master degree project and also during the whole master’s program. Abstract Due to a need of more compact bogies, the brake discs can be mounted on the railway wheels, bolted through the wheel web. Thus, the wheels are drilled and have multiple areas of contact with the brake discs. To establish maintenance procedures that will be applied to the wheels, SNCF used the feedback from experience (as with the train AGC) which gives perfect performance in terms of safety. However, to optimize the maintenance process, numerical simulations may be preferred since they are less conservative. This report describes the numerical simulations, based on the finite element method, that were conducted to determine if the Régiolis wheel complies with the standard EN 13979-‐1 from a mechanical fatigue point of view. In addition, it provides additional insights regarding the loads and damage suffered by the wheel, which are not taken into account in the standard: the damage induced by disc braking and the fretting that may occur at the contact interfaces. This study has been used as a decision support for the first inspection intervals of the Régiolis wheels. Keywords: railway wheel, wheel-‐mounted brake disc, fatigue, fretting, Finite Element Method, Finite Element Analysis, ANSYS, OptiStruct, nCode. 1 Contents ACKNOWLEDGEMENTS ............................................................................................................................ 1 ABSTRACT ................................................................................................................................................ 1 CONTENTS ............................................................................................................................................... 2 CHAPTER 1 INTRODUCTION .............................................................................................................. 4 1.1 CONTEXT ............................................................................................................................................ 4 1.1.1 More compact bogie .............................................................................................................. 4 1.1.2 Rolling stocks at stake ........................................................................................................... 4 1.1.3 Examples of wheel mounted brake discs in the world ........................................................... 5 1.2 STAKES ............................................................................................................................................... 6 1.2.1 Economic stakes ..................................................................................................................... 6 1.2.2 Technical stakes ..................................................................................................................... 6 1.3 LIMITATIONS ....................................................................................................................................... 7 1.4 GOALS ............................................................................................................................................... 7 1.5 METHODOLOGY ................................................................................................................................... 7 CHAPTER 2 LITERATURE AND STANDARD REVIEW ............................................................................. 9 2.1 THE WHEEL-‐MOUNTED BRAKE DISCS ......................................................................................................... 9 2.1.1 General presentation of railway wheels ................................................................................ 9 2.1.2 General presentation of wheel-‐mounted brake discs .......................................................... 11 2.1.3 Characteristics of the wheel-‐disc assemblies ....................................................................... 12 2.2 FATIGUE APPLIED TO WHEELS ................................................................................................................ 16 2.2.1 Fatigue in the standards relative to railway wheels ............................................................ 16 2.2.2 Fretting-‐fatigue .................................................................................................................... 17 2.3 THE FINITE ELEMENT ANALYSIS .............................................................................................................. 18 2.4 PREVIOUS WORK ON THIS SUBJECT ......................................................................................................... 19 2.4.1 Work about the AGC ............................................................................................................ 19 2.4.2 Work about the Régiolis ...................................................................................................... 19 CHAPTER 3 GENERAL PROCEDURE ................................................................................................... 21 3.1 INTRODUCTION .................................................................................................................................. 21 3.2 COMPUTER AIDED DESIGN OF THE COMPONENTS ..................................................................................... 22 3.3 PREPROCESSING ................................................................................................................................. 23 3.3.1 Preprocessor used ................................................................................................................ 23 3.3.2 Meshing: creation of the nodes and elements ..................................................................... 25 3.3.3 Assigning materials and properties ..................................................................................... 27 3.3.4 Creating contact between surfaces ..................................................................................... 27 3.3.5 Application of the boundary conditions ............................................................................... 29 3.3.6 Application of the loading, according to EN 13979-‐1 .......................................................... 31 3.4 SOLVING ........................................................................................................................................... 33 3.4.1 Convergence criteria ............................................................................................................ 33 3.4.2 Contact algorithms .............................................................................................................. 34 3.4.3 Confidence in the results ...................................................................................................... 34 3.5 FATIGUE POSTPROCESSING ................................................................................................................... 35 3.5.1 Criterion and methodology for the fatigue analysis ............................................................ 35 CHAPTER 4 ANALYSIS AND COMPARISON OF THE FATIGUE RESULTS FOR THE RIGID AND THE FLEXIBLE WHEEL-‐DISC INTERFACES ............................................................................................................... 37 4.1 PRELIMINARY CALCULATIONS OF NON-‐EQUIPPED WHEELS ........................................................................... 37 4.2 RESULTS IN THE WEB-‐HUB CONNECTION .................................................................................................. 37 4.3 RESULTS AROUND THE HOLES ................................................................................................................ 38 4.4 RESULTS IN THE CONTACT AREAS ........................................................................................................... 40 2 4.5 FATIGUE POSTPROCESSING ................................................................................................................... 43 4.6 CONCLUSION ..................................................................................................................................... 46 CHAPTER 5 DISCUSSIONS AND SUGGESTIONS FOR FURTHER WORK ................................................ 47 5.1 INFLUENCE OF DISC BRAKING ................................................................................................................ 47 5.1.1 Influence of thermal aspects ................................................................................................ 47 5.1.2 Influence of mechanical aspects .......................................................................................... 49 5.2 OPTIMIZATION OF THE MESH SIZE TO INCREASE THE ACCURACY IN CONTACT AREAS .......................................... 50 CHAPTER 6 CONCLUSION ................................................................................................................ 51 CHAPTER 7 REFERENCES .................................................................................................................. 52 3 Chapter 1 Introduction 1.1 Context 1.1.1 More compact bogie Wheel-‐mounted brake discs are at stake in this master’s thesis. Before presenting the characteristics of the conception of this wheel-‐discs assembly, the aim of this chapter is to deal with the backgrounds of this new design: the emergence of a low floor train. The low floor technology has led to change the design of the systems underneath the carbody. In order to improve the passengers’ accessibility, the entrance floor height must be at the same level as the platform height, which means there is only a horizontal gap between the train and the platform. Thus, the floor is lower than the traditional trains’ floor, where there is often a footboard between the train and the platform. However, beneath the carbody and the passengers’ pathways, there are sometimes axle-‐mounted brake discs and other traction equipment. Since there is less room, the bogies have to be more compact and for example, the brake discs of axle-‐mounted brake discs can be shifted on the wheels. In order to connect the brake discs and the wheel, wheels have been drilled and bolts assemble the wheel and the discs. Drills within the web of the wheel induce concentrated stresses and moreover the static and dynamic loads between the wheel and the discs have given rise to fatigue issues. 1.1.2 Rolling stocks at stake The AGC (Autorail à Grande Capacité, literally high-‐capacity multiple unit) delivered by Bombardier in 2004 and the Régiolis delivered by Alstom in 2014 is the main rolling stocks in focus in this study. Both of them are regional trains and have a maximal speed of approximately 160 km/h. Figure 1-1 - AGC - Source : Bombardier Transport Figure 1-2 - Régiolis - Source : Alstom Transport, A. Fevrier Their wheels are produced by the manufacturers Bonatrans, Valdunes and CAF. The wheel-‐mounted brake discs have been designed and produced by Knorr (previously Freinrail) for the AGC and by Faiveley for the Régiolis. 4 Figure 1-3 – Example of a trailer bogie of AGC - Figure 1-4 - Example of a trailer bogie of Régiolis - Source : Bombardier Transportation Source : Alstom Transport Other rolling stock operated by SNCF have wheel-‐mounted brake discs, for example the NAT (Nouvelle Automotrice Transilien, literally new multiple unit for the Transilien, i.e. the suburban zone around Paris) and the Regio2N (regional train with double deck) of Bombardier. Concerning the design of the brake discs, the AGC and the NAT are similar since they both have ‘rigid’ wheel-‐disc assemblies; and the Régiolis and Regio2N are also similar due to their ‘flexible’ wheel-‐disc assemblies. Thus, there are two distinct designs and more details will be given about them in Section 2.1.3 -‐ Characteristics of the wheel-‐disc assemblies. As written previously, the AGC was delivered 10 years ago and the feedback regarding the ‘rigid’ interface is quite positive. The innovative design of the ‘flexible’ discs used on the Régiolis and the Regio2N has led to a new verification and validation campaign made by SNCF. The work done for this master thesis’ project is included in this validation campaign. 1.1.3 Examples of wheel mounted brake discs in the world In France, there are few wheel-‐mounted brake discs but their number is rising. In Europe, more examples of wheel-‐mounted brake discs can be found, such as the wheels of the AnsaldoBreda E403 locomotive and the wheels of the EMU Regina. Figure 1-5 – Regina bogie The X60 of Alstom, a commuter train also known as the ‘Pendeltåg’ in Stockholm, Sweden, has wheel-‐mounted brake discs with the same design as the Régiolis and the Regio2N. It has been delivered since 2006. 5 1.2 Stakes 1.2.1 Economic stakes In order to prevent a fatigue crack from reaching a critical length, the inspection intervals for preventive maintenance of the AGC wheels are extremely short during their first time of operation. Each 50,000 km, the wheelsets are taken apart and the brake discs are disassembled from the wheels in order to check the appearance of cracks. As it is usually the case within SNCF, the maintenance intervals have been defined based on feedback and learning from experience. This method gives a complete satisfaction regarding the safety issues but it is expensive since the wheelsets have to be removed from the operational fleet, taken apart and assessed. For the latest rolling stock Régiolis, the stakes are not to have that short inspection intervals as it was the case with the AGC. For safety purposes, SNCF must be sure that the wheels are secure regarding fatigue issues and for economic reasons the inspection intervals have to be optimized. This has led SNCF to use numerical methods, in addition to the feedback, in order to achieve their goals. 1.2.2 Technical stakes This project was conducted within the CIM (Centre d’Ingénierie du Matériel, literally rolling stock engineering centre) with belongs to SNCF. Through this project, the goals were to acquire and develop methods that involve the management of contact elements with the Finite Element Analysis (FEA) software products they own, namely ANSYS and OptiStruct, and to enhance their fatigue calculation methods. It is necessary to better know the contact elements since they are used to represent the contacts between the wheels and the brake discs. Until now, the running gear are calculated apart (the wheels on one side, the brake discs on the other side), which does not fully represent the actual behaviour of the wheel-‐discs assemblies. The contact elements of the numerical model are quite delicate to define but they also are decisive to reproduce the actual contacts. The mastery of contacts with the FEA software should help to better understand the relative motion and the forces between the wheel and the discs, and notably to assess the presence of fretting fatigue. Concerning fatigue, the standards EN 13979-‐1+A2 and UIC 510-‐5 require that the wheels should have an infinite life. This requirement is based on the application of conventional loads, so, exceptional loads are not taken into account in the fatigue assessment. However, in its transportation service, a wheel will endure exceptional loads that will create damage. The damage is cumulative which means, one refers to the fact that, every load (the highest as the lowest) creates additional damage and can lead to fatigue crack initiation. Assessing the cumulative damage using post-‐processing software would also be beneficial. To put it in a nutshell, this project aims are bringing decision support for the maintenance that will be applied for the Régiolis. The following aspects will be taken into account: the mechanical behaviour of the wheel-‐discs assembly, the characteristics of the contacts and the fatigue stress state of the wheel under conventional loads. 6 1.3 Limitations The numerical simulations only use structural analysis; this means that thermal loads are not taken into account. The analysis is static or quasi-‐static, so that neither transient effects nor vibration effects are considered. Furthermore, the results of the calculations will assess crack initiation but not crack propagation. Besides, the wheels are subjected to the conventional load cases described in the standard EN 13979-‐1+A2 in the chapter ‘Assessment of the mechanical behaviour’. These conventional load cases represent three common situations: running on a straight track (the wheelset is centred), running on a curve track (the outside surface of the flange is pressed against the rail) and running through switches and crossings (the inner surface of the flange is pressed against the rail). The loads from shoe braking or disc braking are not considered. In addition, the residual stresses from the manufacturing process are not calculated preliminary (there is no residual stress at the initial time) and the same material properties are used for the whole wheel. Finally, the expected methods should be feasible in an industrial setting, that is to say that, the calculations times should be reasonable and that numerical models should be easy to be taken over in the future. 1.4 Goals The main purpose is to get information about the behaviour of the wheel and the brake discs at their interface, for the AGC and for the Régiolis assemblies. For wheels, the stress state at each node of the structure is needed, since a post-‐processing in fatigue is to be conducted in order to see if the wheels fulfil the requirements of the standards EN 13979-‐1 and UIC 510-‐5. As written in the standard: “The purpose of this assessment is to ensure that there will be no risk of fatigue cracking either in the wheel web or in its connections with the hub or the rim during the service life of the wheel“. This constitutes the decision support about the first inspection intervals for the Régiolis’ maintenance. Furthermore, simulation results have to be reliable, robust (if some parameters are changed, the models remain stable and the calculations converge) and fast, while having non-‐linear finite element models. Thus, it will improve the CIM’s methods within structural analysis. 1.5 Methodology This project has been conducted by means of finite element analysis. As I was a beginner, I first trained myself with simple models (for example, the CAD (computer-‐aided design) geometries were rougher, some components were ignored). Once this was done, I made my more complex models, to fit the actual wheel-‐discs assemblies. The simple models enabled me to learn from my errors and thus, not to do it again in the more complex models. 7 While I was building the numerical models, I read the mechanical surveys made by SNCF, Bombardier, Alstom and Faiveley about this subject, so I could compare our hypotheses and our results. Besides, I looked for information about the thermal behaviour of the assemblies in order to enhance the understanding of the wheel-‐discs assemblies’ behaviours. 8 Chapter 2 Literature and standard review 2.1 The wheel-‐mounted brake discs AGC and Régiolis are made of several coaches (from 3 to 6) that are in connection with the track through the bogies. The bogies bring the running gear, the wheelsets and the suspension together. One frame assembles these components. The wheelset is made of an axle on which wheels, brake discs and mechanical transmission components are mounted. Brake discs can also be mounted on wheels. The pictures below illustrate the difference between a bogie with axle-‐mounted brake discs and one with wheel-‐mounted brake discs: Figure 2-1 - Trailer bogie with axle-mounted brake Figure 2-2 - Trailer bogie with wheel-mounted discs - TGV brake discs - AGC In the Finite Elements simulations, the wheel, the brake discs and the bolts are fully modelled. Only one part of the axle is modelled, i.e. the part which is in contact with the wheel hub. In the coming paragraphs, the systems are presented in detail and the specific technical vocabulary is illustrated. 2.1.1 General presentation of railway wheels A railway wheel is a circular component, which rolls and steers the train on the track. There are different kinds of wheels (solid or drilled, monobloc or with encircling tires, cast or forged, etc.) but they are always recognizable. Indeed, they are usually located below the carbody (space for the passengers), generally joined on the same wheel axle and sometimes held by a bogie frame. Several parts of the wheel are distinguished, as it can be seen in Figure 2-‐3. 9
Description:The list of books you might like
