Dipl-Ing Martin Große-Hovest works in the centre for Service Load Simulation & Evaluation at the Fraunhofer Institute for Structural Durability & System Reliability in Darmstadt, where Dr-Ing Gerhard Fischer was Head of the Structures and Components Business Unit. Dr-Ing Bernd Velten works on development and technical consulting for Otto Fuchs KG in Meinerzhagen, and Ing Steven Cervello is Head of Product Development for Lucchini Sidermecchanica SpA of Lovere
AS SPEEDS AND axleloads continue to increase, wear and safety are becoming ever more critical to today's railway operators. New manufacturing techniques and the use of lighter materials offer the potential to reduce significantly the costs of infrastructure maintenance and energy consumption. At the same time, improving bogie dynamics and wheelset design will improve riding comfort.
The Hiperwheel project (HIgh PERformance WHEELset), has brought together researchers, manufacturers and operators to develop an innovative wheelset. As it meets the objectives of the ERRAC Strategic Rail Research Agenda 2020, the work received €3·7m in funding from the EU's Fifth Framework research budget.
The project required the collection of data on design loadings and a fundamental re-examination of the assessment of contact fatigue in wheelsets and axles under operational conditions.
Experimental and theoretical stress analysis led to the development of a demonstration wheelset which is approximately 25% lighter than a conventional steel wheelset and offers a reduction in radiated noise of 5 to 6 dB(A).
Issues and objectives
The use of aluminium alloys for wheel discs and new high-strength steels for axles is a challenging opportunity. Within the scope of the project, the interdisciplinary research included:
- load data acquisition on various tracks regarding noise, riding behaviour, comfort and structural durability;
- design of prototype lightweight steel and hybrid (steel-aluminium) wheelsets using modern experimental and numerical methods;
- characterisation of the main damage mechanisms such as contact fatigue in press-fits and wheel-rail contact;
- verification of the theoretical performance using manufactured prototype wheelsets.
In particular, comprehensive tests were undertaken to validate the structural durability of the wheelset. In addition the research results could be used to support a recommendation on how to overcome the partial deficiencies of the existing European wheel standards.
Because of the wide range of tasks, the Hiperwheel consortium included representatives from the manufacturing industry and research fields as well as railway operators. Table I shows how the various work packages were shared between the different partners.
The hybrid wheelset
Produced by Lucchini and Otto Fuchs, the hybrid wheelset consists of a high-strength steel axle and forged aluminium wheel discs (left). With a total weight of 782 kg the wheelset is approximately 25% lighter than a conventional steel wheelset which weighs 1045 kg. The main technical data for this wheelset are shown in Table II.
The Fraunhofer Institute (LBF) undertook comprehensive measurements in its unique wheelset test rig, using a range of different load assumptions in order to guarantee structural durability despite the known defects in the current norm.
Extensive investigations were carried out with the manufacturers on improving the press-fit connection between wheel and axle2-6. Fig 1 shows the test results for a press-fit using an axle made from 30NiCrMoV12 and an aluminium wheel hub. Woehler tests were done with a constant amplitude and Gassner tests with a variable amplitude, to determine the permissible total damage from fretting fatigue in order to estimate the potential service life under varying operational conditions.
From these investigations it was found that:
- fretting fatigue leads to expectation of different damage behaviour with variable amplitude compared with constant amplitude3,5;
- the aluminium-steel press-fit has a similar effect on reducing fatigue strength to that of a steel-steel connection;
- the increase in fatigue strength obtained by using the higher strength material 30NiCrMoV12 is much smaller in press-fits than is suggested by the relationships of the static fatigue strength values2,8.
Fretting fatigue results in a large scatter. Under constant amplitude tests, there is no fatigue strength endurance limit, contrary to the design recommendations in the various European norms, UIC515-5, EN 13103, EN 13104, and EN 13261, which are based on an 'infinite life' design. To avoid these and other significant shortcomings of the norms, some partners put forward proposals for practice-oriented design8.
The fracture surfaces produced in the Gassner tests (Fig 2b) are comparable with many genuine fractures in service7. The illustrated fracture surface shows the typical beach marks found in operating conditions, which are not similar to the fracture surface produced with constant amplitude (Fig 2a). With the help of further analysis, it is possible to draw conclusions from these fracture surfaces regarding crack growth and to define axle inspection intervals.
The results obtained with component-like material samples were checked using real wheelsets with optimised parameters on the test rig. The Woehler and Gassner curves mostly determined under rotating bending stress must be corrected for use with driven axles, because with combined torsional and bending stress in the press-fits, the effect of fretting fatigue will be greater.
Following the decision to design the prototype wheelsets for an Italian TAF double-deck EMU, maximum design forces for various load cases were derived from the measurement data1,2 and the data of the individual spectra (cumulative frequency distribution). This enabled the development of both design spectra and test spectra for the trials (Fig 3).
For reliable validation of the structural durability of multi-piece wheels, the most important requirement was to ensure a realistic deformation of the axle in the test facility similar to the deformation experienced in operation. It was also necessary to simulate all load cases with realistic correlation of the individual forces. These requirements are not even approximately fulfilled with the simplified testing methods employed for the solid steel wheels involved up to now, and in fact many parts of the wheelset are inaccessible for testing.
On the LBF wheelset test rig it is possible to simulate decisive load cases such as straight runs at up to 350 km/h with high centrifugal forces, curving and passing through switches and crossings4,6. This permits quick and realistic approval for new design variants, materials and production methods.
For time and cost reasons three tests are usually undertaken for experimental validation2,9. Preliminary finite-element modelling was carried out both for the stress analysis and fatigue life evaluation of the aluminium wheel disc, and also in order to detect highly-stressed areas for further experimental stress analyses.
Computer-aided fatigue assessment using the LBF wheel strength software allows the production of pre-optimised prototypes of rail wheels, which in turn enables the number of tests to be minimised. The assessment estimates the crack initiation on the inner wheel disc region on node 484 (Fig 3). The calculated required fatigue strength value (RFS) in this area is 56MPa.
Detailed experimental stress analysis using strain gauges was carried out on the wheel disc on the prototype wheel. The first conclusion to be drawn from this was that the RFS for the highly-stressed aluminium disc area had a value of 57MPa, which coincided well with the value calculated using FEM. This value falls within the existing range of permissible stresses9, but whether or not this is actually achievable must finally be determined by independent tests.
New materials and load assumptions
Yield strength values for medium-strength aluminium alloys meet those of typical wheel disc steels like C22, C35 and 46MnSi4. But if other specific requirements of railway applications, such as a long-life component durability of up to 30 years, are taken into account, additional properties like fatigue behaviour, corrosion resistance and notch sensitivity become relevant.
Considering this combination of requirements as a whole, the alloy EN-AW 6082 T6 was selected as offering the most favourable combination of properties. It is a medium-strength alloy with the best corrosion resistance of all age-hardening alloys. It can be formed by rolling, extruding and forging at high temperatures.
Table III shows the specifications for the TAF EMU wheelsets. The design loads shown refer to the 'classic' durability assessment as defined in UICfiche 510, which were used for basic calculations of the wheel design.
The specific strength (YS/density) of EN AW 6082 is comparable to those of typical wheel disc steel alloys. On the other hand, the wear properties of aluminium and other light alloys are poor, which requires the use of a steel rim to avoid early failure in the wheel-rail contact zone.
This led to the design of a two-part wheel assembled from an aluminium core and a steel rim. With press-fit connections between axle and hub and between disc and tyre, reliable transmission of traction and braking forces will require a good grip in both areas.
The wheel is designed so that old tyres can be cut up for removal from the disc, whilst the dismounting of the disc from the axle is supported by oil pressure applied through a borehole in the hub.
For the fundamental calculations, the overall geometry of the wheel disc was chosen as flat and symmetrical. The symmetry of the tyre seat and the flat web geometry avoid radial bending stresses as a result of the vertical load. This offers considerably lower stress amplitudes compared to a curved geometry.
An additional advantage of the flat web geometry is that only compression stresses, which are known to have a positive effect on the fatigue behaviour, result from the shrink fits. The criterion for the dimension of the axle-disc press-fit is the reliable transmission of moments under all service conditions. As a conservative limit the braking moment for blocked wheels was selected.
Two full-scale validation tests were carried out in the test rig during the research project. Different variations of tyre and axle press-fits were manufactured to simulate the worst case scenarios, with different temperatures resulting from service conditions producing various pre-stresses in the wheel disc.
During one test, a fatigue crack occurred on the highly stressed disc area - as predicted - after the test requirement had been exceeded (Fig 4). Another wheel, optimised in the wheelseat region, achieved a test life of 70000 km without any visible cracks on the disc and axle and without significant wear in either the axle or tyre seats. This corresponds to a service life of more than 10 million km.
Additional validation tests will still be required to verify these single test results. These will need to investigate possible influences resulting from repeated assembly of the wheel disc to the axle and of the tyre to the wheel, the effects of worn tyres, temperatures on the wheel rim/disc interface during braking, and the effect of impact damage on the lightweight wheel disc.
From practical experience with wheelsets that break in service, the conclusion can be drawn that an infinite life is not possible. If newly-developed lightweight designs are to be used widely, new validation procedures will be needed to guarantee safety standards, incorporating better knowledge of operational loads and their spectra.
The major limitations of the existing European specifications in terms of the load and stress levels for modern rolling stock could be avoided by applying this new method of validation testing.
Nevertheless, the Hiperwheel project has established the feasibility of a new wheelset design, using aluminium wheel discs and high-strength steel axle. This offers the potential for a significant reduction in weight and noise compared to a standard steel wheelset.
Table I. Partners in the Hiperwheel project
|Work package||Responsible partner|
|1||Service measurements||Fraunhofer Institute for Structural Durability & System Reliability (LBF), Germany|
|2||Dynamic modelling||Politecnico di Milano, Italy|
|3||Damage mechanisms and design||University of Sheffield, UK|
|4||CAE-based procedure for durability||Fiat Research Centre (CRF), Italy|
|5||Numerical procedure for||Chalmers University of Technology, Sweden|
|6||Development of demonstrators||Lucchini Sidermeccanica SpA, Italy|
|7||Manufacturing of prototypes||Valdunes SAS, France|
|8||Full scale testing on demonstrators||Fraunhofer Institute, Germany|
|9||Management dissemination||Fiat Research Centre (CRF), Italy|
|Wheel production||Otto Fuchs Metallwerke KG, Germany|
Lucchini Sidermeccanica SpA, Italy
|Other partners||TrenItalia, Mechanical Dynamics, Italy|
Table II. Key specifications for hybrid wheelsets
|Wheel discs||Aluminium alloy AlMgSi F31 (EN-AW 6082T6)|
|Axle||Steel 30NiCrMo V12|
|Static axleload, tonnes||16|
|Design life||3 million km (wheel disc)|
15 million km (axle)
|Service temperature range íC||-40 to +70|
Table III. Specifications and load assumptions for TAF EMU wheelsets
|Wheel diameter (new/worn) mm||920/860|
|Axle diameter mm||190|
|Rim profile||UICORE-fiche 510-2|
|Vertical load to be used in design kN||125|
|Lateral load to be used in design kN||60|
|Maximum speed km/h||160|
|Service temperature range íC||-40 to +70|
Picture caption: Two full-scale validation tests were carried out using the Fraunhofer Institute's wheelset test rig. The tyres and axle press-fits were specially manufactured to simulate the worst case pre-stresses which would occur in service
Picture caption: Fig 1. Results of fatigue tests on a press-fit between a steel axle and an aluminium wheel disc
Ts: fatigue strength scatter
TN: fatigue life scatter for probability of survival (Ps) = 90% and 10%
D: axle diameter at wheel seat
d: diameter of axles after fillet radius
h: slope of curve
Picture caption: Fig 2. Variable amplitude tests of 25CrMo4 steel samples produced fracture surfaces exhibiting 'beach marks' (b, right), similar to fractures which have occurred on wheelsets in service. Constant amplitude fracture surfaces are visibly different (a, left), and analysis enables the estimation of crack growth rates and defintion of inspection intervals
Picture caption: Fig 3. Calculation of the required fatigue strength value (RFS) on the hybrid steel-aluminium wheelset
Picture caption: Fig 4. As expected, fatigue cracking occurred on the highly stressed disc once the test wheelsets were subjected to loads above those which would be encountered in reality
1. Hiperwheel - development of an innovative high performance railway wheelset. European Community research project GRD1-1999-11028, final report, 2005
2. Fischer G and Grubisic V. Dimensioning of wheelset axles - influencing parameters and procedure for the structural durability validation. Report No. FB-226-e, Fraunhofer Institute for Structural Durability and System Reliability, Darmstadt, Germany, 2005
3. Waterhouse R B. Fretting Corrosion. Pergamon Press, 1972
4. Fischer G, Grubisic V, and Widmayer H. Fatigue Tests on Wheelsets under Simulated Service Stress Spectra. 9th International Wheelset Congress, Montréal, 1988
5. Fischer G, Grubisic V and Buxbaum O. The influence of fretting corrosion on fatigue strength of nodular cast iron and steel under constant amplitude and load spectrum tests. Standardisation of Fretting Fatigue Test Methods & Equipments, ASTM Special Technical Publication No 1159. 1992
6. Fischer G and Grubisic V. Service-like durability approval of wheelsets. 12th International Wheelset Congress, Qingdao, 1998
7. Fischer G, and Grubisic V. Versagen von Radsatzwellen und dessen Ursache. ZEVrail - Glasers Annalen, March 2006
8. Grubisic V and Fischer G. Betriebsfeste Bemessung von Radsatzwellen. Eisenbahntechnische Rundschau 55, 2006
9. Sonsino C M, Berg-Pollack A and Grubisic V. Structural durability proof of automotive aluminum safety components - present state-of-the-art. SAE Technical Paper Series 2005-01-0800