Project A1 - Appraisal of Track/Sub-Base Design using Modern Geotechnical Principles

INVESTIGATORS: Professor William Powrie (University of Southampton); Professor Chris Baker (University of Birmingham)
RESEARCHER(S): Dr Jeffrey Priest, Dr Li-Ang Yang, Dr Daren Bowness (University of Southampton); Dr Michael Burrow (University of Birmingham)
RESEARCH STUDENTS: Louis Le Pen (University of Southampton); Symeon Konstantelias (University of Birmingham)

INDUSTRIAL COLLABORATORS:

• Balfour Beatty
• British Geological Survey
• Great Central Railway
• Network Rail (Channel Tunnel Rail Link) Ltd
• Scott Wilson Pavement Engineering
• SPOORNET

Objectives: The objectives of the research carried out in Project A1 were to:

• Assess the robustness of traditional track foundation design using a more fundamental modern soil mechanics approach
• Explore the potential of a more fundamental approach to guide design decisions
• Explore the potential of a more fundamental understanding to assess performance, maintenance and remediation

Background: The major costs of track maintenance are linked to the deformation and failure of the formation, but our current understanding of trackbed and subsoil performance, particularly in the context of a high speed and heavily trafficked railways, is poor. Our current state of knowledge does not allow reliable estimates of the life of the formation. Therefore, the development of a proper scientific understanding of the dynamic load deformation response and track/sub-base interactions is required for the economic design of new and replacement track systems, the development of remediation and maintenance strategies, the assessment of track system performance and the optimisation of whole life and whole system costs.

Brief summary of research methods: The development of a more scientific understanding into the performance of the subgrade below a railway line requires a detailed knowledge of the behaviour of the ground when subjected to the dynamic loading induced by a train passage. Research has therefore focused on:

• Appraisal of current methods of track system design
• Development and verification of suitable instrumentation to measure track and subbase deflections by means of field trials (on railways carrying trains of different speeds and axle loadings, including the CTRL and a South African heavy haul line), laboratory calibration and validation tests
• Field measurements of the track system and foundation response to moving train loading
• Development of numerical analyses to estimate stress paths followed by representative soil elements in the sub-ballast and formation
• Advanced soil element testing (using torsional hollow cylinder/cyclic triaxial) to investigate the effect of these stress paths on soil behaviour

Additional research strands have included:

• Use of finite element models to replicate non destructive field measuring techniques and train induced in situ loading conditions
• Assessment of subgrade stiffness before and after remedial work using site investigation techniques

Summary of Outcomes: The outcomes of the individual research may be summarised as follows:

• Appraisal of current methods of track system design. A parametric study was carried out to determine the thickness of ballast required for a given set of variables such as subgrade condition, axle load, speed and cumulative tonnage using a variety of international design codes currently used as a basis for the design of both new railway track and for the renewal of existing track. The methods considered were those that are being, or have been, used in the United States (the American Railway Engineering Association (AREA) method, a method proposed by Raymond (1978) and one suggested by Li and Selig (1998a and 1998b)); the United Kingdom (the former British Rail method and the current Network Rail design standard) and in Europe (the UIC 719R Code). The results showed that for all but one design method the thickness of ballast varied considerably and was not related to any of the factors examined. This highlighted the empirical nature of current design codes and the shortfalls with all codes in designing railways which will be subject to increasing axle loads and faster line speeds. Areas where further research is required were identified, including traffic characterisation, accurate determination of insitu material properties and material behaviour under realistic loading.
• Development and verification of suitable instrumentation by means of field trials. Two independent and innovative techniques for measuring track displacements have been developed. One system combines remote video monitoring with particle image velocimetry (PIV), using a webcam and a small telescope. The second uses sleeper mounted geophones that give a voltage output proportional to the velocity of motion, which can be filtered and integrated to calculate displacements. Laboratory validation tests show that the video monitoring system can measure peak-to-peak displacements to within 0.04 mm from a distance of 15 m for frequencies less than 4 Hz. The geophones measure peak-to-peak displacements to within 0.07 mm for frequencies as low as 1 Hz. Field measurements were conducted on a variety of lines within the UK (Crewe, NR; Ashford, CTRL; Loughborough, GCR). It was found that processing of the geophone data was inaccurate for train velocities which gave a dominant deflection induced by the train bogies less than 1Hz. The accuracy of the remote video monitoring system was limited by the frame rate of the camera to around 90km/h. It is anticipated that cameras with a quicker frame rate may overcome this limitation. In addition, field trials were conducted on the GCR, using a low energy laser source and meters attached to the sleepers. Initial results suggest that this may also offer an effective method of measuring absolute deflection of the track system, although its accuracy has yet to be defined.
• Field measurements of the real loads applied by trains and the track system and foundation response.From the parametric study on current design guides, it was concluded that to improve current design methods it is necessary to include more realistic loading within a model to determine stresses within the subgrade and that the stress path applied during soil element testing should be representative of that applied during a train passage. To that end a site investigation was conducted to provide data to validate numerical models being developed to ascertain the change in stresses within an element subjected to a moving load. This involved using instrumentation developed to measure track displacements (arrays of geophones and accelerometers) to measure ground velocities and accelerations within the ground at different depths, and at different positions across the track, for a railway line subject to heavy axle loads. Deflections were calculated from both the velocity and acceleration response of the ground. The results show how vertical displacement reduces rapidly with depth and lateral distance from the track. Analysis of the power spectrum obtained from a geophone attached to a sleeper showed a dominant frequency of ~1Hz, relating to a coupled pair of bogies from adjacent wagons. Additional frequencies at ~2 Hz and ~ 6 Hz related to individual bogies and individual axles respectively. This response was attenuated with depth such that at ~800mm below the bottom of the ballast, the individual axles were not discernible and by ~1800mm below the base of ballast only the coupled bogie loading was evident in the displacement history. Results from horizontal movement measurements showed a similar behaviour although the magnitude of the axle loading event on the sleeper was greater than in the case of the vertical response.
• Development of numerical analyses to estimate stress paths followed by representative soil elements in the sub-ballast and formation. As highlighted in the review of design codes the design of railway track formations has traditionally been empirically rather than analytically based, with ballast and sub-ballast layer thicknesses specified mainly on the basis of previous practice. Although recent methods are more scientific they still rely on input parameters= based on resilient modulus obtained from laboratory testing using cyclic triaxial test apparatus, which do not recreate the loading conditions experienced by a soil element in the field. Two- and three-dimensional numerical analyses have been carried out to investigate the stresses experienced by soil elements in the natural ground below a railway track during train passage. The effects of the depth of the element, the initial in situ stress state of the soil and the elastic parameters used to characterise the soil have been investigated. Results show that both the magnitude and the number of loading cycles during train passage reduce with depth, as measured during the field trials. Also, during loading the soil element is subjected to a rotation of the principle stress. This has enabled the stress paths to be modelled in a cyclic hollow cylinder apparatus, which is able to better represent the stress changes experienced in the field than a conventional triaxial test.
• Advanced soil element testing (using torsional hollow cylinder/cyclic triaxial) to investigate impact of principal stress rotation, load frequency, etc. The laboratory testing aims to simulate the stress conditions imposed in the field, and to investigate the behaviour of a soil specimen subject to PSR. A stress path for an element below the centreline of the track was used in Hollow Cyliner Apparatus (HCA) tests on reconstituted kaolin clay specimens. HCA tests with and without PSR have shown that PSR can result in larger permanent plastic strains and deformations, which is consistent with the findings of earlier research. Also, a smaller resilient modulus was obtained from the HCA test with PSR compared to the test without.
• Use of finite element models to replicate non destructive field measuring techniques and train induced in situ loading conditions. 3D dynamic numerical models have beeen developed to represent train induced loading and Falling Weight Deflectometer (FWD) dynamic loading. These numerical models have been used to better understand track performance and help assess the accuracy of design methods. The FWD model has been used to back analyse the resilient moduli and layer thicknesses of a number of existing railway structures including poorly performing sites, those performing well and those made from concrete slab track.
• Assessment of subgrade stiffness subject to remedial work using site investigation techniques. Methods of subgrade stiffness assessment for old railway track before and after remedial work using site investigation techniques have been developed. Future Work:The work initiated within project A1 are continuing within RRUK2 project A4, or as stand alone projects by groups associated with Project A1. The techniques developed for measuring dynamic displacements of the track system are currently being used to assess ballast behaviour on sections of the WCML associtaed with tilting Pendolino trains. In addition, testing is planned to commence work on CTRL1 to investigate aspects of track behaviour on this high speed line. A new high speed digital camera has been purchased and software written to enable high speed acquisition of sleeper displacements for analyses.Finite element models that considered static loading on track behaviour are being enhanced to account for the dynamic motion of the wheel. The results of these models will be validated against filed data and also used to determine actual stress paths for use in the HCA apparatus. A particular focus of Project A1 is the interaction between vehicle dynamic behaviour and track response, which is being investigated in a collaboration between University of Southampton and Manchester Metropolitan University.

OUTPUTS

Project Reports:

• Bowness, D., Priest, J.A., Powrie, W. (2005). Dynamic Sleeper Displacement Measurements at the Alsops Road Tunnelling Site. RRUK Report A1/03 (Prepared for CTRL)
• Bowness, D., Powrie, W., Richards, D.J., Lock, A.C., Clayton, C.R.I. (2004). Report on field observations of the performance of railway track – location, Crewe. RRUK Report A1/02 (prepared for Balfour Beatty Rail)
• Powrie, W., Bowness, D. (2003). Appraisal of track/sub-base performance using modern geotechnical principles. RRUK Report A1/01 (prepared for RSSB)

Publications:

• Priest, J.A., Le Pen, L., Powrie, W., Mak, P., Burstow, M. (forthcoming 2008) Performance of Canted Ballasted track during curving of high speed trains.
• Bowness, D., Lock, A.C., Powrie, W., Priest, J.A., Richards, D.J. (2007). Monitoring the dynamic displacements of railway track. Proc. IMechE Vol. 221-1 Part F: J. of Rail & Rapid Transit. pp. 13-22.
• Burrow, M.P.N., Bowness D., Ghataora, G.S. (2007). A comparison of railway track foundation design methods. Proc. IMechE Vol. 221-1 Part F: J. of Rail & Rapid Transit. pp. 1-12.
• Powrie, W., Yang, L.A., Clayton, C.R.I. (2007). Stress changes in the ground below ballasted railway track during train passage. Proc. IMechE Vol. 221-2 Part F: J. of Rail & Rapid Transit. pp. 247-262.
• Burrow, M.P.N., Evdorides, H.T., Ghataora, G.S., Gunn D.A. (2006). A Comparative Analysis of Railway Foundation Design Principles: A Case Study. Railway founations, RailFound 06, proceedings of the first international conference on railway foundations. University of Birmingham, UK. pp 316 -327.
• Clayton, C.R.I., Grabe, H., Powrie, W. (2006). Ground Investigations and Monitoring for Track Formation Problems. Railway founations, RailFound 06, proceedings of the first international conference on railway foundations. University of Birmingham, UK. pp 1 -22.
• Ghataora, G.S, Burns, B., Burrow, M.P.N., Evdorides, H.T. (2006). Development of an index test for assessing anti-pumping materials in railway track foundations. Railway foundations, RailFound 06, proceedings of the first international conference on railway foundations. University of Birmingham, UK. pp 355 -366.
• Ghataora, G.S., Burrow M.P.N. (Editors) (2006). Railway Foundations. RailFound 06, Proceedings of the First International Conference on Railway Foundations. University of Birmingham Press. ISBN 0704426005 / 9780704426009.
• Gunn, D.A., Nelder, L.M., Chambers, J.E., Raines, M.R., Reeves, H.J., Boon, D., Pearson, S., Haslam, E., Carney, J., Stirling, A.B., Ghataora, G.S., Burrow, M.P.N., Tinsley, R.D., Tinsley, W.H., Tilden-Smith, R. (2006). Assessment of railway embankment stiffness using continuous surface waves. Railway Foundations, RailFound 06, proceedings of the first international conference on railway foundations. University of Birmingham, UK. pp 94 - 106.
• Gunn, D.A.1, Nelder, L.M.1, Ghataora, G.S, Stirling, A.B., Konstantelias, S., Burrow, M.P.N. (2006). Geophysical properties of the railway subgrade at a site in Leominster. Journal of the Permanent Way Institution, 124, (3), 131-135, July 2006.
• Huang, S., An, M., Burrow, M.P.N., Ghataora, G.S., Baker, C.J. (2006). A Potential Application of the Fuzzy Reasoning Approach to Railway Foundation Maintenance. Railway Foundations , RailFound 06, proceedings of the first international conference on railway foundations. University of Birmingham, UK. pp 134 - 151.
• Powrie, W. (2006). RRUK and Track Systems Research. Railway foundations, RailFound 06, proceedings of the first international conference on railway foundations. University of Birmingham, UK. pp 187 -189.
• Bowness, D., Lock, A.C., Richards, D.J., Powrie, W. (2005). Innovative remote video monitoring of railway track displacements. Applied Mechanics and Materials, Vol 3-4. pp. 417-4223.
• Bowness, D., Powrie, W., Clayton, C.R.I. (2005). Formation dancing: factors affecting railway trackbed. Ground Engineering, 38 (5), 30-33, ISSN 0017-4653. Reproduced by kind permission of Ground Engineering.
• Burrow, M.P.N., Ghataora, G.S., Stirling, A.B.S. (2005). A rational approach to track substructure design. Railway Track Integration Conference, Derby, U.K, 25-26 November 2005.
• Naito, S., Burrow M.P.N., Madelin, K.B. (2005). The development of a network level railway track condition model. Railway Engineering 2005, London, U.K.
• Priest, J.A., Bowness, D., Clayton, C.R.I., Powrie, W., Richards D. J., Gräbe, P. J., Maree, J. S. (2005). Instrumentation testing and geophysics on the Richards Bay coal line, South Africa, Railway Engineering 2005, London, U.K.
• Yang, L., Clayton, C.R.I., Powrie, W., Bowness, D. (2005). Preliminary investigations into the effect of principal stress rotation on railway subgrade behaviour. Railway Engineering 2005, London, U.K.
• Burrow, M.P.N., Ghataora, G.S., Bowness, D. (2004). Analytical track substructure design. ISSMGE TC3 International Seminar on Geotechnics in Pavement and Railway Design and Construction NTUA - Athens, 16-17 December 2004.
• Burrow, M.P.N., Ghataora, G.S., Stirling, A.B., Madelin, K.B., Konstantelias, S., Kingham. P., Bowness, D., Powrie, W. (2004). Track subbase monitoring and performance. Railway Engineering 2004, London, U.K.
• Powrie, W., Bowness, D. (2003). RRUK – a new start for track systems research. Eur Railway Review 2003, Issue 3, 57-61. ISSN 1351-1599

More detailed information and access to UK events associated with this project will be available to members of the Engineering Interfaces Theme Network