A full-scale laboratory investigation into railway track substructure performance and ballast reinforcement
MetadataShow full item record
To reduce railway track maintenance costs and meet the growing demand for rail travel the railway industry needs to significantly increase the performance of old existing tracks and design any new tracks accordingly. In this thesis, a new full-scale laboratory Geopavement & Railway Accelerated Fatigue Testing (GRAFT) facility at Heriot-Watt University is developed to study the performance of both unreinforced and reinforced railway track substructure systems. The new GRAFT facility enables accelerated testing of full-scale railway tracks and innovative railway products under realistic railway loading conditions. The unreinforced track systems represent typical railway tracks in the UK while the reinforced track systems represent sections of track implemented with various geosynthetic products. GRAFT consists of a track constructed within a steel tank. The track comprises a 750mm clay subgrade layer overlain by a clay formation layer overlain by a 300mm ballast layer. The track includes three hardwood sleeper sections overlain by an I-section steel beam which has similar stiffness properties to a BS 113 A rail section. Cyclic loading is applied to the track from a hydraulic testing machine with the centre sleeper directly under the loading actuator. The loading mechanism replicates a repeated quasi static single wheel load on the central sleeper of one half of a 3m long section of railway track. Based on the results found from the testing programme in GRAFT empirical relationships are developed between the unreinforced track performance in terms of track settlement and stiffness and the subgrade modulus, applied load and number of applied cycles. These relationships fit the GRAFT data presented in this thesis well and it is thought that they could be used (tentatively) to estimate track settlement on track after tamping/ballast renewal/new track. These relationships are shown to be consistent with other well known track settlement models and they indicate that subgrade stiffness and applied vertical load are two of the most significant parameters that influence track substructure deterioration. The results found from the reinforced track tests quantify the improvement in track performance available with each product under various track conditions. Two ballast ii reinforcement products have been tested; XiTRACK reinforcement and geocell reinforcement, along with a reinforced geocomposite used primarily for separation at the ballast/subgrade interface. In addition, a geocomposite product designed to replace a traditional sand blanket, used on the tracks where severe subgrade erosion conditions prevail, has been tested in GRAFT under flooding conditions. The most significant results show that XiTRACK reinforcement can considerably improve the performance of railway tracks while the performance of the track implemented with the sand blanket replacement product indicates that currently a traditional sand blanket with a geotextile separator is the recommended option for tracks with subgrade wet spots. From all the data recorded empirical settlement models are proposed for each of the geosynthetics compared for reinforcement purposes. These models form the basis for reinforced track design graphs that could potentially be used to form part of an initial cost-benefit analysis of different track reinforcement techniques considered for improving track performance and reducing maintenance. In order to use the track settlement design graphs developed within this thesis (in the field) a reliable measure of subgrade stiffness needs to be made on track. A reliable insitu measuring device could enhance railway site investigations. Several in-situ measuring devices that could potentially be used to measures subgrade stiffness and strength in the field have been tested within GRAFT. The devices studied include the Dynamic Cone Penetrometer (DCP), Light Falling Weight Deflectometer (LFWD), Pocket Penetrometer and Proving Ring Penetrometer. The accuracy of these devices is compared to Plate Load Tests (PLT) and unconfined compression strength tests and suggestions towards the use of such devices on track made. The results indicate that the DCP has the potential to be a quick and accurate in-situ measuring device for railway track site investigations. The GRAFT facility and the results found in GRAFT have been validated using a basic static 3D FE computer model termed SART3D (Static Analysis of Railway Track 3D). The program has been calibrated to GRAFT by modifying the FE mesh for the dimensions of GRAFT and inputting the GRAFT track properties. The validated results from this thesis have direct practical implications to the railway industry in terms of iii design recommendations on how best to investigate and improve key geotechnical parameters that influence railway track performance and hence reduce maintenance costs and extend asset life. A review of current design procedures used in the railway industry is given and a new design procedure is suggested to reduce track maintenance and offer an optimised design and maintenance strategy.