Experimental and numerical analysis of high-speed railway infrastructure
Esen, Ahmet Furkan
MetadataShow full item record
The demand for higher train speeds and heavier axle loads proved that the performance of ballasted tracks and conventional embankments are suboptimal. Such train loading triggers high stress levels and elevated vibration levels of the track when subjected to forces of high-speed and heavy axle trains. An effective superstructure type, known as slab track, was developed in the last decades and has already been in use all over the world. In addition to the state-of-the-art superstructure types, improvements in substructures have also become increasingly popular. The experimental research work presented in this thesis evaluates the performance of a Geosynthetic Reinforced Soil Retaining Wall (GRS-RW) system as an alternative to the conventional railway embankment. Significant savings in Carbon emissions, cost and time could be achieved if the capital costs of the track construction and the land take could be reduced. The GRS-RW substructure offers this opportunity. However, such technology requires significant performance evaluation and the development of appropriate design guidelines before the rail industry can justifiably implement it in projects. Full-scale testing is carried out on three-sleeper sections of ballasted and slab tracks by simulating moving loads at 360km/h in the Geopavement and Railway Accelerated Fatigue Testing (GRAFT-II) facility. The tracks are supported by a low-level fully confined conventional embankment and a GRS-RW substructure. First, a three-sleeper section of a precast concrete Max-Bögl slab track was tested under controlled laboratory conditions, followed by a ballasted track. Both superstructures are supported by a 1.2m deep subgrade and frost protection layer, in accordance with high-speed railway design standards. Two different axle load magnitudes are applied statically, and then cyclically/dynamically, using 6 actuators to replicate moving train axle loads. The overall aim is to assess the performance of the tracks, in terms of transient displacements and total settlements, as well as stress levels at different locations of the substructures. The experimental results show that the pressure levels on the GRS-RW wall are negligibly small for the particular test setup, proving the GRS substructure under the action of compaction reached its active state. This means that the reinforced soil was self-supporting under its self-weight and train loads, implying there was minimal pressure on the walls. Therefore, GRS-RW systems have the potential to be better alternatives to traditional earth embankments due to the enhanced soil stabilisation and lower land take. It is concluded that the slab track performs significantly better than the ballasted track in terms of elastic and plastic deformation, under both static and cyclic loading. The full-scale experimental work is complimented by a numerical part, which aims to develop three-dimensional train-track-soil models using the finite element (FE) method. The commercial software Abaqus was used to create these models. First, the GRAFT-II tested samples were modelled replicating all the track components and geotechnical parameters. The New Ballastless Track (NBT), developed by Alstom, was also simulated in the FE models, which were calibrated using the laboratory results, under the considered cyclic loading. The calibrated models are then extended to create 3D linear dynamic models, considering train-track-soil interaction, simulating train passages at various speeds. The Ledsgård case was used to validate the models. Trains travelling at low and high speeds are considered to investigate the track deflections and the wave propagation in the soil. The issues associated with critical speeds were observed in the presence of both ballasted and slab tracks.