No abstract

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. Recent design methods are more scientifically based, and for the most advanced design methods currently in use, input parameters are typically determined from cyclic triaxial testing. The changes in stress experienced by an element of soil below a railway track as a train passes are complex, involving (for example) a cyclic rotation of the principal stress directions. In these conditions, soil element testing in uniaxial compression may lead to the underestimation of vertical strains. Testing in a hollow cylinder apparatus, which can impose the rotations in principal stress direction likely to be experienced by a soil element in the field, may therefore be preferable to triaxial testing. However, there are as yet no data to guide the designer to a rational specification of a testing programme in this more complex apparatus. This article reports the results of finite element analyses carried out to investigate the stress changes experienced by an element of soil beneath a ballasted railway track during train passage. The effects of element location, the initial in situ stress state of the soil, and the elastic parameters used to characterize its behaviour are investigated, and the modelling of the stress paths in a cyclic hollow cylinder apparatus is discussed.

This paper presents the results of a detailed investigation into the ground deformations that occur under a railway line during the passage of a train. Four horizontal boreholes were installed at different depths below a ballasted railway track. Ground deformations were measured using geophones at set distances from the centreline of the track within each borehole. The results show vertical displacements reducing with depth, from a maximum at the sleeper. Sleeper displacements are dominated by pairs of bogies at the ends of adjacent wagons (which have a frequency of loading 1 Hz), although the effects of individual bogies (2 Hz) and axles (6 Hz) are also apparent. Higher loading frequencies attenuate with depth so that at a depth of 0 . 780 m below the sleeper soffit no axles are visible within the displacement data and by a depth of 1 . 98 m only the combined effect of pairs of adjacent bogies is apparent. In contrast, longitudinal horizontal motion is greatest at a depth of 0 . 78 m below the sleeper soffit, and the longitudinal horizontal displacements at the sleeper and at a depth of 0 . 78 m are dominated by the individual axles (,6 Hz). By a depth of 1 . 98 m, the longitudinal horizontal motion is dominated by the bogie pairs. A dynamic linear-elastic two-dimensional finite element model was developed and validated using the measured displacements.

No abstract

No abstract