Behaviour of two layer railway track ballast under cyclic and monotonic loading.

New railway track is laid to a specified level and alignment. However, with time and
trafficking movements occur in the ballast bed and the underlying subgrade, and frequent
maintenance is necessary to re-establish the correct geometry. This is currently done by the
process of tamping, where extra material is vibrated under the sleeper to raise its level.
Unfortunately this is not permanent, and the sleeper eventually reverts to its premaintenance
state. Prior to mechanised tamping, track was relevelled by hand shovel
packing in which the sleepers were raised and fine aggregate shovelled into the space under
the sleeper, giving a permanent lift. A prototype of a mechanised version of this process,
called the Stoneblower, has been developed and is currently undergoing field trials. It lifts
the sleeper and blows single size stone smaller than the ballast into the void space, creating a
two layer granular foundation for each sleeper.
In order to get a better understanding of the behaviour of a two layer granular foundation
when subjected to repeated loading a laboratory study involving large scale cyclic triaxial
testing of single size and two layer ballast has been carried out. This has been
complemented by a large scale model study using a 'half sleeper rig' in which ballast beds
similar to those produced by stoneblowing have been subjected to repeated loading
simulating long term railway trafficking. These tests have been used to carry out a
parametric study into stone angularity, stone size and layer geometry to assess their
influence on the layered system, with the results being assessed primarily in terms of the
resilient modulus and plastic deformation undergone by the material.
In the triaxial tests it was found that the layer of smaller material was the major controlling
factor for the deformation. However, the layer of larger material tended to act as a restraint,
effectively reducing the HID ratio of the specimen. In the model testing, it was found that
the smaller material dictated how the load was transmitted to the ballast bed, and this was
then responsible for the majority of the settlement.