This project focuses on the optimal properties of a railway track, with the design parameters for optimization being
related to the railway sleeper and those track components which interface the sleeper. The optimization is multicriterion in the sense that it targets both the failure strength of the sleeper and the geometric stability of the track.
Since track properties vary significantly in space and time, a stochastic approach to the optimization is taken. The
optimization will rely on fast simulations, and dual level modelling will be used. At the first stage, a detailed model
based on first principles will be calibrated and validated to test results. At the second stage, a substantially simplified
surrogate model will be established, with input-output relations that mimic those of the detailed model.