Stochastic analysis of dynamic stress amplification factors for slab track foundations

The relationship between railway track-bed stresses and speed is essential when performing track foundation design. Current approaches to describe this relationship are typically empirical and multiply the static track response by a dynamic amplification factor to account for speed effects. This paper develops an alternative method for dynamic amplification factor generation instead of numerical simulation. A vehicle-track dynamics model is presented, capable of simulating the subgrade dynamics considering random track irregularities. First, samples of irregularity are generated based on a mean vertical track profile power spectral density. A range of train speeds and track-spectrum cumulative probabilities are computed using the numerical model before analyzing the resultant surface stresses. The resulting dynamic amplification factors follow a Gumbel distribution. Data normalization is performed before developing a bespoke dynamic amplification function. Based on speed, track-spectrum cumulative probability, and p-values, the proposed model yields dynamic amplification predictions within 5% of the simulated values.