The influence of frozen climatic conditions on railway track settlement rate

Railway trackbeds in cold climates undergo repeated freezing and thawing that can destabilise the ballast matrix structure and accelerate differential settlement, leading to increased track maintenance requirements. Further, climate projections show weather patterns are likely to shift in the future, meaning it is important to be able to assess how changes in ballast freezing and thawing may affect future maintenance requirements. Despite this, no research has been undertaken to simulate the effect of freezing on the primary driver of ballasted track maintenance, vertical track geometry. Thus, this paper presents a novel framework for predicting winter-related settlement and track geometry deterioration with explicit representation of freezing persistence and freeze-thaw cycling. To do so, a 2.5D FEM-PML train-track-ground interaction model, capable of simulating track geometry changes over long timespans is coupled with (i) an air-temperature-driven thermal/ice-content workflow and (ii) a mechanistic-empirical, ice-content-dependent settlement law calibrated from large-scale triaxial tests, in which accumulation varies continuously with ice content rather than a binary frozen/unfrozen assumption. Track geometry is updated iteratively under repeated axle passages and the standard deviation (SD) over 200 m is used as a maintenance-relevant indicator. The iterative deterioration framework is validated against longitudinal level data from a mainline railway track. Winter climate effects are then explored via scenario simulations forced with hourly air temperatures from five winters under identical traffic/loading. It is shown that the resulting track geometry has a weak relationship with frozen-day totals but a positive relationship with freeze-thaw transition count, supporting transition-based winter risk screening and tamping prioritisation. Sensitivity analysis shows effective freezing temperature governs frozen-day counts, while moisture availability controls mean ice content. The model enables the performance of comparative analyses for the purpose of informing future track geometry changes under differing climate scenarios. • Ice-content method links freeze–thaw to ballast settlement and geometry change. • Air temperature drives a thermal model to estimate ballast ice content continuously. • Traffic loading updates geometry to predict growth in geometry standard deviation. • Calibrated with 2018–2023 UK mainline geometry data from an untamped section. • Freeze–thaw transitions explain deterioration better than total frozen days.