On non-linear methods of bogie stability assessment using computer simulations

Abstract

Stability assessment of rail vehicles is probably the most widespread form of dynamic analysis in railway vehicle engineering. The computer simulations using fully non-linear three-dimensional vehicle models constructed in a modern multi-body simulation tool allow detailed non-linear stability analysis for the specified conditions. However, high sensitivity to the wheel/ rail contact conditions and different definitions of stability in mechanics and in railway practise can lead to significant differences between prediction and the measurement. Different methods of non-linear stability analysis, which may be used in industrial applications, are introduced and compared on selected examples of contact geometry wheel set/track with high equivalent conicity. The comparisons show that the linearization of the contact geometry wheel set/track can enable a better assessment of the non-linear stability analyses. A decreasing equivalent conicity function in the range of amplitudes below ∼ 3 mm leads to supercritical Hopf bifurcation with small limit cycles and consequently to largest differences between the methods compared. When excluding the contact geometries leading to supercritical Hopf bifurcation, the results achieved are closer each other, but still with differences in the range of up to 10 per cent. This uncertainty in the stability prediction caused by the method applied must be taken into consideration, in addition to other uncertainties related to vehicle parameters, modelling, etc.