Abstract This study systematically investigates shimmy dynamics in a towed-wheel system under stochastic excitations, bridging critical gaps between laboratory experiments and model-based numerical predictions. A purpose-built experimental rig is constructed featuring a towed wheel on a variable-speed conveyor belt, an electromagnetic exciter, and synchronized multi-sensor data acquisition. When the bifurcation parameter is the towing speed, noise-induced transitions and stochastic P-bifurcations are quantified in a bistable regime between the fold and Hopf bifurcation points of the deterministic system. Apart from the marginal and joint probability density functions of the wheel yaw angle, the distribution of the first passage times is also evaluated when violent shimmy oscillations are likely to occur. The experimental observations are compared with the Monte-Carlo simulation results based on data-driven dynamic modeling of the shimmying wheel. The results provide a guideline for estimating the probability of accidents on motorways where stochastic perturbations appear, such as road irregularities and crosswinds.
Du et al. (2026) studied this question.