Bump steer effects in a 10 ton truck with a leaf-spring suspension

Abstract

Abstract

A 10 t truck with a leaf-spring suspension and beam axles was known to have suffered an incident during normal road operation that was believed to have been related to the phenomenon of bump steer. Experimental investigations using a four-post rig established that the vehicle was particularly excited while traversing, at approximately 25 m/s, a sinusoidal undulating road profile that had a wavelength of approximately 12 m. Bump steer was evident under these conditions, and the excited frequency of approximately 2.4 Hz raised the possibility that driver reaction times (a lag of approximately 0.2 s) could have compounded the problem when trying to counter the effect of the bump steer.

A handling model of the vehicle has been used to investigate the bump steer effects. A three-degree-of-freedom handling model of the vehicle was created that included a non-linear tyre model, which was able to simulate the differing cornering stiffness effects created by varying tyre loads, friction limits, and slip angles. A driver model was created that made use of lateral displacement and intended vehicle-path return rate feedback loops (effectively proportional and differential feedback) but incorporated a neuromuscular lag element and steer rate and steer angle limiting functions.

Simulations of violent high-speed double-lane-change manoeuvres enabled the driver's return rate and lateral displacement feedback gains to be benchmarked for these conditions. Further simulations using these feedback gains showed that, when coupled with the bump steer of the vehicle at an excitation frequency of 2.4 Hz, a typical neuromuscular lag of 0.2 s created an approximately 150° out-of-phase response from the driver that resulted in an unstable oscillatory handling system. If the neuromuscular lag was reduced to near-zero, the system remained stable in spite of the high feedback gains. Alternatively, reducing the driver's feedback gains by 50 per cent also resulted in a stable system, in spite of the neuromuscular lag, thereby demonstrating that the correct response to the bump steer effects was not to respond to them. Actual vehicle trials conducted under controlled conditions at various road locations (including the incident location) and other test centres demonstrated that this was indeed the correct response should bump steer be encountered.