Division of Information Technology, Engineering and the Environment
Agricultural Machinery Research Design Centre
Agricultural Machinery Research Design Centre
Design and Dynamics of some innovative passenger and light cargo vehicles based on the motorcycle.
A motorcycle towing a single-wheeled trailer forms a single-track dynamically balancing articulated vehicle. Some trailers produced commercially have suffered poor lateral stability. With improvement, the vehicles may carry loads at good speed and low cost on narrow rough tracks, particularly in developing countries or remote areas of Australia.
Literature on the lateral dynamics of motorcycles has described successful modelling and experimental verification of oscillatory normal modes such as weave and wobble. Modelling techniques have evolved from manual derivation of equations to use of computer codes for analysis of multi-body systems. Experimental studies have measured the motorcycle's steady state responses or transient response to step, impulse or harmonic excitation. System identification has been applied to problems in vehicle dynamics but rarely if at all to motorcycles.
In this work, Lagrange's Equation was evaluated in a symbolic algebra computer package to obtain equations describing the lateral dynamics of a motorcycle towing a single-wheeled trailer. The equations included degrees of freedom for motion in the yaw articulation of the hitch and roll deflection due to compliance of the tow bar. Using values for the vehicle constants measured on prototype designs of trailers, linearised equations predicted;
Four designs of single-wheeled trailer were tested experimentally, three of which had novel asymmetric tow bars connecting underneath or to the side of the motorcycle gearbox and one commercial design known to have poor stability at high speed. Test runs were made at a range of speeds and loads in each trailer towed behind an off-road motorcycle instrumented for steering torque input, yaw and roll rates, and steering and yaw articulation positions. During test runs, the rider excited the system by shaking the steering to induce small erratic perturbations from straight line running.
System identification techniques were used to fit linear time-invariant, autoregressive, state space models to the resulting single input, multiple output time series data. The main findings were as follows:
