Vehicles equipped with rubber track systems feature a high level of performance
but are challenging to design due to the complex components involved and the
large number of degrees of freedom, thus raising the need to develop validated
numerical simulation tools.
In this article, a multibody dynamics (MBD) model of a continuous rubber track
system developed in Part 1 is compared with extensive experimental data to
evaluate the model accuracy over a wide range of operating conditions (tractor
speed and rear axle load). The experiment consists of crossing an instrumented
bump-shaped obstacle with a tractor equipped with a pair of rubber track systems
on the rear axle. Experimental responses are synchronized with simulation
results using a cross-correlation approach.
The vertical and longitudinal maximum forces predicted by the model,
respectively, show average relative errors of 34% and 39% compared to
experimental data (1–16 km/h). In both cases, the average relative error is
lower for tractor speed from 1 to 7 km/h, namely 20% and 35%. The model and
experimental amplitudes spectra of the force signals are compared using the
coefficient of determination r
2. In the 1 to 7 km/h tractor speed range, the average vertical and
longitudinal coefficients of determination are, respectively, 0.83 and 0.42. The
coefficients, respectively, reduce to 0.27 and 0.14 for speeds over 7 km/h.
In summary, the model can predict the maximum vertical and longitudinal forces in
addition to the amplitude spectrum of those signals for operating conditions up
to 7 km/h, regardless of the rear axle load, with accuracy acceptable for many
applications, such as load case determination for preliminary structural design.
Several factors affecting the accuracy of the model at higher tractor speed are
identified for future work including suspension creeping, suspension compression
characterization at high strain rates, and temperature dependence of material
properties.
