Research and modeling of the tractor tire enveloping behavior

Abstract

Vibration characteristics are one of the most important criteria to assess performances of agricultural tractors. Since vibration excitation primarily arise from short wavelength ground unevenness, tire enveloping behaviour plays significant role in determining effective excitation transmitted to the tractor and human operator. In order to be able to influence tractor vibration behaviour in early stage of computer simulation based development, appropriate tire model is needed that is able to account for this phenomenon, characterized at the same time by high computational speed. In this work, two quasistatic enveloping models of the tractor tire were developed and compared to each other. Main purpose of quasistatic enveloping model is to pass effective excitation to the dynamic model of tire viscoelastic structure. For one of the models, novel approach that makes use of artificial neural network for empirical modelling is introduced. In the second case modelling approach is of physical nature, based on analytical description of the simplified tire structure represented by radially distributed springs. It was concluded that neural network based model fulfils requirements for use in vehicle dynamics simulations, due to high computational speed and satisfactory accuracy. Physically based model, due to low computational speed, is not appropriate for vehicle dynamics but can serve as auxiliary model to obtain broader set of empirical data for neural network model development, hence reducing laboratory work needed.