Evaluation and Optimization of Dozing Operations for Small-Scale Wheeled Vehicles with Simulation and Validation Using a Reference Platform

Abstract

Currently the majority of dozing operations are conducted using tracked vehicles with a fixed blade cutting angle. There is interest from industry in exploring alternatives to tracked vehicles for small scale dozing operations because of the limitations of the platform. One alternative is a six-wheeled vehicle with independent drives and an active suspension. This alternative has higher mobility and controllability than tracked vehicles, but generally produces less traction. This decrease in traction leads to lower dozing ability when compared to tracked vehicles. The purpose of this thesis is to study the process of robotic dozing and understand the value and limitations of dozing with wheeled vehicles. This study focuses on the evaluation and design of the blade and vehicle suspension as major contributors in dozing operations. To begin, the blade-soil fundamentals were studied and it was found that there could be large increase in dozing performance if the blade was changed from a traditional vertical angle to a variable attack angle. Therefore, a variable angle blade design that would decrease the power required for dozing application was investigated. The forces experienced during dozing were derived and an optimization was developed to change the blade angle in response to different dozing conditions. This new design and control is shown to increase the dozing ability when compared to a vertical blade. For shallow dozing, this a variable angle blade can reduce the overall push force required up to 75% while reducing the total volume per swipe by only 15%. Following the evaluation of a new blade design the second component, the suspension, was investigated. First, a simplified version of the reference robot being studied was constructed in CAD and imported into SimMechanics. Next, in a joint MatLab SimMechanics interface, a simulation of varying suspension properties and the robot operating environment was created. The simulation was then validated with physical testing of the robot. Following validation, an optimization program was written such that the optimal suspension properties are determined for different operation scenarios. The use of this semi-active suspension has shown to increase the applied force for the wheels up to 16 percent. To conclude the study, three case studies which combine the new blade design and the SimMechanics suspension optimization are presented.