Modeling Fundamental Processes in Soil-Machine Interaction

Jin, Zhefei

doi:https://doi.org/10.21985/n2-mnc9-js40

Work

Modeling Fundamental Processes in Soil-Machine Interaction

Public

Download PDF

Download All Files (.zip)

Soil-Machine Interaction (SMI) is ubiquitous on Earth and other planets. Over decades, ground-engaging equipment has been developed and widely used in various engineering applications, including civil construction, agriculture, terrestrial mobility, and even space exploration. However, systematic studies of SMI processes and the corresponding soil behavior are inadequate, which obstructs further advances in the automation and optimization of the machines. This study focuses on the development of experimental methodologies and efficient computational techniques for promoting the fundamental understanding of loading histories and soil deformation patterns in SMI processes. The work starts from developing and evaluating a novel experimental system consisting of an industrial robot and fluidized bed to enable automated and efficient experimental studies in SMI. The fluidized bed allows a repeatable preparation of a uniform sand bed with a broad range of relative density, and the robot functions as a versatile actuator while simultaneously tracking multiple components of soil reaction acting on the tools. Experiments on fundamental penetration and cutting processes were performed on the newly developed system to observe the soil deformation evolution and measure force-displacement responses. With the help of the test results on soil deformation, a sequential kinematic method is developed. This method consists of evaluating the incremental displacement field based on the kinematic method of plasticity and then sequentially updating the deformed configuration. The model was calibrated against experimental results to investigate three fundamental problems: biaxial compression, cutting, and penetration. It is demonstrated that the model can accurately predict the complex response observed in the SMI processes, including soil deformation and force-displacement histories. Although the computational cost is affordable compared to conventional methods such as FEM and DEM, the run time is still high and cannot satisfy the requirement for design optimization and control algorithms used in robotics. To overcome this limitation with respect to efficiency, a semi-analytical (also known as macro-element) modeling framework, inspired from the generalized yield envelopes widely adopted in geotechnical engineering, is developed. Despite its simplicity, the proposed model can generate force-displacement data in seconds for the loading scenarios of a plate moving translationally and rotationally, considering large deformation in sand. By integrating the experimental studies with the numerical investigations, this work opens opportunities for a comprehensive study on soil-machine interaction in an efficient way.

Creator