Date of Award

1997

Publication Type

Doctoral Thesis

Degree Name

Ph.D.

Department

Mechanical, Automotive, and Materials Engineering

Keywords

Engineering, Mechanical.

Supervisor

North, W. P. T.,

Rights

info:eu-repo/semantics/openAccess

Abstract

During the past two decades there has been a steady increase in the number of computer numerically controlled (CNC) machine tools. Such problems as dimensional accuracy and surface finish are recently receiving widespread attention. This theoretical investigation explores the possibility of decreasing the surface roughness by suppressing radial tool-workpiece relative displacements via control of the depth of cut. In order to achieve this goal various control strategies are developed and evaluated via simulation. An integrated model of the turning process in the most flexible component which manifests itself in the surface roughness, i.e., the radial direction was generated from an extensive literature survey and is presented. Three major programs were developed in Matlab so as to simulate the turning process and to evaluate the effectiveness of controllers based on classical and modern techniques. First of all a proportional plus integral plus derivative (or simply PID) controller is designed and a typical machine turning centre is chosen from published literature in order to evaluate the effectiveness of this controller. Secondly, optimal control strategies were designed based on Linear Quadratic, LQ, and Linear Quadratic Gaussian (LQG) methods. Their effectiveness was also verified via simulation. Pole-placement and frequency shaped designs were carried out. Improvements superior to the PID controller were achieved especially with the frequency shaped LQG design. Finally, a parameter adaptive controller was designed and its performance evaluated via simulations. A self tuning regulator (STR) based on LQG methods was developed. All controllers were stable and robust with high enough bandwidths to easily accommodate cuffing speeds around 2000 rpm at 20 samples per revolution while satisfying the Nyquist sampling theorem. If a higher speed is required the number of samples per revolution should be reduced so as to be consistent with the Nyquist theorem. The same holds if a higher number of samples per revolution is desired, i.e., the cuffing speed must be reduced. (Abstract shortened by UMI.)Dept. of Mechanical, Automotive, and Materials Engineering. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis1996 .L525. Source: Dissertation Abstracts International, Volume: 62-10, Section: B, page: 4746. Adviser: W. P. T. North. Thesis (Ph.D.)--University of Windsor (Canada), 1997.

Share

COinS