Date of Award

Fall 12-1-2004

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Industrial Engineering

First Advisor

M. Affan Badar

Second Advisor

Marion D. Schafer

Third Advisor

Ming Zhou

Abstract

Efficient part feature verification using Coordinate Measuring Machine (CMM) requires efficient sample point selection as well as accurate zone fitting. This work deals with flatness tolerance evaluation and the Least- Square (LS) technique is used to compute tolerance zone. In regard to the sampling strategy there are several methods commonly employed in Coordinate Measuring Machine inspection literature. The appropriate method is selected according to the accuracy requirement, the geometry features of the work piece, and the condition of the machine producing the work piece. This research focuses on the five sampling methods: Hammersley sequence sampling, Halton-Zaremba sequence sampling, Random sampling, Uniform sampling and Tabu search method. The objective of this work is to compare the estimation accuracy of these methods with respect to sample sizes 4, 16, 64 100 and 256. Two- factor experimental design was carried out to achieve the objective. Based on the mean accuracy obtained for all sample sizes (4, 16, 64, 100, and 256) combined, Halton-Zaremba was found to be the best and for all sampling methods (Uniform, Random, Hammersley, Halton-Zaremba, and Tabu search) combined, sample size of 256 was found to be more accurate.

Download

Share

COinS