Date of Award

Winter 1996

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Mechanical Engineering

Committee Director

Gene Hou

Committee Member

Chuh Mei

Committee Member

Stephen G. Cupschalk

Committee Member

Jen-Kuang Huang

Abstract

Extensive benchmarking on competitive cars has shown that the durability of vehicles, assembly-induced variations, and assembly cost are directly related to the quality and number of spot welds. On most luxury cars, approximately 4,500 to 5,000 spot welds are found. On less expensive cars, the range is from 3,500 to 4,500. These cars usually inherit more noise, rattles, squeaks, and other quality-related problems. These problems have motivated the search for a new capability to systematically place welds in the most critical areas, with a minimum number of welds for car bodies, while satisfying all performance requirements (e.g., stiffness, stress distribution, load paths, and durability). This research represents an initial attempt to develop this aforementioned capability.

The main thrust of the research is to develop the framework for a numerical approach for optimal pattern design of spot welds and to assess the applicability of the approach in an industrial environment.

The pattern design of spot welds in this study is viewed as a combinatory design optimization problem and is solved by a genetic-algorithm-based search method incorporated with an efficient reanalysis technique. The reanalysis technique models the spot welds as multiple-point constraints and uses the Sherman-Morrison identity to recursively calculate the new solution of the structure subjected to modification on the joint and support conditions. This optimization procedure is verified with several numerical examples. The results show that the proposed optimization procedure is effective and can be extended to realistic applications for pattern design of spot welds.

DOI

10.25777/9as2-5324

ISBN

9780591262193

Share

COinS