Grammatical Evolution Based Tool for Predicting MultivariableResponse Surface for Laser Lap Welding

Document Number: 2008-01-1372

Date Published: April 2008

Author(s):
D. Kaur - Univ. of Toledo
Dominic Wilson - Univ. of Toledo
Mariana G. Forrest - Chrysler LLC
Feng Lu - Chrysler LLC

Abstract:
The problem of predicting the quality of weld is critical to manufacturing. A great deal of data is collected under multiple conditions to predict the quality. The data generated at DaimlerChrysler has been used to develop a model based on grammatical evolution. Grammatical Evolution Technique is based on Genetic Algorithms and generates rules from the data which fit the data. This paper describes the development of a software tool that enables the user to choose input variables such as the metal types of top and bottom layers and their thickness, intensity and speed of laser beam, to generate a three-dimensional map showing weld quality. A 3D weld quality surface can be generated in response to any of the two input variables picked from the set of defining input parameters. This tool will enable the user to pick the right set of input conditions to get an optimal weld quality. The tool is developed in Matlab with Graphical User Interface for the ease of operation.

File Size: 189K
Product Status: In Stock

Included in: SP-2170

See other papers presented at SAE World Congress & Exhibition, April 2008, Detroit, MI, USA, Session: Reliability and Robust Design in Automotive Engineering: (Part 9 of 12) Model Validation and Verification

SAE World Congress & Exhibition, April 2008, Detroit, MI, USA, Session: Reliability and Robust Design in Automotive Engineering: (Part 9 of 12) Model Validation and Verification

Purchase more technical papers and save! With TechSelect, you decide what SAE Technical Papers you need, when you need them, and how much you want to pay. Learn more >