Combining Noise Factors and Process Parameters in a Response Surface

摘要

This presentation covers the strategy and analysis of an experiment to211u001echaracterize a gas tungsten arc welding process. The experiment combined four 211u001euncontrolled noise factors and four controlled process parameters. A 211u001enontraditional response surface design was employed. Multiple responses were 211u001emodeled. Optimal settings for the process parameters to successfully weld the 211u001ewidest range of the pertinent product features were identified. Thus, the process 211u001ewas made against noise factors. Comparisons are made between the experimental and 211u001eanalytical approach taken versus the Taguchi style of experimentation and 211u001eanalysis. This comparison is mainly done with respect to the information gained, 211u001esuch as product design criteria, incoming material specifications, and process 211u001eadjustments for nonconforming material.