This presentation was made at the 2016 NAFEMS India Conference.

Resource Abstract

This paper presents a case study requiring the user to perform a design synthesis task on a validated FE model. The objective is to meet NVH performance targets expressed in decibel units. The challenging task in this particular scenario is that the FE model is huge and involves creation of several bode plots for phase and magnitude graphs in the postprocesser to predict inertance transfer function over several nodes in the FE model. It is subsequently required to parse the ordinate values off each of these many bode plots and locate the nodes where the ordinate values exceed the targets over desired frequency ranges by superposing target values using Legacy calculations on the same graph. Further, it is required from user perspective that these identified locations are directly highlighted and tagged in the Pre-processor in a graphic manner. This is to facilitate the user to graphically interact with the model rather than interpret tacit graphs and then toggle across the Pre-processor and Post processor, scouting for locations to make design modifications which is a laborious task. The author hereby presents an innovative tool developed through application of Lean Six Sigma principles to automate this entire design synthesis task by seamless integration of the preprocessor with postprocessor that phenomenally speeds up the process. In addition, the knowledge based tool also captures legacy calculations and data interpretation heuristics through efficient use of Perl, native file formats, and native macro commands. Finally, the Lean Six Sigma Statistics for measurement of success metrics is also presented to quantify the productivity achieved.