This presentation was made at the 2018 NAFEMS India Conference.
The quest to develop quality products and consumables and thereby create a hallmark in the market by establishing any organization's brand name is getting more and more competitive by the day. In order to develop such high quality products engineers today rely on the results from high fidelity virtual simulation and analysis tools for shrinking the product development lifecycle. Naturally CAE users need to squeeze all available engineering functionality features offered by the commercial CAE applications for performing their complex domain design and simulation tasks, once the platform is deployed. One very important activity prior to full scale deployment of such commercial CAE applications is pre-deployment solver validation to establish confidence in results. Normally it would be expected to consider all input parameter combinations and permutations and for each case, validate the solution against four types of benchmarks. The first type of the benchmarks are the theoretical benchmarks such as those provided by NAFEMS. The second type of benchmarks are the physical testing benchmarks from authorized test agencies such as ARAI. The third type of benchmarks are the use of closed form solutions obtained by independent mathematical treatment of simplified cases through hand calculations. The fourth type of benchmarks are the acceptable analyses results obtained from prior versions of the commercial CAE applications. Thus it is needed to collate all four types of benchmarks as well as perform all types of analyses to validate them using the benchmarks. To do this entire activity would be a herculean and nearly impossible task given the shrinking product development lifecycle and intricacies involved in the high end commercial CAE applications across all their inputs. Six Sigma Quality methodologies can address this situation effectively. It involves intelligently arriving at a specific set of finite number of input parameter combinations that can address the whole set, by use of widely acclaimed methods such as Taguchi and Orthogonal arrays for optimal coverage. Further, these specific finite number of data driven simulation runs can also be performed seamlessly through process automation, for effective reuse in deployment of future versions of the CAE applications. This standardized framework is thus based on a combination of Six Sigma quality methodology and process automation and also conforms to the NAFEMS QSS for IS0 9001. Deploying such a framework would provide optimal coverage of evaluating the solver's capabilities and establish a good amount of confidence in results delivered by the solver prior to its full scale deployment across the organization.
|Date||20th July 2018|
|Organisation||The Automotive Research Association of India|