This paper was produced for the 2019 NAFEMS World Congress in Quebec Canada

Resource Abstract

We present a collaborative business platform for design, analysis, simulation and data management. It enables enterprises to manage and deploy flexible and scalable product development processes on premise or in the cloud.



Design of Experiments is a tried-and-true product development process that is used to understand the cause and effect between the input and output parameters of a process (virtual or experimental). Instances of the input parameters are sampled within a range, and for each of these samples the output behavior of interest is computed and stored. The results are then post processed to determine input/output relationships.



At first glance, it seems straightforward to implement DOE in SPDM by simply creating very large numbers of input samples and computing the results on a large high performance cluster. However, there are challenges to build model instances that are robust in terms of input variations and there are challenges to converge the DOE iteratively to an optimal design for complex design spaces.



Therefore, we developed a new iterative process to find the best achievable performance trade-off for a product in a given milestone period by leveraging all resources on the platform in the most efficient manner possible:

• Platform & Applications that make it easy for users to collaborate, author, execute and post processes models in an iterative automated workflow.

• Users who create reusable models to deploy simulation earlier in the product development cycle, and make adjustments as more information becomes available after every iteration.

• Hardware which enables the platform and its applications to scale iteration time with computational resources.



The iterative automated DOE workflow will enable the solution of computationally intensive design and operational problems through efficient and parallel use of hardware together with user-interaction and guidance.



To illustrate this approach, we applied it to the design of 20 components of a conceptual automotive structure in a 55 km/h frontal crash scenario. The purpose of the design optimization is to reduce the driver accelerations to 35 g and limit the structural deformations to 550 mm. Such a structure is hard to optimize because changes in a single component can change the collapse sequence of the beams, thus creating a bifurcation in the crash response. We achieved good results in just 5 Adaptive DOE iterations, each iteration submitting all 40 frontal crash simulations in parallel, each simulation on multiple cores. This reduced the frontal crash optimization process to just a few days.