A Simulation-Driven System Design Methodology with Manufacturing Constraints

This paper presents a holistic simulation-driven system design methodology considering multiple performance objectives, performance constraints including formability criterion defined herein, using a genetic algorithm based multi-objective optimization software GDOT, developed in-house. This tool treats multiple objective functions separately without combining them in any form. A decision-making criterion is subsequently invoked to select the “best” subset of solutions from the obtained non-dominated Pareto optimal solutions under multiple performance constraints along with a formability index. Geometric properties, associated material properties (yield strength / plastic strain to failure) are considered as design variables. An example involving the frontal impact on a rail section is used to demonstrate the methodology. This process can further suggest requirements for synthesizing new materials that will result in optimal product performance. The objective of this study is to establish an ‘optimized’ set of design parameters with the dual aim of (i) minimizing the structural weight and (ii) maximizing energy absorption efficiency of the front rail system during frontal impact. The performance constraints being maximum transmitted force, maximum intrusion, pulse efficiency and formability criterion. This study also looks at the effect of parameter uncertainty on the optimal design. This study is composed in two stages. The first stage attempts to solve the multi-objective optimization problem, which is attempted using proprietary GDOT optimization code. Stage two performs reliability-based multi-objective optimization to generate a ‘reliable’ pareto optimal front. A 2nd order meta model is developed using responses, including formability index, computed from physics-based finite element models using LS-DYNATM analysis code. Looking at a broader picture, this methodology can potentially fill the gap between numerically optimized system development and simulation-driven digital product development. This, in turn, will help realize numerical simulation-driven product development process by aiming to achieve designs that are “first time right”.