Multi-Disciplinary Design Optimization in Support of the Functional Performance Engineering Process
Virtual prototyping can lead to shorter design cycles, reduced design costs, and products with superior performance. Computer-aided design methods often focus on product data, form and fit. However, within the virtual prototyping process, there is a key requirement to identify the design features that have most influence on the functional performance. These are the design features needed to achieve the desired targets. This paper discusses the use of a multi-disciplinary optimization program, LMS OPTIMUS, to solve industrial optimization problems, using existing legacy simulation software. The methodology is based on Design of Experiments and Response Surface Modeling, coupled to numerical optimization techniques. Available algorithms include Gradient methods, Genetic Algorithms and Simulated annealing - covering both the areas of local and global optimization. Robust design techniques like Monte Carlo analyses enable the variability inherent to any industrial process to be taken into account. By using parallel processing capabilities, an open software architecture that facilitates integration with multiple analysis codes, and an extended toolbox of methods, users are able to get the most out of their analyses. Examples include durability and crashworthiness, vehicle ride and handling, noise and vibration and harshness, structural integrity and weight. Details are given of a particular application of LMS OPTIMUS alongside LMS FALANCS and standard linear and non-linear structural FE analyses, for mass optimization under durability and impact loading constraints.