engleski

Enhanced Reverse Engineering using Genetic-Algorithms-based Experimental Parallel Workflow for Optimum Design

Shape optimization is numerically very intensive due to multidisciplinary objectives and constraints, many shape variables, non linear models, geometric infeasibility of candidate designs, etc. It involves participation of numerical optimizers, computer- aided geometric modelers and subject-related simulators as well as their coupling at the process- and data levels. This paper develops a simple experimental workflow which employs existing commercial software for computer-aided design, finite element analysis and evolutionary optimization modules. It sets up parallel execution of multiple simulators to reduce the execution time, which is implemented inexpensively by means of a self-made .net- based cluster. Shape optimization is introduced in the generic context of ‘enhanced’ reverse engineering with optimization whereby the initial solution can be obtained by 3D optical scanning and parameterization of an existing solution.

enhanced reverse engineering; genetic algorithms; optimum design

DOI: 10.1007/978-3-642-20520-0_18