Cost-Based Design Optimization of a Structure using a Genetic Algorithm

Abstract

The traditional design optimization of a system focuses on the performance of the system and subsystems. For example, if one is designing wing of a missile, the design is mostly dictated by aerodynamics, structural stiffness, and strength criteria. This paradigm does not account for the production cost, which is one of the most important drivers behind any reallife design. Ignoring cost parameter at the design stage could lead systems exceeding budget constraints. Furthermore, it creates inconsistency among subsystems as their performance have not weighed against their cost and trade-offs have not been adequately made. In this work, we integrated a cost module in a CAD tool and used the cost information of design along with the performance criteria to find an optimum design. We chose SolidWorks as the CAD software and SEER-DFM as the cost estimation software. Using the APIs of these two software tools, we integrated SEER-DFM in SolidWorks environment. This led to a parametric cost estimation capability. We used ANSYS for calculating the stress on the parts. The ANSYS script file was written in such a way that all the model creation commands were directly applied on geometric entities. Therefore, as the design was changed in SolidWorks, the corresponding maximum stress was calculated using ANSYS without manual intervention. The optimization was done using a genetic algorithm. In the following sections, we describe the main accomplishments of this work. The cost-based design optimization was illustrated using an I-beam example. At the end, we present the main conclusions of this work.