Structural Optimization with Surrogate-Based Optimization Method for an Aircraft Win

Abstract

In today’s world, simulation technology has high reliability which takes plenty of time. Due to the restricted schedule of the study, extended simulation time is always a problem. Surrogatebased optimization (SBO) improves a meta-model for an optimization process. It approximates objective and constraint functions by allocating diverse designs of experiment points. The methods of optimization have been developed to yield the best results in the last two decades. To have the best option, the meta-model is improved with an adequate number of design of experiment samples. A surrogate model helps to obtain a mathematical model of simulations. As a numerical form of analysis, the meta-model is a practical way to optimize structural parts. SBO is an efficient way to optimize several simulation data, such as the results of computational fluid dynamics (CFD) simulations for complex engineering problems. This study extends a 2D-validated CFD setup to a 3D wing model, ensuring the use of a realistic structural model. Then, it effectively validates the optimization algorithms through a Speed Reducer test case. The research employs SBO in conjunction with a parametrized finite element model (FEM) by focusing on mass optimization while considering constraints of frequency and compliance. An optimum is achieved in about 36 iterations with results suggesting promising outcomes, although the results are not definitive best candidates. The optimization method's validity is further supported by comparison with established optimization methods by highlighting its potential in structural optimization.