This conference paper was submitted for presentation at the NAFEMS World Congress 2025, held in Salzburg, Austria from May 19–22, 2025.

Abstract

In the lifetime of a vehicle, ease of use and quality of appearance is as an important a goal as the longevity of the vehicle: engineers do not only need to manufacture a car whose various mechanisms will remain functional without defects after continuous use, but they must also ensure that the users will be able to operate it comfortably. This study aims to optimize the design and manufacturing of the tailgate component of a vehicle on two fronts: manufacturing quality and user comfort. The process involves the odification of the gas lifter components positions in order to perform Multi-Body Dynamic simulations followed by durability analyses. The goal is to maintain the deformations of the tailgate component at reduced levels resulting in optimum external appearance regarding panel gaps, as well as comfortable user operation. Machine Learning predictive models (also refereed as predictors) are employed in order to accelerate the product design and evaluation process. Engineers can explore various what-if scenarios and extract the necessary key responses for each modification applied to the vehicle, to estimate its improved performance and usability, without sacrificing the design time. At the same time Machine Learning predictors are employed in Optimization studies, replacing the FE (Finite Element) Solver, in order to reach the optimum design in an automated and faster way, thus, improving the product development time. In this study three optimization approaches are presented, utilizing machine learning methods that predict simulation results for two different analyses. Compared to the established "Direct" optimization method (design updates, FE analysis, post processing), the Machine Learning assisted Optimization methods significantly reduced the optimization time while maintaining similar levels of accuracy. This allowed for more optimizations studies resulting in reduced product development time and increased product performance.