Master Thesis Project: Evaluation of AI-based Surrogate Models for Structural CAE Analysis

Artificial Intelligence is rapidly being introduced into engineering simulation workflows. This thesis aims to evaluate the applicability and potential efficiency gains of AI-based surrogate models in structural CAE analysis. The study should investigate prediction accuracy compared to traditional FE simulations, required amount of training data, potential reduction in simulation turnaround time, and applicability for implicit and explicit analyses. The final objective is to establish guidelines for when AI-based surrogate models should be used and when traditional FE simulations remain the preferred approach. The study will use a finite element structure model as a case study. Surrogate models will be developed using data from both implicit and explicit analyses, and their predictive accuracy, computational efficiency and required engineering effort will be compared with conventional FE simulations. The thesis project will be performed in Trollhättan, in the department CAE & Vehicle Safety, as part of the team performing strength, durability and NVH analysis. The company aims to find two students to do this project together.

Responsibilities

Evaluate AI-based surrogate models for structural CAE; Assess accuracy and potential time savings; Evaluate applicability for implicit and explicit analyses; Establish guidelines and recommendations for industrial use

Requirements

Master Student at Applied Mechanics or equivalent program