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

AI-based surrogate models could speed up structural engineering simulations by predicting results that would otherwise require finite element analysis. This thesis at AFRY will assess their accuracy, efficiency and suitability for structural CAE. The aim is to clarify when surrogate models are useful and when conventional FE simulations are the better choice.

The assignment

• Use a finite element structural model as a case study and develop AI-based surrogate models using results from implicit and explicit analyses.
• Assess prediction accuracy and the amount of training data needed.
• Compare computational performance, simulation turnaround time and engineering effort with conventional FE simulations.
• Evaluate the suitability of surrogate models for both implicit and explicit analyses.
• Document the findings in a report with recommendations and guidelines for industrial applications.

Who we are looking for

• Master’s student in Applied Mechanics or an equivalent programme.
• Interested in undertaking the project together with another student.

What we offer

• Join AFRY’s CAE & Vehicle Safety team in Trollhättan, working with strength, durability and NVH analyses for future vehicles and systems.
• Gain exposure to an engineering environment serving local and international customers.

Practical details

The project will be carried out in Trollhättan. AFRY is seeking two students to work together. For questions, contact Ilia Dokhani, Senior CAE Engineer, at ilia.dokhani@afry.com, or Robert Hansson, Lead CAE Engineer, at robert.hansson@afry.com.