Are you passionate about solid mechanics and intrigued by machine learning? We are seeking a qualified candidate who can merge expertise from both fields to develop innovative tools for analyzing deformation and fracture in solid materials.

JOB DESCRIPTION

Physics-informed machine learning is emerging as a promising avenue for solving problems in physics and engineering. Particularly, physics-informed neural networks (PINNs) have shown remarkable potential to solve problems governed by differential equations while accounting for measured experimental data. While a strong limitation of PINNs was their low speed, recent research has shown that efficiency and accuracy can be significantly improved by leveraging domain-specific knowledge of the problem at hand. The present project aims to further explore this research direction with a focus on 2D problems involving deformation and fracture in solid materials. Specific tasks will include:

• Contributing to developing efficient PINN-based models for deformation and fracture of solid materials
• Combining the developed theoretical models with full-field measurement techniques like digital image correlation
• Collaborating with key partners at other EU academic institutions

The candidate will be affiliated with the Damage and Fracture research group, and the research activities will be carried out in close collaboration with two larger projects: “Strength-of-materials informed neural networks” and “Network-inspired models to predict the strength of heterogeneous materials”.

WHO WE ARE

Aarhus University is the largest academic institution in Denmark and is consistently ranked among the Top 100 universities worldwide across various global rankings. The Mechanical and Production Engineering Department at Aarhus University has a strong research profile, leading both fundamental and applied research projects across multiple disciplines.
The Damage and Fracture research group focuses on achieving mechanistic understanding of the structure-properties relations for complex heterogeneous materials by exploiting synergistic combinations of computational, experimental and machine learning methods. A key theme is the development of efficient computational models for the prediction of the strength and fracture properties based on experimental data describing the size, shape and location of the defects/heterogeneities at the micro-scale. Applications range from additive manufactured metals to cast materials and composites.

• Contributing to developing efficient PINN-based models for deformation and fracture of solid materials
• Combining the developed theoretical models with full-field measurement techniques like digital image correlation
• Collaborating with key partners at other EU academic institution