PhD - Machine Learning-based Surrogate Modeling for Computationally Efficient Multiphysics Simulation

Hundreds of components are manufactured and continuously optimized across industries worldwide. Intense market competition demands more efficient design and simulation methods to support engineering innovation. Current design workflows struggle to handle the complexity of interacting parameters affecting friction and wear, relying heavily on empirical data or oversimplified models. In contact dynamics and elastohydrodynamic lubrication (EHL), the computational cost of accurate multiphysics simulations hinders their integration into practical design processes. Moreover, no comprehensive, experimentally validated framework exists to address these tribological design challenges systematically.

• Your role will be to develop and establish the scientific foundations for a machine learning-based multiphysics framework, using surrogate models trained on validated EHL simulations.
• You will also create a novel, computationally efficient, data-driven design protocol for lubricated components.
• Furthermore, you will dramatically accelerate the design process for complex EHL problems, enabling the development of more robust, efficient, and reliable tribological components for critical industrial applications.
• You will be at the forefront of integrating AI into classical engineering design.
• Last but not least you will also become an expert in applying machine learning to complex engineering challenges, a skill set that will make you exceptionally valuable for leading roles in both industry and academia.