Machine Learning Engineering for Multiphysics Design and Life Prediction of Turbomachinery

Purpose and goal:

The purpose of this project is to increase the use of machine learning in solid mechanics calculations in turbomachinery with the goal of achieving overall faster structural integrity calculations. Machine learning will be applied in predicting fatigue life in high temperature components as well as in material modeling of high temperature materials.

Expected results and effects:

This project is expected to result in significantly faster calculations in structural integrity with the effect that components can be designed faster and better. The project is expected to result in machine learning based computational methods for predicting the elasto-plastic strain field and the fatigue life. In addition, the developed computational methods are expected to be generalized to apply also to industrial applications other than turbomachinery.

Approach and implementation:

The project is carried out as a three year PhD project with partners from academia, the turbo machine industry and the industrial software industry. The project is coordinated by Linköping University in collaboration with Siemens Energy AB and Siemens Industry Software AB. The project will begin by developing machine learning-based calculation methods for fatigue life prediction and then focus on the equivalent for elasto-plastic material modeling.