ASW4DTs - Automating Simulation Workflows for Digital Twins: Assessing the Influence of Defects on Aerodynamic Performance
Project duration
May 01, 2026 - April 30, 2029
Project partners
Funding
The project is funded by the Helmholtz School for Data Science in Life, Earth, and Energy of the Helmholtz Association and Forschungszentrum Jülich.
Project description
Reverse engineering (RE) is essential for analyzing manufactured components, which often show defectsfrom process variations or damage such as cracks, bumps, or corrosion. In wind turbine maintenance, RE simulation workflows can predict component performance; similar applications include aircraft turbines and additive manufacturing. Traditional RE requires manually converting 3D scans into computer-aided design (CAD) models for subsequent simulations - a time-consuming process due to extensive manual work.
This project develops an automated workflow to process measurement and scan data: Coarse meshes of aerodynamic components are generated from CAD models and then enriched with real-world data. Using adaptive mesh refinement (AMR) from the t8code [1] library, the meshes are refined to simulation resolution while incorporating deviations from the original CAD. The result is an automated digital twin of the object, enabling direct comparison of simulations with design models.
The project aims to integrate design data and measured defects as geometry information into the RE workflow, and adapt the simulation meshes to resolve these features. Level-set methods track interfaces such as crack propagation, while mesh adaptation and aerodynamic performance evaluation are performed at run time. This allows direct comparison between original CAD models and damaged components without recreating entire CAD models and meshes, significantly accelerating the design process.