ECCOMAS 2024

Multi-fidelity Modelling for Digital Twins of Fast Processes

  • Dodt, Miriam (KU Leuven)
  • Marelli, Stefano (ETH Zurich)
  • Persoons, Augustin (KU Leuven)
  • Faes, Matthias (TU Dortmund University)
  • Sudret, Bruno (ETH Zurich)
  • Moens, David (KU Leuven)

Please login to view abstract download link

A digital twin is a virtual representation of a physical process, which is updated according to changes in said process but that also has the ability to control the physical process. This enables a level high of automation and control in fast production processes, which is especially interesting for nonstationary processes. In this contribution the use of multi-fidelity modelling for digital twins is proposed and discussed on the example of resistance spot welding. This allows for a good compromise between fast but also accurate evaluations, enabling reliability assessment even in areas with low data availability.