SDTools is involved in several communications at the ISMA24 (International Conference on Noise and Vibration Engineering) hosted by KULeuven. Etienne Balmes will talk about Hybrid FEM/test twin building challenges when it comes to an electric drive application⚡.
Keeping a rational parametrized reduced order modelling approach based on component mode synthesis, a fast model is built integrating a full FEM with refined geometry and sensors for experimental state estimation and prediction.
🔧Many tools are required to achieve such task, including
👉 Signal processing to extract features from measurements and simulation
👉 Signal processing to extract features from measurements and simulation
👉 Adequate parametrization of junctions, materials, loads
👉 Error metrics for estimation, localization, identification, updating
👉 Efficient reduction regarding CPU, storage, but keeping display and post-treatment capabilities
![](http://www.sdtools.com/wp-content/uploads/2023/11/eGMP_Expand.gif)
🎙️ Don’t miss Etienne’s presentation, session Model Updating (MU) on Tuesday 10th September at 16:05 in the Model Update session.
📝 Summary:
“An electric engine case history illustrates issues in the hybrid use of models and test data. Having fast online models for many query evaluations of responses of interest is a traditional topic. Rather than focusing on IA related inverse methodologies, one evaluates the performance of current versions of component mode synthesis leading to direct generation of parametric reduced order models. Discussions of parametrization address geometry, contact and high modal densities. Finally, model and test errors are considered for hybrid estimations of motion.”