Parameter Estimation
How to tame a vast number of combinations
use A I to learn material properties for e.g.
- spotwelds with different material combinations
- composites with different fiber or matrix materials
training results of machine learning models for parameter prediction
check model validity
odd coverage, load distribution of experiments
analyzing input space
Your model parameters look perfect, but is your model valid within your load case?
- find sparse areas where no experimental data exist
- check if your simulation result is covered by experimental data
use unsupervised learning methods