Propagation of Uncertainty for Model Validation of Substructured Spacecraft
Abstract
In many situations, it is impossible to perform a system vibration test. If modeling and analysis are to replace system tests, then it is imperative to have confidence in the results. Correlation/validation is the path to providing this confidence and determining the predictability of models used in the decision making process. Within the new validation paradigm, there is no system level test data. Therefore, a probabilistic system correlation must be performed. The product of this research project is a complete and systematic procedure for studying the effects of substructure uncertainty on the test-analysis correlation of complex spacecraft that are validated on a substructure-by-substructure basis, using test and analysis comparisons. The uncertainty is quantified in terms of accepted modal test-analysis correlation metrics, and covariance and relation matrices associated with the differences in the test and FEM frequency response. Linear perturbation analysis is used to relate uncertainty in correlation metrics to uncertainty in substructure matrices. Covariance propagation is then used to propagate substructure uncertainty into the expected correlation metric uncertainty for the system using a Craig-Bampton based component mode synthesis approach.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 2013
- Accession Number
- ADA579035
Entities
People
- Daniel Kammer
- Dimitri Kratiger
- Sonny Nimityongskul
Organizations
- University of Wisconsin–Madison