The focus of the course is on the quantification of uncertainty in multiscale multiphsyics simulations for engineering analysis. Though engineering simulation has become the mainstay of academic and industrial analysis in recent years, there has been little emphasis in understanding and quantifying the source of uncertainty in predictions. These uncertainties may arise for a variety of reasons: lack of mesh-independence, inadequate physical models, uncertainties in geometry, operating conditions and material properties, among others. The course introduces the student to the concepts of verification and validation, sensitivity analysis, and uncertainty propagation using sampling methods, polynomial chaos and reliability based methods. An introduction to uncertainty quantification software is also given. Emerging topics in uncertainty quantification of microsystem simulations are presented.
Researchers should cite this work as follows:
Birck Nanotechnology Center, Purdue University, West Lafayette, IN
|Lecture Number/Topic||Online Lecture||Video||Lecture Notes||Supplemental Material||Suggested Exercises|
|Uncertainty Quantification in Experiments||View Flash||View||Notes|
|An overview is provided of how experimental data should be reported with true uncertainties. Examples from experiments on gas damping measurements in RF switches and for estimation of nanoparticles …
|Uncertainty Propagation in a Multiscale Model of Nanocrystalline Plasticity||View Flash||View||Notes|
|Uncertainty Quantification of Molecular Dynamics Simulations||View Flash||View||Notes|