Course Description

Finite Element Model Validation, Updating and Uncertainty Quantification

Upon completion of this course, the student will be able to:

	Describe the model validation paradigm of sensitivity analysis, test/analysis correlation, and uncertainty analysis
	Summarize the history of finite element model test/analysis correlation and model validation
	Discuss current techniques for global sensitivity analysis and parameter screening
	Describe the process for selecting and computing appropriate response features from the model outputs
	Explain the role of DOE and ANOVA in model validation
	Define appropriate test/analysis correlation metrics for model revision and parametric updating

(Note: All course instruction in English only.
Course outline subject to change without prior notice)

Introduction

	Description of model validation process
	Definitions of key terminology
	Overview of finite element modeling
	Input output models: Parameters and features

Historical Overview

	Discipline specific applications (Aerospace, civil, automotive)
	Linear test/analysis correlation methods (modal parameter correlation, time history comparisons)
	Example of test/analysis correlation applied to linear structure
	Problems of interest for nonlinear model validation

Feature Extraction: “What Should the Model Predict?”

	Desirable properties of response features
	Selection process for response features
	Features commonly used for different types of model response (e.g. modal parameters, peak strains, temporal moments)

Metamodeling: “Simplifying the Response Space”

	Overview of surrogate modeling / metamodel forms
	Parameter space sampling
	Model regression & error estimation
	Optimization using metamodels

Sensitivity Analysis: “What is the Model Output Saying?”

	Global sensitivity analysis/ local sensitivity analysis
	Parameter screening
	Design of experiments (DOE)/ Analysis of variance (ANOVA)

Design of Validation Experiments: “Measuring Reality”

	Screening of variables to study experimentally
	Control and measurement of experimental variables
	Placement of instrumentation
	Variability: Replication of experiments
	Estimating and accounting for experimental uncertainties

Test/Analysis Correlation: “How do they agree?”

	Comparisons between predictions & experimental data
	Definitions of feature fidelity metrics
	Statistical testing
	How good is good enough?

Model Revision and Updating: “How to improve the Prediction?”

	Revising conceptual model form and assumptions
	Model parameter calibration
	Independent parameter assessment
	Optimization under uncertainty

Uncertainty Quantification: “How much Confidence do we Have?”

	Classification of simulation uncertainties
	Estimation of input parameter distributions
	Propagation of input parameter uncertainties

Sample Applications from the Notes:

Linear

	Aerospace structures and vehicles
	Automotive engine assembly
	Steel & concrete highway bridge

Nonlinear

	Impact of polymer foam layer
	Explosively driven impulse on threaded joint
	Three-dimensional shell buckling and crush

Demonstration of Software Packages:

	Design Expert®
	Matlab®-based model validation tools

Concluding Remarks

Course evaluation / Informal discussions