A principal component analysis (PCA) decomposition based validation metric for use with full field measurement situations

Abstract

A validation metric that involves principal component analysis (PCA) decomposition of simulation and test data is developed for potential use in the quantification of margin and uncertainty (QMU) for an upcoming USAF design exercise (OEM Aero-Thermo-Structures Design Study—USAF Task Order 0015: “Predictive Capability for Hypersonic Structural Response and Life Prediction”). This validation metric allows for use of nearly full-field, simulation and test data over a wide range of spatial realizations (3-D responses over multiple input conditions) and temporal (time or frequency) information, as needed. A demonstration example utilizing two datasets explains how the validation metric is formed and how it can be used to quantify the margin between the simulation and the test data as well as how it can quantify the uncertainty. The primary advantage of the proposed PCA validation metric is that it preserves the engineering units (EU) of the original data so that the quantifications of margin and uncertainty can be made in EU. A second advantage of the PCA validation metric is that it can be used over a wide range of temporal information. While the proposed method shows considerable promise, future work is identified in terms of exploring other decomposition methods commonly used in fingerprint, iris and facial biometric pattern recognition.