Inverse Airfoil Design Using Proper Orthogonal Decomposition on Computed Flowfield Snapshots

The gappy proper orthogonal decomposition (POD) proposed and explained in detail in Bui-Thanh et al.1 which allows one to cast an subsonic and transonic inverse airfoil design task as a gappy data problem simply and efficiently is outlined here. First a collection of flowfield snapshots (in prac-tice, these could also be obtained from computational simulations, experimental results, tabulated data, or a combination thereof) is constructed by choosing a set of arbitrary airfoil shapes and obtaining their cor-responding surface pressure distributions using com-pressible flow solvers. In order to solve the inverse airfoil design problem using the gappy POD method, these snapshots are redefined to contain not only the flowfield variables but also the corresponding airfoi