- Due to the computational burden of fluid dynamic simulations that require numerous repetitions, reduced-order models are used to reduce computational time. Reduced-order models are often based on proper orthogonal decomposition. Through proper orthogonal decomposition, the variable is decomposed into a time-averaged, spatially dependent function and a linear combination of spatially dependent basis functions weighted by time coeffcients. These reduced-order models have been shown to perform poorly when a deformable mesh is used; however, deformable meshes play an important role in nonlinear computational aeroelasticity. In this paper, modifications to proper orthogonal decomposition, which include the use of a dynamic average and dynamic basis functions, are proposed to improve the fidelity of the reduced-order model when simulating nonlinear flows and deforming meshes. 2013 by Brian A. Freno and Paul G. A. Cizmas.