POD-accelerated CFD analysis of wind loads on PV systems

Abstract

Paper presented to the 10th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics, Florida, 14-16 July 2014.

A real-time response framework based on the Proper Orthogonal Decomposition (POD) method is proposed to provide a solution that would not only take advantage of the great detail and accuracy of a grid-converged 3D computational fluid dynamics (CFD) analysis but also calculate, in real-time, flow features and loads that result from wind-induced drag and lift forces on Photo-Voltaic (PV) systems. The key is to generate beforehand and off-line an extensive set of solutions, i.e. pressure and shear stress distributions over the PV system surface, using CFD within a predefined design space (module sizes, wind speeds, topographies, roof dimensions, pitch, etc.). These solutions are then organized to form the basis snapshots of a POD decomposition matrix. An interpolation network using radial-basis functions (RBF) will be employed to predict the solution from the POD decomposition given a set of values of the design variables. The entire POD matrix and RBF interpolation network are stored in a database that can be accessed remotely by the wind-load calculator tool and therefore predict the solution, flow features and loads, in real time. The trained POD-RBF acts as a multifaceted interpolation that preserves the physics of the problem and has been tested and validated by performing the fast algebraic interpolation to obtain the pressure distribution on the PV system surface and comparing them to actual grid-converged fully-turbulent 3D CFD solutions at the specified values of the design variables (wind speed and angle).