Speeding-up the computation of high-lift aerodynamics using a residual-based reduced-order model

Abstract

In this article, we propose a strategy for speeding-up the computation of the aerodynamics of industrial high-lift configurations using a residual-based reduced-order model (ROM). The ROM is based on the proper orthogonal decomposition (POD) of a set of solutions to the Navier–Stokes equations governing fluid flow at different parameter values, from which a set of orthogonal basis vectors is evaluated. By considering an initial set of few snapshots at different angles of attack, a ROM is constructed which is used to predict a solution at an angle of attack which is just outside the space spanned by the POD basis vectors. The ROM solution is subsequently used to initialize the flow solver for an accurate calculation of the aerodynamics at the same flow condition. This procedure is conducted repetitively for a series of angles of attack, whereby for each and every ROM prediction, the snapshots set is augmented with the latest CFD computed flow solution. Using this strategy, a considerable reduction in the total number of iterations to reach the converged steady-state solution is achieved when compared with conventional computational techniques used in industry for a series of computations such as drag polar computations. The methodology is applied and demonstrated on a two-element airfoil and a body-wing aircraft in high-lift configuration. Furthermore, an investigation is conducted on the behavior of the reduced-order modeling approach at angles of attack close to and within the static stall region, where aerodynamic hysteresis may occur and the aerodynamic coefficients are found to be multiple-valued functions of the angle of attack. It is revealed that by constructing the ROM from an appropriate set of basis vectors, it is also possible to model the resulting bifurcation.