RBF interpolation with improved data reduction algorithm — A meshfree method for fluid-structure coupling and mesh deformation

摘要

In the present work radial basis function (RBF) method, combined with improved data reduction algorithm, is presented as a unified methodology for efficient computational static aeroelastic simulations and volume mesh deformation. This method has been implemented as an extension of in-house hybrid unstructured Reynolds-Averaged Navier-Stokes solver coupled with an open source finite element solver for aeroelastic behavior predictions. The interpolation is performed on arbitrary point clouds and does not have any type of connectivity constraint between the fluid and structure mesh. Improvements have been made in the interpolation method by constructing a multi-level subspace radial basis function interpolation based on ‘double-edge’ Greedy algorithm to create an approximate interpolation for all moving boundary points. The method is equally efficient for both the structured and unstructured meshes; preserves orthogonality, computationally efficient, has no dependency on the type of flow solver and can be readily parallelized. Typical deformation problem of DLR-F6 wing-body-nacelle-pylon configuration based on its mode shapes has been selected as the test case for demonstration of the volume mesh deformation. Results show that the present mesh deformation method has good efficiency and robustness even for large deformations. Static aeroelastic simulations have been performed for HIRENASD wing/body configuration. Good qualitative and quantitative agreement has been achieved between the predicted results and the available experimental data. This method has shown its effectiveness in accurately predicting the aeroelastic behavior and preserving global grid quality after deformation.