Higher order and dynamic CFD for aeroelastic simulations

Abstract

This dissertation develops and analyzes higher order computational fluid dynamics simulations for coupled fluid-structure interactions as well as large deflection vehicle motions. The flow regime ranges across subsonic, transonic and supersonic Mach numbers and inviscid and viscid boundary conditions.

The basic Euler frame equations of motion alone are not sufficient for simulating boundary motions. For rigid body dynamics simulations, quaternion tracked orientation is robust. The Galerkin method is expensive and rate-limits finite element based CFD. Increasing order reduces the grid element count at the expense of grid sensitivity. Higher order CFD increases the numerical and geometrical complexities; stabilization and boundary conditions are especially affected. There is no maximum-timestep advantage for compressible Lagrangian CFD over Euler frame CFD. The prediction efficiency of super maneuvering, deforming and constantly remeshing unsteady Navier-Stokes computational fluid dynamics was found to improve by moving from linear to 2nd and 3rd order simulation methods.