A Mesh Motion Methodology for Efficient Flutter Simulations

Abstract

Abstract. Aeroelastic simulations of turbomachinery involve complex fluid–structure interaction analyses, which require solving three problems simultaneously: the fluid problem, the structural problem, and the mesh motion problem. Due to their complexity, aeroelastic simulations are more CPU intensive than standard computational fluid dynamics simulations. This computational overhead primarily stems from the mesh motion process, where the time-dependent mesh node coordinates lead to the recalculation of areas, volumes, and surface normal vectors of the cells at each time-step. In contrast, structural calculations remain relatively inexpensive, as they are often reduced to a limited number of modes of interest, the number of which is significantly smaller than the number of nodes in the fluid mesh. In this study, a novel method is introduced to reduce the computational burden of the mesh motion problem. The proposed approach involves precomputing mesh properties, including area, volume, and normal vectors, by imposing harmonic displacements on the turbomachinery blades. This precomputed data allow for the reconstruction of the mesh for arbitrary displacements without the need for real-time recalculations of cell properties. Therefore, this method is suitable for both prescribed motion calculations and free-flutter simulations. The method has been tested on a fan blade, including tip clearance. It is shown that the cost of free-flutter simulations can be greatly reduced by precomputing a reference grid and rescaling the displacements according to the structural solver. The impact on predicting the critical damping ratio is minimal. Moreover, the method is robust since mesh distortions can be predicted in advance.