Research

Mesh Generation

The ability to modify a pre-existing computational mesh is a common requirement in many areas of computational fluid dynamics (CFD) such as aeroelasticity and shape optimization. CFD tools are applied to the aerodynamic mesh whereas computational structural mechanics methods (CSM) to the structural one. In order to connect both meshes, efficient numerical tools are necessary:

1) To transfer structural deformations to the aerodynamic mesh, and aerodynamic loads to the structural nodes, and 2) to deform the CFD flowfield mesh after a new aerodynamic surface has been computed.

Another interesting field of application is aerodynamic shape optimization, where reliable and efficient algorithms for mesh movement are becoming of increasing importance. 

In the last years we have been working in developing an interpolation tool using radial basis functions (RBF) to solve data transfer problem between structure and aerodynamics. We have devised two strategies to deal with two special cases of interest. First, when the structure is represented by a beam model and there isn’t a proper structural mesh close to the aerodynamic surface. Second, when managing a full configuration aircraft and the problem has been splitted in some small blocks in such a way that there are aerodynamic nodes belonging to more than one block (Fig. 1).

Figure 1
(a) Structural and aerodynamic input meshes and (b) interpolated aerodynamic mesh from the deformed structural mesh.

After the arodynamic has been deformed, the second stage consist in moving the flow field volume mesh. To deal with large meshes, we have developed an efficient and robust tool based on domain decomposition using octree algorithms (Fig. 2).

Figure 2
(a) Wing torsion (b) detail of the interoplated volume mesh at 80% of the span and the wing leading edge.

In this context, we are working in octree based mesh generation for high order applications as well (Fig. 3).

Figure 3
Quadrilateral mesh generated with octree decomposition techniques