[SanFer]

Hi everybody,

I am running the transient case of an external compressible (M=0.6) flow around a 3D flat ellipsoid. The body, which has a circular planform, a relative thickness of 0.2 and no sharp trailing edge, has to perform two sine-wave movements: a lateral asymmetric movement and a foreaft symmetric movement. The solution is performed using the inviscid equations and the density-based solver in both full-model and half-model meshes.

The full-model (no symmetry) case is initialized with the steady solution. However, when the initial transient effects disappear, the lift monitor does not show a periodic behaviour, and some erratic fluctuations appear in both lateral and foreaft movements. On the other hand, when a half-model is used with a symmetry boundary condition, the forces in the model show a stable periodic behaviour (foreaft movement).

I need to use the complete model for performing the lateral movement solution. Can this behaviour be due to the lack of symmetry in the mesh (tetra-mesh with ANSYS Meshing)? Or could there be any other source of instability, like the absence of a sharp trailing edge? Any possible solution?

I appreciate any help!

Regards,
Emilio

[Blanco]

I think that a non-symmetric mesh can be the root cause of the described behavior, even if you have to check however for solver convergence at each time-step.

Btw, what about trying to obtain a full-scale mesh starting from the half-model mesh? If I remember well you can mirror the mesh before exporting it in ANSYS, so you can still use your half-model mesh, mirror it, merge vertices and boundaries where needed and then obtain a full-model mesh for your simulation (this way assuring a mesh symmetry along a desired axis).