[gtg627e]

Hello OpenFOAM Forum,

I would like to analyze the effects of a vibrating structure on the surrounding flow. The set up would be say a portion of a wing's outer skin, oscillating at 1000-2000 Hz surrounded by flow moving at Mach 0.1-0.3. I would like to see what effects these oscillations have on the fluid.

I expect some separation and possibly vortex shedding, so RANS may produce significant error,
event hough unsteady RANS models are supposed to average turbulent fluctuations while preserving fluctuations from coherent structures.

So the alternative is to adopt a LES model. I am wondering if anybody in this forum has ever tried anything similar. I have access to a cluster of of 6 workstations with 2 Cpu's on each (Dell PowerEdge 1425), and I think a LES simulation would run for weeks with my set up. What is your experience, as far as simulation time, with LES?

I see that the mesh-moving capabilities have been extended to the turbulent solver in release 1.4.1. Is it possible to create an oodles solver + mesh-motion?

Lastly, I would like to impose displacements on a portion of the wing (mode shapes from modal analyses), by interpolating/mapping such displacements onto the wing profile. In file ...OpenFOAM-1.4.1/src/fvMotionSolver/fvMotionSolvers/displacement/laplacian/disp lacementLaplacianFvMotionSolver.C I noticed the following:

pointDisplacement_
(
IOobject
(
"pointDisplacement",
fvMesh_.time().timeName(),
fvMesh_,
IOobject::MUST_READ,
IOobject::AUTO_WRITE
),
pointMesh_
),

Instead of specifying uniform(1 0 0) for Dx for example, could I write a file with list<vector>:
specifying the displacements as a function of position and then interpolate the new mesh boundary given such list?

Thank you,

Alessandro Spadoni
Georgia Institute of Technology