My analysis is similar to a flapping wing. How can I give a predefined deformation to a body (instead of applying a force and then getting a deformation)? I want to use that deformation to disturbe a surrounding fluid. My biggest problem is creating a mesh that moves the way I want it to with time. Can this be done in CFX/ANSYS? If not, do you know a software that can?

Thank you.

Yes, you can turn on mesh motion, and then prescribe mesh location on a given wall boundary. The motion can be CEL expression and a function of time.

thanks johnny. Now let me ask this: what if the mesh deforms with different increments at different locations, how can I have that mesh deform in a synchronized motion? I m thinking setting up some keypoints or uploading a different mesh for each time step though I m not sure if that's possible.

Hi,

The ball valve tutorial in the CFX Tutorial manual does mesh deformation by both CEL expressions and different mesh files from memory. Have a look at this tutorial.

Glenn Horrocks

Thanks, but in that tutorial, the entire mesh move uniformly. I am trying to move the mesh non-uniformly. Say I had a cantilever beam and applied a force on it, the beam would deform non-linearly. How can I recreate that deformation without applying a force but by applying a displacement? In the help files it says you could apply a displacement to each node of your mesh using CEL. Does anyone know how to select individual nodes?

Hi,

Where can I find more information about "reading different mesh files from memory"? The manual only says something like that the topology must be the same.

Can I use tet-mesh? Have anyone here done something like this before?

regards

Michelle

My understanding is that you can apply the displacement as a CEL function f(x,y,z,t) but if you need to move individual nodes different amounts in a way that cannot be described as a function, then you may need to look at something other than CEL. I have written CEL functions that do prescribed rotation of a body as functions of time and space and it has worked quite nicely.

A note of caution - take care to ensure these functions are as smooth as you can. I have seen some very jumpy behavior in pressure fields at changes of slope of these mesh motion functions, and could not get discrete data motion to work at all. I generally work with things moving in water, which is not very forgiving in terms of pressure fields, but have seen it to some extent in all mesh motion problems I have tried.

Best of luck,

-Alex