I've read quite a few papers on ALE "arbitrary Lagrangian Eulerian" formulation and I'm trying to modify my FVM solver to work in a moving object.

I have an idea but I'm not sure if it is correct. These are the steps:

1. use the momentum eqn in ALE form, which involves using the relative convective velocity. no change to the poisson eqn

2. calculate the grid velocity following GCL

3. solve the NS eqns

4. deform the mesh to new time step

5. rezoning ie do a remeshing if mesh quality is poor

6. remapping ie transfer old mesh cell center values to new mesh thru interpolation.

7. repeat from step 2

Are these the general steps? I hoped I have not missed out anything. Tks!

Dear ALE Newbie,

Your idea is right and you do not seem to have missed out the steps. But in each step you must be quite careful while implementing the ALE form, specifically while replacing the fluid velocity with the relative velocity. Also, note that the BCs (esp. on the wall) should be treated properly.

Regards,

Ganesh

Hello. I'm also trying to modify my FVM solver to work in a moving object, now i'm thinking about the method ,can you give me some advices? or we can change ideal with each other!

Dear ptaotao,

I could send some papers, a good concise paper would be of Venkatakrishnan and Mavriplis. The broad steps are same as Newbie has pointed out. I could send some useful papers if you could send me your mail id.

Regards,

Ganesh

Hi Ganesh,

i'm interested in the papers. Can you please send me too? my email is zonexo@gmail.com.

Thanks alot!

Dear Ganesh,

That's exactly what I am looking for. it is very much appreciated.I think those papers will help me a lot,My mail is kkndttt@sina.com .

Thanks very much

Sincerely, Ptaotao

It seems several people are interested in this, including myself.

Last Spring I was looking at time advancement using a twist on finite elements in time, and in time and space, using Hermite elements. I know that finite difference in time is the standard approach, but this F.E. approach, going one time step at a time, can generate a number of standard integration methods/formulae, as well as some novel methods. The obvious extension was to problems with moving boundaries. The obvious problem is incompressible fluid flow in an elastic duct/hose/... . The application is to blood flow.

I am not familiar with ALE beyond the name, so that has been on my "to do" list to learn. I don't know where to get started in researching the literature. References would be helpful to me too.

Pls send it to me as well at quarkz@126.com.

Thank you very much. Regards