Hello all,

I have just started research in the field, where there is a system of bodies moving in the fluid on the computational mesh. The movement of the bodies is the result of the fluid itself (e.g. pressure exerts force on the surface of the solids, which may be easily calculated by proper integration).

I am just looking for some references (Internet sites and maybe books or papers) for such cases. The problem here is the fact that the bodies move on the computational grid as well, and it requires a special treatment.

So far I have been working in the "normal" CFD, where the boundaries where fixed and have no experience in that. I will be grateful for your help.

Best wishes, Pawel

I am too looking for such sort of information about moving boundaries to model the effect of temperature increase and then the pressure on an ICE piston. So far after contacting a friend he told me that FLUENT supports sliding mesh and strongly recommended that software. So I think you had better drop this question in the fluent discussion board so you might get a fast answer there. GoodLuck

In order to use moving boundaries, one commonly employed technique is the Arbitrary Lagrange Euler (ALE) formulation. In this formulation, the flow equations are written in a frame with an arbitrary grid velocity. If the grid velocity is fixed to 0, the system degenerates in Eulerian formulation, if it is equal to the material velocity, into Lagrangian formulation.

You can get some ideas about ALE formulation on our website www.mcube.fr under the M-Explicit Validation section.

Hope it helps!

Dimitri

Yes, a nice technique and I used to know it (although it was a long time ago) and thanks for reminding me about it. The only problem is the fact that I have a whole system of moving bodies travelling in diffrent directions. Their movement is a result of the fluid surrounding them but also they bounce into each other from time to time, which changes their directions even more drastically.

The problem is the interaction between the fluid and the bodies. I prefer the computational grid to be constant and only update the position of the bodies.

Well, it looks more complicated, doesn't it?

Best wishes, Pawel

I prefer the computational grid to be constant and only update the position of the bodies.

IMMERSED BOUNDARY METHOD Ref: Peskin (lots of publications!!)

Yusof (Works really well for moving solid

boundaries)

Udaykumar et al (Iowa), Choi's group (Korea)

CTR (Stanford) and many more

hope this helps MT

Yes. It is really a lot. Thanks very much!

Best wishes,

Pawel

Here I have one paper you may be interested in:

Peskin AP and Hardin GR, Moving particles through a finite element mesh, J. Res. Natl. Inst. Stand. Technol. 103, 77-91 (1998).

I can sent the e-copy to you if you want.

Tony

Of course. I will take it with pleasure.

My e-mail adress is pkosi@poczta.onet.pl.

Thanks for help

Best wishes, Pawel