[morne]

Hi Guys
I'm looking at modelling a burning surface in a combustion chamber(solid rocket motor) and I was wondering if anyone has any tips ideas or suggestions on the best way to go about it. The analysis will involve a moving mesh, the moving burning surface which regresses at a rate which is dependent on the pressure at the surface and more specifically the pressure at a particular cell face. The amount of mass added (the combustion products) is dependent on the volume/mass of the propellant burnt during the last time step.
The analysis is transient with compressible gases and the flow out of the combustion chamber is choked so sonic conditions occur at the throat.

As you can see it's quite a big ask and I'm not even sure how much of it is realistically possible to simulate. Modelling the combustion process in combination with the variable boundary conditions, moving mesh, and sonic flow, may be too much and it's likely I'll have to compromised and simplify to get a good result. But I plan to tackle each of these elements individually then try and tie it all together.

So I put the question to the forum for any comments or suggestions.

What solvers may be useful, RhopsonicFoam modified with combustion elements and moving mesh? Do I need a turbulence model?

Would I be able to move the individual points on the burning surface accurately with an unstructured grid or do I need to have structured hexes.

In general I'd just like to get an idea of what I'm up against, an opinion from some of the experts as to how much of a challenge this would be?

Thanks in advance for any comments

Morné

[hartinger]

this sounds interesting, would be a damn good phd, or two, if you actually get reasonable results

[pierre]

Three PhDs. And reasonable results are way too overated :-)

[hjasak]

Heya,

I don't want to comment on combustion, I'm sure you know what you're doing at that side. Variable boundary mesh, sonic flow and moving mesh are no problem - all standard part of FOAM.

As for the automatic moving mesh solver, this is "just as made" for you: the solver takes the motion of the boundary vertices and the boundary condition and calculates the motion of internal vertices for meshes without causing overlaps. The surface will move "pretty slowly" in terms of simulation and all is well. For an example of what is possible, have a look at the PhD from dr. Zeljko Tukovic (unfortunately, still only in Croatian, but the picces speak a thousand words) :-) He did free surface flows with mesh deformation, which is very much your kind of mesh motion problem.

You don't need structured, hexes, the motion solver works on arbitrary polyhedra with no trouble. You may have some issues with mesh generation, but this is not strictly within OpenFOAM remit and should not be too hard - I suspect the initial configuration of the solid fuel mayl be a bit tricky to mesh.

Could you keep us posted on you progress - I'd really like to see the kinds of stuff you plan on doing (I can offer limited help in return for some pictures and similar).


Good luck,

Hrv

[morne]

Thanks for the comments guys I really appreciate it.

I'm still a relative newbie to FOAM and CFD in general, but I'll keep you updated as I progress. It's encouraging to hear that you guys think the simulation is possible.

Thanks and I'll keep in touch.
M

[PicklER]

Hi Morne

I am working on almost the exact same project. Have tried almost everything, moving mesh, layer addition, combination of mesh motions, but have not figured it out yet.

Did you make progress (finish it)?

Looking forward to hear from you
Pickler