[li siye]

Dear Foamers

Now I'm going to do a simulation about the transient flow field of a rocket launching, after studying different tutorials, i think the following grid methods can be used.

1. dynamic mesh

2. Overset grid

3. Adaptive mesh

Because i have done the case of flow around a cylinder moving up and down before, so i plan to use the dynamic mesh method to simulate the flow, but in the previous cylindrical dynamic mesh example's 0/pointDisplacement, the boundary condition of the cylinder is as follows:


cylinder
{
type oscillatingDisplacement;
omega 0.5;
amplitude (0 1 0);
value uniform (0 0 0);
}

And the constant/dynamicMeshDict is

FoamFile
{
version 2.0;
format ascii;
class dictionary;
location "constant";
object dynamicMeshDict;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

//dynamicFvMesh staticFvMesh;
dynamicFvMesh dynamicMotionSolverFvMesh;

motionSolverLibs ("libfvMotionSolvers.so");
solver displacementLaplacian;

displacementLaplacianCoeffs
{
diffusivity inverseDistance (cylinder);
}



It is a basic tutorial about dynamic mesh, and i want to set the conditions for a rocket to move upward at a certain acceleration.However, I am not familiar with the mesh motion parameter setting of pointDisplacement in the 0 file. At present, OpenFoam prompts me that the boundary type of pointDisplacement is as follows

{angularOscillatingDisplacement
angularOscillatingVelocity
calculated
codedFixedValue
cyclic
cyclicACMI
cyclicAMI
cyclicRepeatAMI
cyclicSlip
empty
fixedNormalSlip
fixedValue
nonuniformTransformCyclic
oscillatingDisplacement
oscillatingVelocity
processor
processorCyclic
slip
solidBodyMotionDisplacement
surfaceDisplacement
surfaceSlipDisplacement
symmetry
symmetryPlane
timeVaryingMappedFixedValue
timeVaryingUniformFixedValue
uniformFixedValue
uniformInterpolatedDisplacement
value
waveDisplacement
wedge
zeroGradient}

Since I am not familiar with these boundary types, I choose the codedFixedValue or solidBodyMotionDisplacement boundary conditions. How can I set up a rocket to move up at a certain acceleration? (I've searched GitHub for examples of dynamic mesh, but most of them are rotation cases using overset grid. There's no one like me that moves in a straight line with a certain acceleration)

Thank you very much for some suggestions， help and relevant resources!

If any Foamers who have any ideas about overset grid and adaptive grid, welcome to put forward some solutions. Thank you! (I saw a case of rocket launching simulation with CONVERGE using adaptive grid on YouTube https://www.youtube.com/watch?v=JYGJbhRHAzU )

[Saleh Abuhanieh]

Hi,


To mange such large mesh deformation in the standard release of OpenFOAM, you have only the overset. It will not be very easy and you may need collar meshes to manage the interacted walls at the beginning of the simulation. Check the OpenFOAM overset tutorials on YouTube.

As far I know, there is no active library in OpenFOAM that can handle the motion with refinement.


If your flow is incompressible (I don't think so), the foam-extend fork has more option for dynamic meshes which allows topological changes.



Hope that was useful


Saleh

[li siye]

Thank you very much Saleh!

I have used overset grid in OF-v2006, and I also found it in foam-extend 4.1.

The flow i want to simulate is compressible flow, and i plan to add the density (rou) into the incompressible solver.

I have two questions now:
1.what the difference of the overset grid between OF-v2006 and foam-extend 4.1?
2.if i just add the rou into the incompressible solver, are there some problems that i should take care of?

Thanks a lot! Best wishes to you.

li siye

[Saleh Abuhanieh]

Hi,


The overset implementations in both forks are totally different. According to what have been posted in this forum and my own observations, the foam-extend is faster. However, in foam-extend, there is no standard compressible solver with overset capability.


Converting an incompressible solver to a compressible one involves more complicated tasks than just adding only the density field. Unless you know very well the theoretical background and the implementation details, I don't recommend you to try.


Regards,
Saleh

[li siye]

Thanks a lot for your valuable advice Saleh!

According to your advice, maybe i could just only simulate the situation with OF-v2006.

And there must be one day that i should know the implementation details, and now i just can change some codes in xxxFoam.C (or .H) dictionary and wmake them. Are there some sources or that can help me know the implementation details, i really want to know them!

Thanks a lots again! Best wishes to you!