search

AndyShor / RD_107

CFD case for simulation of RD107 rocket engine
2 stars 1 forks source link

blockMesh not possible #1

Open jcoelho5 opened 2 years ago

jcoelho5 commented 2 years ago

Hi. First of all, thak you for this incredible tutorial. I was looking for something like this for a long time, trying to use the best combination of BC to achieve something very similar to this case. However, I realised this case does not have the blockMeshDict, which I was also curious about how you did the mesh, and trying o adapt to the case I am working on. I know that the case is ready to run, as the mesh is already created and one should simply ignore the blockMesh command. Nevertheless, if possible, I would love to see the blockMeshDict and check the steps to achieve this mesh. Again, thank you. Best regards, João

EDIT: I just found that notebook has details for it. Nevermind. Thank you for the great work!

AndyShor commented 2 years ago

Hi João, Looks like you found the answer, and i am a bit too late, anyway - the blockmeshdict is located at https://github.com/AndyShor/RD_107/blob/main/model/costant/polyMesh/blockMeshDict Verticies are generated using outputs from https://github.com/AndyShor/RD_107/blob/main/notebooks/rd107_blockMeshDict_generation.ipynb It does not provide fully formatted blockmeshdict file, but I copy-pasted chunks of input from there. As for the boundary conditions - a found the most important to have enough space for plume (mine is still on the shorter side), otherwise some kind of not physical reflection from boundary is likely to happen (shock wave from the nozzle reflects from the boundary and travels backwards). I am still not super happy with the effects of ambient pressure. This is the reason why I introduced radial outlet boundary condition, where ' distance to infinity' is shorter compared to the axial outlet BC where most of the plume goes. This should enforce ambient pressure near the nozzle exit. Even with that I had hard time getting the exactly matched flow at the pressure where it is supposed to happen (RD107 should be matched at ambient pressure of some 0.3-0.4 atm).

Whish you all the best for your project. Hope this example was helpful

jcoelho5 commented 2 years ago

Hi Andy. I know this is not directly related to this issue, but it was the way I thought would be best to ask you this: do you have any basis for choosing those values for the turbulence properties at the inlet (namely the omega - 1; and the k - 1000) ?

Once again, thank you for all your help.

AndyShor commented 2 years ago

Hi João, In short - no, not really. In more detail - being not a fluid dynamics expert myself (by far) I started from reproducing/adopting what I found in other peoples examples. A couple were especially helpful. One is model of Merlin engine - https://github.com/Interfluo/OpenFOAM-Cases-Interfluo/tree/main/OpenFOAM-sonicFoam-2D-Merlin-Nozzle-4-main/OpenFOAM-sonicFoam-2D-Merlin-Nozzle-4-main (there is also a youtube video on it https://www.youtube.com/watch?v=PMZK0IZiRjs). So K value is taken from there. Omega value may not even be even in use since the final version uses kEpsilon model (I played with solvers and models a bit). Again looks like with resolution I worked and in that flow - wall effects, shear and related turbulence are rather minor contributors and not resolved by either model (again - this comes from a non fluid dynamics specialist)