[snappyHexMesh] distorting internal mesh on a pipe when snapping

MetEng · December 2, 2020, 12:49

Hello,

I am trying to run an MPPIC simulation on a pipe but have ran into an issue that is new to me when trying to generate the mesh. My block mesh is roughly square and when i run snappy the first pass before snapping the mesh after removing external cells is still nicely orthogonal inside. however after it snaps the mesh to the surface it is creating highly distorted areas inside of the mesh (surface mesh is still good).

I have tried messing with a bunch of the mesh quality setting and increasing the surface refinement on my geometry but can't seem to get it to work.

Has anyone ran into this before/know how to fix it?

Thanks,

AtoHM · December 3, 2020, 01:42

Looks like the typical Paraview issue. How does it look if you check the option "crinkle slice/clip"? Or does checkMesh point out bad quality cells?

MetEng · December 10, 2020, 20:56

It still seems a little strange as it pops it out in the normal direction but I managed to get my problem to solve my reducing my time steps.

Thanks

B

AtoHM · December 11, 2020, 01:35

Uhm, from your post I am not sure if it is actually solved now. Did it help to switch the options in Paraview? I don't understand what want to say about the time steps. Let's make sure future readers of this thread will find a solution here. If it took you anything else to solve, it would be great if you posted the steps here.

MetEng · December 11, 2020, 09:27

Ahh.. Yes it might help to have some context.

The solution I was running was a lagrangian particle flow. I was using implicit methods to apply properties to the particles and had my max courant # set greater than 1 which was making my solution unstable and it kept failing to converge. I lowered the max courant # to 0.8 and despite the potential mesh quality issue the solution was able to converge without issue.

I am not sure that it solved the problem applying the crinkle to the slice. When I did it made the odd pattern in the mesh go away (for the most part) but it made the slice rough normal to the plane of the slice to do so. in all of the other meshes I have made so far I have not had to apply the crinkle to it to get generally orthogonal slices of the cells so I am still not convinced that there isn't an issue with the mesh.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[snappyHexMesh] snappyHexMesh does not create any mesh except one for the reference cell	Arman_N	OpenFOAM Meshing & Mesh Conversion	1	May 20, 2019 17:16
Too diffusive result on Delaunay type mesh for pipe	AlmostSurelyRob	OpenFOAM Running, Solving & CFD	7	October 8, 2018 13:34
[snappyHexMesh] Snapping not enough (even with surfaceFeatureExtract)	ironclad	OpenFOAM Meshing & Mesh Conversion	1	November 8, 2013 02:58
[GAMBIT] 3D mesh for pipe intersection	pranab_jha	ANSYS Meshing & Geometry	3	May 28, 2010 01:33
Convergence moving mesh	lr103476	OpenFOAM Running, Solving & CFD	30	November 19, 2007 14:09

December 3, 2020, 01:42		#2
AtoHM Senior Member M Join Date: Dec 2017 Posts: 642 Rep Power: 12	Looks like the typical Paraview issue. How does it look if you check the option "crinkle slice/clip"? Or does checkMesh point out bad quality cells?

December 10, 2020, 20:56		#3
MetEng New Member Join Date: Dec 2020 Posts: 15 Rep Power: 5	It still seems a little strange as it pops it out in the normal direction but I managed to get my problem to solve my reducing my time steps. Thanks B

December 11, 2020, 01:35		#4
AtoHM Senior Member M Join Date: Dec 2017 Posts: 642 Rep Power: 12	Uhm, from your post I am not sure if it is actually solved now. Did it help to switch the options in Paraview? I don't understand what want to say about the time steps. Let's make sure future readers of this thread will find a solution here. If it took you anything else to solve, it would be great if you posted the steps here.

December 11, 2020, 09:27		#5
MetEng New Member Join Date: Dec 2020 Posts: 15 Rep Power: 5	Ahh.. Yes it might help to have some context. The solution I was running was a lagrangian particle flow. I was using implicit methods to apply properties to the particles and had my max courant # set greater than 1 which was making my solution unstable and it kept failing to converge. I lowered the max courant # to 0.8 and despite the potential mesh quality issue the solution was able to converge without issue. I am not sure that it solved the problem applying the crinkle to the slice. When I did it made the odd pattern in the mesh go away (for the most part) but it made the slice rough normal to the plane of the slice to do so. in all of the other meshes I have made so far I have not had to apply the crinkle to it to get generally orthogonal slices of the cells so I am still not convinced that there isn't an issue with the mesh.