|
[Sponsors] |
[cfMesh] tichkness Transition top of boundary layer |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
September 17, 2018, 13:18 |
tichkness Transition top of boundary layer
|
#1 |
New Member
FredPieri
Join Date: Mar 2018
Posts: 16
Rep Power: 8 |
Hello ,
Using cfMesh , I wondering how can I smooth the transition in thickness between the top boundary layer cell and the first cell of the mesh ? I mean it looks like my boundary layer is always one mesh size thick , and it is divide by the number of layer I ask. So teh transistion is size is big. Please find attach a picture , with 5 and 15 layers. Overall thickness looks the same. Reading the user Guide I tried to play with the parameters: featureSizeFactor and relThicknessTol . But without sucess. I found this thread, but it doesn't give me the answer. boundary layers What tricks do you use? Thanks in advance. Special thanks to franjo_j ans the CfMesh developers. Fred ps: Here my boundarylayer MeshDict: Code:
boundaryLayers { // untangleLayers 0; patchBoundaryLayers { w_wall { nLayers 1; thicknessRation 1.1; allowDiscontinuity 1; } } optimiseLayer 1; optimisationParameters { nSmoothNormals 5; relThicknessTol 0.15; featureSizeFactor 0.4; reCalculateNormals 1; maxNumIterations 5; } } Last edited by Fredvol; September 17, 2018 at 13:19. Reason: cleaning |
|
September 17, 2018, 14:19 |
|
#2 |
Senior Member
Join Date: Mar 2014
Posts: 112
Rep Power: 12 |
I guess the source of your problem is a wrong keyword.
Try "thicknessRatio" instead of "thicknessRation" Then you can play with values... |
|
September 18, 2018, 06:09 |
|
#3 |
New Member
FredPieri
Join Date: Mar 2018
Posts: 16
Rep Power: 8 |
Hi mzzmrt
Thanks for your reply and the typo mistake. unfortunatly , I continue to have a problem. I think it is geometric problem. If I use this naming ( see overview.jpg): The position on each layer can be found using a geometric progression: https://en.wikipedia.org/wiki/Geometric_progression I use the mesh ( without BL) size as référence unit. If I want 10 layers with an expension ratio to 1.2. the last boudarylayer is only 0.198 thick. Which make a big gap with the next cell. If I accept the last boundary cell size to be at least 0.6 thick. I only have 2 choices: - nb layer = 101 expension ration = 2.5 ! - expension ratio = 1.2 but only 2 layers. Please see the other attachement. am I missing something? Is there a way to spread more the boundary layer toward the mesh? Thanks. Fred |
|
September 18, 2018, 07:22 |
|
#4 |
Senior Member
Join Date: Mar 2014
Posts: 112
Rep Power: 12 |
Well, I am not an expert for the cfMesh and have limited experience with it.
But you can also use the "maxFirstLayerThickness" keyword for declaring it according to your yPlus needs and also try to increase nLayers... A sample: w_wall { nLayers 20; maxFirstLayerThickness 0.002; // use a sutiable size for the geometry here thicknessRatio 1.2; allowDiscontinuity 1; } |
|
September 24, 2018, 16:54 |
|
#6 |
New Member
FredPieri
Join Date: Mar 2018
Posts: 16
Rep Power: 8 |
here an other thread I found having this questions:
2015: version 1.1 - Total boundary layer thickness |
|
October 18, 2018, 04:57 |
|
#7 |
Senior Member
KHB
Join Date: Aug 2010
Location: Singapore
Posts: 118
Rep Power: 16 |
It seems CFMesh is coded so that the total boundary layer is always around 1 cell height (I hope I am wrong though).
But if the gradient in that area is not high, this should be acceptable I guess? maxFirstLayerThickness is to set the maximum first layer thickness and it seems it will not give what you want.. |
|
October 19, 2018, 10:03 |
|
#8 | |
New Member
FredPieri
Join Date: Mar 2018
Posts: 16
Rep Power: 8 |
Hello Lava12005
Yes I contacted CfMesh on this topic , here their answer: Quote:
As it looks very promising, but with this way of doing the boundary layer, I had to swith back to snappyHexMesh. Regards fred |
||
October 21, 2018, 10:22 |
|
#9 |
Senior Member
KHB
Join Date: Aug 2010
Location: Singapore
Posts: 118
Rep Power: 16 |
Ow, thanks for the information..
Yes CFMesh is a nice tool to have, hope they improve the open-source version.. |
|
Tags |
boundary layer, cfmesh, tickness |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
y+ = 1 boundary layer mesh with snappyHexMesh | Arzed23 | OpenFOAM Running, Solving & CFD | 6 | November 23, 2022 16:15 |
Wrong flow in ratating domain problem | Sanyo | CFX | 17 | August 15, 2015 07:20 |
Boundary Layer roughness/ low reynolds wall treatment | Luigi_ | STAR-CCM+ | 1 | March 14, 2012 09:40 |
Boundary Layer Transition | Tom A | FLUENT | 2 | January 21, 2012 02:12 |
errors | Fahad | Main CFD Forum | 0 | March 23, 2004 14:20 |