|
[Sponsors] |
Quads vs Triangles for lid-driven cavity flow |
![]() |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
![]() |
![]() |
#1 |
Member
Giorgio
Join Date: Mar 2023
Posts: 57
Rep Power: 3 ![]() |
Dear Foamers,
I am comparing quads and triangles for a lid driven cavity flow. I am using blockMesh to generate the quadrilateral (hexa) mesh, and Salome for the triangular (prismatic) mesh. The viscosity is 1e-5 and I am using pisoFoam with the k-epsilon model. After running the simulation for 5 seconds, the results are very different: the flow computed with quads is much more developed, see pictures below. QUADS: ![]() TRIA: ![]() I am obviously using the same parameters and schemes for the two simulations. The spacing is delta x = 0.02 for both meshes. Does anybody have any idea what can cause the difference? |
|
![]() |
![]() |
![]() |
![]() |
#2 |
Senior Member
Join Date: Apr 2020
Location: UK
Posts: 792
Rep Power: 14 ![]() |
Yes - with a triangulated mesh, there is much more non-orthogonality to the flow through the mesh faces. You therefore need to adjust your solver settings, as per solving on a tet mesh. In the end, it's often just simplest with OF to avoid tet/triangulated meshes.
|
|
![]() |
![]() |
![]() |
![]() |
#3 |
Member
Giorgio
Join Date: Mar 2023
Posts: 57
Rep Power: 3 ![]() |
Indeed, I thought it was non-orthogonality but then it should at least get better when using orthogonality corrections. And it's not, the solution looks exactly the same with orthogonality corrections. How do you suggest to set up the solver for tets?
|
|
![]() |
![]() |
![]() |
![]() |
#4 |
Senior Member
Join Date: Apr 2020
Location: UK
Posts: 792
Rep Power: 14 ![]() |
As I said earlioer - I avoid using tets in OpenFOAM; I have found that it is just too difficult to get stable and accurate simulations with tet meshes. You can google and find threads like Openfoam is not succesful for tetrahedral meshes, which contain some advice (esp on gradient schemes etc.) BUT be warned - the stabilising emasures add numerical diffusion (ie degrade the solution) and you still might not get good convergence or stability.
Just avoid tris/tets if you can and use quads or polys. |
|
![]() |
![]() |
![]() |
Tags |
comparison, lid driven cavity, quads, salome, tria |
Thread Tools | Search this Thread |
Display Modes | |
|
|
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
Boundary Conditions for FORTRAN Lid Driven Cavity Error | BaklavaBaby29 | Main CFD Forum | 0 | January 23, 2023 07:49 |
Open channel, Gravity driven flow over and through a porous bed | narendrapatel111 | FLUENT | 0 | June 8, 2020 05:14 |
Lid Driven Cavity and Turbulent Flow | Mayzie12 | FLUENT | 1 | November 8, 2017 07:28 |
Limitations on time the step for lid driven cavity flow | ja0335 | Main CFD Forum | 12 | October 21, 2016 20:13 |
Flow over a cavity inconsistent | Sparow89 | FLUENT | 0 | June 20, 2016 13:02 |