Mesh of Square Cylinder

Yeats · May 28, 2015, 09:14

Hello everyone!
I am dealing with vortex shedding effects in a transonic flow regime around a square cylinder. For a start my study shall remain in 2D.
In order to discretise the domain, I applied the Directed Mesher (PBM). I understand that the same meshing approach has been applied to these kind of geometries before, so I didn't expect any issues in that regard.
However, once I started the simulation and observed the velocity field, it'd become apparent that the mesh has a big influence on the simulation result. It seems that regions with small cell size somehow act as a boundary, which yields unphysical results.
I have attached some Screenshots of the mesh and one of the velocity field, where the mesh influence is clearly visible.
Does anyone have a clue why this problem could occur?Any help or hint is greatly appreciated

Capture du 2015-05-28 15:03:30.png

Capture du 2015-05-28 15:02:49.png

Capture du 2015-05-28 15:13:15.jpg

flotus1 · May 28, 2015, 11:55

Looks like a classic example for a solution that is not converged.
If you are using an implicit solver, run more iterations / iterations per time step.

A better meshing approach would be an o-grid around the block.
This way you could avoid small / high aspect ratio cells in the free stream region that cause the poor convergence behavior

May 28, 2015, 09:14	Mesh of Square Cylinder	#1
Yeats New Member Join Date: May 2015 Location: Austria Posts: 4 Rep Power: 10	Hello everyone! I am dealing with vortex shedding effects in a transonic flow regime around a square cylinder. For a start my study shall remain in 2D. In order to discretise the domain, I applied the Directed Mesher (PBM). I understand that the same meshing approach has been applied to these kind of geometries before, so I didn't expect any issues in that regard. However, once I started the simulation and observed the velocity field, it'd become apparent that the mesh has a big influence on the simulation result. It seems that regions with small cell size somehow act as a boundary, which yields unphysical results. I have attached some Screenshots of the mesh and one of the velocity field, where the mesh influence is clearly visible. Does anyone have a clue why this problem could occur?Any help or hint is greatly appreciated Capture du 2015-05-28 15:03:30.png Capture du 2015-05-28 15:02:49.png Capture du 2015-05-28 15:13:15.jpg

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[blockMesh] 3-D Mesh in a cylinder	Nikunj.R	OpenFOAM Meshing & Mesh Conversion	32	July 9, 2023 13:22
[Gmsh] GMSH: Structured mesh of a square inside a circle	wc3speler	OpenFOAM Meshing & Mesh Conversion	3	September 18, 2019 16:13
[snappyHexMesh] SnappyHexMesh for internal Flow	vishwa	OpenFOAM Meshing & Mesh Conversion	24	June 27, 2016 08:54
[blockMesh] mesh generation inside/around 3D cylinder	Cloudy1988	OpenFOAM Meshing & Mesh Conversion	10	August 23, 2012 14:13
[snappyHexMesh] Layers:problem with curvature	giulio.topazio	OpenFOAM Meshing & Mesh Conversion	10	August 22, 2012 09:03

May 28, 2015, 11:55		#2
flotus1 Super Moderator Alex Join Date: Jun 2012 Location: Germany Posts: 3,399 Rep Power: 46	Looks like a classic example for a solution that is not converged. If you are using an implicit solver, run more iterations / iterations per time step. A better meshing approach would be an o-grid around the block. This way you could avoid small / high aspect ratio cells in the free stream region that cause the poor convergence behavior