Force and Mesh resolution

bjavila1409 · January 4, 2024, 15:14

I am trying to do a mesh resolution study and test if the resolution of my mesh changes the results of my case. To do that, I am computing the forces (using the forces function in OF) on my geometry and plotting over time. I noticed the forces increase with increasing number of grid points. I am not sure if that is really the case (drag force with the highest resolution is of order 400N which seems wrong for basically a foil in 0.6*body length m/s fluid). I am wondering if I should be dividing by some factor like the number of grid points in the geometry.

Also, I would accept any other way to prove the mesh resolution doesn't change the results. It seems the drag force oscillates until a certain time and then converges to about the same value for each mesh resolution, but it takes more time to converge for the larger number of grid points.

For more context, I am running simpleFoam, using k-omega-SST RAS turbulence. I am using OF8.

MarcusAng · March 12, 2024, 10:23

Hi Bailey,

It is not uncommon for forces to fluctuate due to mesh refinement. It is difficult to say if 400N is a lot, therefore I would suggest converting them to coefficients.

If the coefficients continue to change (more than ~5%) due to mesh refinement there is likely something wrong with your mesh or solver. Some things you can check/try,
1. Is the solver converging?
2. Is the boundary conditions and turbulence model settings physical?
3. Is the mesh getting rougher?
4. Try increasing mesh resolution just near the boundary layer.

Marcus

January 4, 2024, 15:14	Force and Mesh resolution	#1
bjavila1409 New Member Bailey Join Date: Jun 2021 Posts: 14 Rep Power: 5	I am trying to do a mesh resolution study and test if the resolution of my mesh changes the results of my case. To do that, I am computing the forces (using the forces function in OF) on my geometry and plotting over time. I noticed the forces increase with increasing number of grid points. I am not sure if that is really the case (drag force with the highest resolution is of order 400N which seems wrong for basically a foil in 0.6*body length m/s fluid). I am wondering if I should be dividing by some factor like the number of grid points in the geometry. Also, I would accept any other way to prove the mesh resolution doesn't change the results. It seems the drag force oscillates until a certain time and then converges to about the same value for each mesh resolution, but it takes more time to converge for the larger number of grid points. For more context, I am running simpleFoam, using k-omega-SST RAS turbulence. I am using OF8.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
parallel run in foam extend	CRI_CFD	OpenFOAM CC Toolkits for Fluid-Structure Interaction	5	March 28, 2024 08:36
Position and Forces in a Transient Problem with a Moving Mesh	astrakhan	FLUENT	3	July 17, 2023 09:22
[ICEM] How to force surface mesh respect the line element of already existing shell mesh	Sina.Li	ANSYS Meshing & Geometry	3	November 16, 2017 12:32
Compute drag force on a mesh (complete beginner!)	matteo.monti	Main CFD Forum	0	October 10, 2014 20:07
[snappyHexMesh] Layers:problem with curvature	giulio.topazio	OpenFOAM Meshing & Mesh Conversion	10	August 22, 2012 09:03

March 12, 2024, 10:23		#2
MarcusAng New Member Marcus Join Date: Jul 2022 Posts: 8 Rep Power: 4	Hi Bailey, It is not uncommon for forces to fluctuate due to mesh refinement. It is difficult to say if 400N is a lot, therefore I would suggest converting them to coefficients. If the coefficients continue to change (more than ~5%) due to mesh refinement there is likely something wrong with your mesh or solver. Some things you can check/try, 1. Is the solver converging? 2. Is the boundary conditions and turbulence model settings physical? 3. Is the mesh getting rougher? 4. Try increasing mesh resolution just near the boundary layer. Marcus