Grid Independence Study, how to start?

Clau.77 · August 28, 2021, 06:53

Hey guys,

for my steadyState multiRegion case I would like to perform a grid convergence study. The case is a flow around an object.

I have never done that before, so I am a bit confused on how to start. I was thinking about first doing a coarse grid and then two finer grids. The question is: what about adding layers near the wall?
Am I first doing a study without any layers, find the best mesh and then do another study to find the right amount of layers??

Tobermory · August 30, 2021, 06:49

You are right to be cautious here - essentially you have two different grid effects that you are discussing here:
1. Wall resolution of the boundary layer(s); this is dependent primarily on the cell y+ values, but also potentially their shape (are you running RANS, LES, ... ?).
2. Grid resolution of the flow (and scalar) gradients in the main body of the flow.

The nature of your test case (aerofoil flow? bluff body separation? mixing vessel? etc etc) will determine the relative importance of the above two effects. Have a think about what is dominant in your flow, and make sure you test these independently (eg keep wall y+ fixed while you vary the mesh size in the interior of the flow, or vice versa), since otherwise it will get very confusing for you. Good luck!

August 28, 2021, 06:53	Grid Independence Study, how to start?	#1
Clau.77 Member Claudia Join Date: Mar 2021 Posts: 43 Rep Power: 5	Hey guys, for my steadyState multiRegion case I would like to perform a grid convergence study. The case is a flow around an object. I have never done that before, so I am a bit confused on how to start. I was thinking about first doing a coarse grid and then two finer grids. The question is: what about adding layers near the wall? Am I first doing a study without any layers, find the best mesh and then do another study to find the right amount of layers??

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[ANSYS Meshing] How to: Grid independence study for structured mesh	Dronzer	ANSYS Meshing & Geometry	8	March 31, 2018 23:37
[ICEM] Mesh Independence study	Carlos Modesto	ANSYS Meshing & Geometry	4	July 7, 2017 06:57
Grid independence study problem	Bisht	STAR-CCM+	8	February 2, 2016 13:28
Grid Independence of Boundary layer	Luigi_	Main CFD Forum	0	December 14, 2011 13:42
A doubt on grid independence study	G.Balakrishnan	FLUENT	4	November 21, 2000 11:05

August 30, 2021, 06:49		#2
Tobermory Senior Member Join Date: Apr 2020 Location: UK Posts: 668 Rep Power: 14	You are right to be cautious here - essentially you have two different grid effects that you are discussing here: 1. Wall resolution of the boundary layer(s); this is dependent primarily on the cell y+ values, but also potentially their shape (are you running RANS, LES, ... ?). 2. Grid resolution of the flow (and scalar) gradients in the main body of the flow. The nature of your test case (aerofoil flow? bluff body separation? mixing vessel? etc etc) will determine the relative importance of the above two effects. Have a think about what is dominant in your flow, and make sure you test these independently (eg keep wall y+ fixed while you vary the mesh size in the interior of the flow, or vice versa), since otherwise it will get very confusing for you. Good luck!