How to understand differences between simpleFoam and icoFoam results for airfoil

Jay_CFD · December 24, 2023, 22:14

Hi, everyone:
I'm an OpenFoam learner; I'm learning how to use OpenFoam to simulate the airfoil test case.

The first solver I adopted is the icoFoam (I learned it from the flow around cylinder example). But the flow becomes quite steady after several iterations (first attachment). Then I found some documents online (https://www.openfoam.com/documentati...irfoil-2d.html), it suggests to use simpleFoam for this test case. I try to adapt my mesh data to the simpleFoam and get results shown in the second attachment. There are some vortexes, which might be the phenomenon as expected. I am wondering why the results look quite steady for the icoFoam results.

I noticed there are some documents about different solvers(https://www.openfoam.com/documentati...andard-solvers), but i'm still not quite understand about these details, are they supposed to solve the same Navier-Stokes equation, why the results are so different?

Here are my OpenFoam configurations for
airfoil icoFoam test case (https://github.com/wangzhezhe/5MCST/...irfoil_icoFoam) and
airfoil simpleFoam test case (https://github.com/wangzhezhe/5MCST/...oil_simplefoam)
Thanks a lot for the help!

anubasu · December 25, 2023, 03:37

check whether you are running a steady state simulation (Without any time-step or deltaT) in any of these solvers. It seems you are running a steady state simulation in one of these cases.

Jay_CFD · December 25, 2023, 20:58

Previously, I use icoFoam to run the flow around cylinder case(https://github.com/wangzhezhe/5MCST/...cular_cylinder), the results looks ok (there are some interesting phenomenons) But I'm not sure why it does not work well for the airfoil case (quite steady and no vortex).

Tobermory · December 29, 2023, 12:19

What turbulence model are you running? Presumably a RANS model, in which case the icoFoam solution (steady) is the correct one, and the simpleFoam solution is junk. Try adjusting your solver settings to get a stable simpleFoam solution.

Note that icoFoam is a transient solver, using the PISO algorithm, and so is stabilised by taking a small enough time step, i.e. suitable Co number. simpleFoam uses the SIMPLE algorithm, and so requires underrelaxation to stabilise.

Ultimately, the two approaches will give the same solution, if run properly.

Jay_CFD · December 29, 2023, 23:19

After some exploration, I found that rotate the airfoil to a certain degree can get some interesting phenomenons. In the first animation(https://drive.google.com/file/d/1bCI...ew?usp=sharing), the left side one is the airfoil with 15 degree, the right side one is 30 degree, they use same icoFoam parameters. For the second animation(https://drive.google.com/file/d/1uOK...ew?usp=sharing), it is the results using simpleFoam with 15 degree. I'm still a little bit curious why there are so different and how to judge if simulation parameters make sense from these output. (why the results are junk numbers). Could you provide some guidance about how to adjust these parameters and what are expected output results. Thanks a lot for the help!

Tobermory · December 31, 2023, 07:18

I think that the best thing I can do here is give you some advice - CFD is just a tool ... and like any tool it needs knowledge and skill to wield effectively. You cannot expect to pick up a tool and generate quality results without knowledge or experience.

So for you, my friend, based on your earlier posts, can I suggest that you do a lot of reading of the basics of CFD & Fluid Mechanics, or take some relevant courses? Then, you will be in a good position to start using the tool.

All the best, and Happy New Year!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
divergence problem with icoFoam or simpleFoam	salicTo	OpenFOAM Running, Solving & CFD	0	July 28, 2023 11:01
icoFoam diverges when simpleFoam converges	leizhao512	OpenFOAM Running, Solving & CFD	2	September 11, 2015 16:02
icoFoam after simpleFoam to check for unsteadiness?	casesam	OpenFOAM Running, Solving & CFD	2	October 12, 2014 13:52
error while solving motorBike with simpleFoam or icoFoam	anjansir	OpenFOAM Running, Solving & CFD	1	December 12, 2011 14:51
Different results with icoFoam and simpleFoam..why???	nileshjrane	OpenFOAM Running, Solving & CFD	11	September 13, 2010 16:08

December 25, 2023, 03:37		#2
anubasu Member Anurag Join Date: Feb 2023 Posts: 78 Rep Power: 3	check whether you are running a steady state simulation (Without any time-step or deltaT) in any of these solvers. It seems you are running a steady state simulation in one of these cases.

December 29, 2023, 12:19		#4
Tobermory Senior Member Join Date: Apr 2020 Location: UK Posts: 745 Rep Power: 14	What turbulence model are you running? Presumably a RANS model, in which case the icoFoam solution (steady) is the correct one, and the simpleFoam solution is junk. Try adjusting your solver settings to get a stable simpleFoam solution. Note that icoFoam is a transient solver, using the PISO algorithm, and so is stabilised by taking a small enough time step, i.e. suitable Co number. simpleFoam uses the SIMPLE algorithm, and so requires underrelaxation to stabilise. Ultimately, the two approaches will give the same solution, if run properly.

December 29, 2023, 23:19		#5
Jay_CFD New Member Jay Wang Join Date: Dec 2023 Posts: 3 Rep Power: 2	After some exploration, I found that rotate the airfoil to a certain degree can get some interesting phenomenons. In the first animation(https://drive.google.com/file/d/1bCI...ew?usp=sharing), the left side one is the airfoil with 15 degree, the right side one is 30 degree, they use same icoFoam parameters. For the second animation(https://drive.google.com/file/d/1uOK...ew?usp=sharing), it is the results using simpleFoam with 15 degree. I'm still a little bit curious why there are so different and how to judge if simulation parameters make sense from these output. (why the results are junk numbers). Could you provide some guidance about how to adjust these parameters and what are expected output results. Thanks a lot for the help!

December 31, 2023, 07:18		#6
Tobermory Senior Member Join Date: Apr 2020 Location: UK Posts: 745 Rep Power: 14	I think that the best thing I can do here is give you some advice - CFD is just a tool ... and like any tool it needs knowledge and skill to wield effectively. You cannot expect to pick up a tool and generate quality results without knowledge or experience. So for you, my friend, based on your earlier posts, can I suggest that you do a lot of reading of the basics of CFD & Fluid Mechanics, or take some relevant courses? Then, you will be in a good position to start using the tool. All the best, and Happy New Year!