second order time scheme in interFoam

lifeinhand · January 17, 2018, 23:31

Hi,

I was running the interFoam tutorial case "les/nozzleFlow2D". The default setting in ddt was Euler which is a first order scheme. I was trying to use a second order scheme, somehow seems like backward and CrankNicolson are not supported.

When I try backward, the error message is "only Euler and CrankNicolson are supported"

When I use CrankNicolson, the error message is "Sub-cycling is not supported with the CrankNicolson ddt scheme"

any suggestions?
Thanks!

TobM · January 18, 2018, 03:43

Seems like you have to switch nAlphaSubCycles in fvSolution to 1 to switch off subcycling. Consequently you will have to reduce maxAlphaCo.

lifeinhand · January 19, 2018, 03:16

Thanks for the advice! Unfortunately, after I change nAlphaSubCycles to 1, the solution diverges. (With maxAlphaCo reduced to 0.3)

I did a little bit research, finding that this error (the sub-cycling and CrankNicolson) does not exist in OpenFOAM 2.3.1. This is strange to me. Somehow, newer version loses second order ddt capablity......

TobM · January 19, 2018, 03:23

I don't have any experience with second order time for interFoam.
However, just a few ideas:
- you can try to reduce maxAlphaCo=0.1
- reduce the initial time step
- recheck boundary conditions

AliE · January 19, 2018, 04:20

Hello,

Let me put my 2cents. The alpha equation is treated explicitly in interfoam since it is hyperbolic. Sometimes to improve stability you need to split the actual time step in more sub-steps and this is called subcycling. Now, cranck nicholson is implicit by definition and in theory unconditionally stable, so maybe that's why you cannot use subcycles... This is my odea, but maybe someone more expert on of can shed ligth on this

lifeinhand · January 19, 2018, 18:46

Thank you! Will try those.

January 17, 2018, 23:31	second order time scheme in interFoam	#1
lifeinhand New Member Han Li Join Date: Jul 2015 Location: College Station Posts: 16 Rep Power: 11	Hi, I was running the interFoam tutorial case "les/nozzleFlow2D". The default setting in ddt was Euler which is a first order scheme. I was trying to use a second order scheme, somehow seems like backward and CrankNicolson are not supported. When I try backward, the error message is "only Euler and CrankNicolson are supported" When I use CrankNicolson, the error message is "Sub-cycling is not supported with the CrankNicolson ddt scheme" any suggestions? Thanks!

January 19, 2018, 04:20		#5
AliE Senior Member Join Date: Dec 2017 Posts: 153 Rep Power: 8	Hello, Let me put my 2cents. The alpha equation is treated explicitly in interfoam since it is hyperbolic. Sometimes to improve stability you need to split the actual time step in more sub-steps and this is called subcycling. Now, cranck nicholson is implicit by definition and in theory unconditionally stable, so maybe that's why you cannot use subcycles... This is my odea, but maybe someone more expert on of can shed ligth on this M.W.G. likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
bash script for pseudo-parallel usage of reconstructPar	kwardle	OpenFOAM Post-Processing	42	May 8, 2024 00:17
Extrusion with OpenFoam problem No. Iterations 0	Lord Kelvin	OpenFOAM Running, Solving & CFD	8	March 28, 2016 12:08
Floating point exception error	lpz_michele	OpenFOAM Running, Solving & CFD	53	October 19, 2015 03:50
Moving mesh	Niklas Wikstrom (Wikstrom)	OpenFOAM Running, Solving & CFD	122	June 15, 2014 07:20
Could anybody help me see this error and give help	liugx212	OpenFOAM Running, Solving & CFD	3	January 4, 2006 19:07

January 18, 2018, 03:43		#2
TobM Member Join Date: Sep 2014 Location: Germany Posts: 88 Rep Power: 12	Seems like you have to switch nAlphaSubCycles in fvSolution to 1 to switch off subcycling. Consequently you will have to reduce maxAlphaCo.

January 19, 2018, 03:16		#3
lifeinhand New Member Han Li Join Date: Jul 2015 Location: College Station Posts: 16 Rep Power: 11	Thanks for the advice! Unfortunately, after I change nAlphaSubCycles to 1, the solution diverges. (With maxAlphaCo reduced to 0.3) I did a little bit research, finding that this error (the sub-cycling and CrankNicolson) does not exist in OpenFOAM 2.3.1. This is strange to me. Somehow, newer version loses second order ddt capablity......

January 19, 2018, 03:23		#4
TobM Member Join Date: Sep 2014 Location: Germany Posts: 88 Rep Power: 12	I don't have any experience with second order time for interFoam. However, just a few ideas: - you can try to reduce maxAlphaCo=0.1 - reduce the initial time step - recheck boundary conditions

January 19, 2018, 18:46		#6
lifeinhand New Member Han Li Join Date: Jul 2015 Location: College Station Posts: 16 Rep Power: 11	Thank you! Will try those.