Large source term in species equation

MACFD · January 3, 2011, 18:59

Hi ,
I am trying to add a large source term in species equation. I have two species one has a mass fraction of 3e-5 (particle phase). Diffusivity of particles is really samll (D=1e-8 m^2/s).

Fluent does not converge when my source term is really large compared to diffusion term. I tried to gradually add the source term as 1/50 of the term, then 1/40,... It converges up to 1/4 of the source term and does not work afterward.

Could anybody help me resolve this issue. My guess is that divergence happens due to upwinding. my source term is not defined for upwinding. Any though is appreciated.

Thanks,
MA

ComputerGuy · January 3, 2011, 22:39

MA,

Have you tried lowering the under-relaxation factors for the species mass fractions, or perhaps lowered your time step size? Decrease the time step by an order of magnitude and see if you can't get it to work.

ComputerGuy

MACFD · January 4, 2011, 02:23

COMPUTERGUY,

Thanks for your prompt reply. My case is steady state.
I lowered the under-relaxation factor and it converged. I should check the results though to see if they are rational.
Your suggestion was great. Thanks a lot.

MA

ComputerGuy · January 4, 2011, 08:58

Definitely check the results. Any time you lower the URF, the residuals will drop. Thus, you can occasionally "force" the residuals lower than your convergence criteria. To be on the safe side, you can set the residual criteria to some very low value, then monitor the residuals to see that they plateau (or, monitor a value of interest and wait until it stabilizes).

ComputerGuy

MACFD · January 4, 2011, 15:16

I lowered URF value for the species. As you mentioned, it lowered the residuals and I defined a really small value for convergence criterion.
Unfortunately, the results are not rational. Center line values look fine but near wall I get really high concentration values (particles) that do not seem correct.
I wonder if you have any other suggestions. Thanks a lot,

MA

January 3, 2011, 18:59	Large source term in species equation	#1
MACFD New Member Join Date: Jan 2011 Posts: 3 Rep Power: 15	Hi , I am trying to add a large source term in species equation. I have two species one has a mass fraction of 3e-5 (particle phase). Diffusivity of particles is really samll (D=1e-8 m^2/s). Fluent does not converge when my source term is really large compared to diffusion term. I tried to gradually add the source term as 1/50 of the term, then 1/40,... It converges up to 1/4 of the source term and does not work afterward. Could anybody help me resolve this issue. My guess is that divergence happens due to upwinding. my source term is not defined for upwinding. Any though is appreciated. Thanks, MA

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem of SOURCE term gradient in UDS	wind	Fluent UDF and Scheme Programming	6	December 1, 2022 15:21
momentum source term	zwdi	FLUENT	14	June 27, 2017 16:40
Reaction Source Term in Steady Species transport equation	cfdvenkatesh	Main CFD Forum	0	August 19, 2010 10:28
OpenFOAM on MinGW crosscompiler hosted on Linux	allenzhao	OpenFOAM Installation	127	January 30, 2009 20:08
species source term reaction	Andrew	FLUENT	3	May 27, 2005 13:35

January 3, 2011, 22:39		#2
ComputerGuy Senior Member Real Name :) Join Date: Jan 2010 Location: United States Posts: 192 Rep Power: 16	MA, Have you tried lowering the under-relaxation factors for the species mass fractions, or perhaps lowered your time step size? Decrease the time step by an order of magnitude and see if you can't get it to work. ComputerGuy

January 4, 2011, 02:23		#3
MACFD New Member Join Date: Jan 2011 Posts: 3 Rep Power: 15	COMPUTERGUY, Thanks for your prompt reply. My case is steady state. I lowered the under-relaxation factor and it converged. I should check the results though to see if they are rational. Your suggestion was great. Thanks a lot. MA

January 4, 2011, 08:58		#4
ComputerGuy Senior Member Real Name :) Join Date: Jan 2010 Location: United States Posts: 192 Rep Power: 16	Definitely check the results. Any time you lower the URF, the residuals will drop. Thus, you can occasionally "force" the residuals lower than your convergence criteria. To be on the safe side, you can set the residual criteria to some very low value, then monitor the residuals to see that they plateau (or, monitor a value of interest and wait until it stabilizes). ComputerGuy

January 4, 2011, 15:16		#5
MACFD New Member Join Date: Jan 2011 Posts: 3 Rep Power: 15	I lowered URF value for the species. As you mentioned, it lowered the residuals and I defined a really small value for convergence criterion. Unfortunately, the results are not rational. Center line values look fine but near wall I get really high concentration values (particles) that do not seem correct. I wonder if you have any other suggestions. Thanks a lot, MA