Convergence problem

lalupp · July 20, 2010, 06:31

I am developing a 3d code with SST turbulence model (incompressible,unsteady) which is built based on OPENFVM. My problem is that my k and omega equations are giving convergence for the first time step but diverges in the next iteration onwards (segregated approach).

Is it because of lack of proper updation procedure or
if some modification in the linear equation especially in the source term is necessary for the implementation of wall boundary condition
I rechecked my k and omega equation several times but not giving
My problem is flow past a sphere

but please help
Thanks in advance

artemiss1984 · July 21, 2010, 18:01

Dear friend your statement is general.
It can occur because of several reasons.

What is is your initialization for k and omega?
What is flow Reynolds number? Probably your simulation has vortex shedding process.
The SST turbulence model uses k-w near the wall and k-e far from the wall.what is your methodology for matching this two turbulence models?
be more specific for your problem.

QUOTE=lalupp;268153]I am developing a 3d code with SST turbulence model (incompressible,unsteady) which is built based on OPENFVM. My problem is that my k and omega equations are giving convergence for the first time step but diverges in the next iteration onwards (segregated approach).

Is it because of lack of proper updation procedure or
if some modification in the linear equation especially in the source term is necessary for the implementation of wall boundary condition
I rechecked my k and omega equation several times but not giving
My problem is flow past a sphere

but please help
Thanks in advance[/QUOTE]

artemiss1984 · July 21, 2010, 18:02

Dear friend your statement is general.
It can occurs because of several reasons.

What is is your initialization for k and omega?
What is flow Reynolds number? Probably your simulation has vortex shedding process.
The SST turbulence model uses k-w near the wall and k-e far from the wall.what is your methodology for matching this two turbulence models?
tell more detail about your problem.

QUOTE=lalupp;268153]I am developing a 3d code with SST turbulence model (incompressible,unsteady) which is built based on OPENFVM. My problem is that my k and omega equations are giving convergence for the first time step but diverges in the next iteration onwards (segregated approach).

Is it because of lack of proper updation procedure or
if some modification in the linear equation especially in the source term is necessary for the implementation of wall boundary condition
I rechecked my k and omega equation several times but not giving
My problem is flow past a sphere

but please help
Thanks in advance[/QUOTE]

lalupp · July 23, 2010, 01:33

Thanks
I got my k and omega equation from the following link
http://turbmodels.larc.nasa.gov/sst.html

For near wall , omega was assigned a large value (equation below) and k was set to zero. I used a blending function to blend k-w and k-epsilon equation as per Nasa site (which depends on the distance to wall) . Other boundary conditions are inlet , outlet etc

I checked my code several times but not giving convergence
Re ~= 1e6 , vortex shedding present
Thanks

artemiss1984 · July 23, 2010, 07:33

Dear friend.

my first suggestion is about your numerical simulation if your code is well prepared.

1.check the CFL condition (the relation between time step and grid cell sizes)
2. what is the properties of fluid? at given Re number use of air with density of 1kg/m3 help to better convergence rather than water with density of 1000 kg/m3.
3.if you are using iterative solution check for the under relaxation values.Don't select high values like usual commercial software(s).I doubt the reality of this great values in commercial software(s).

July 20, 2010, 06:31	Convergence problem	#1
lalupp Member lalupp Join Date: Jul 2010 Location: India Posts: 47 Rep Power: 16	I am developing a 3d code with SST turbulence model (incompressible,unsteady) which is built based on OPENFVM. My problem is that my k and omega equations are giving convergence for the first time step but diverges in the next iteration onwards (segregated approach). Is it because of lack of proper updation procedure or if some modification in the linear equation especially in the source term is necessary for the implementation of wall boundary condition I rechecked my k and omega equation several times but not giving My problem is flow past a sphere but please help Thanks in advance

July 23, 2010, 01:33	Re Convergence problem	#4
lalupp Member lalupp Join Date: Jul 2010 Location: India Posts: 47 Rep Power: 16	Thanks I got my k and omega equation from the following link http://turbmodels.larc.nasa.gov/sst.html For near wall , omega was assigned a large value (equation below) and k was set to zero. I used a blending function to blend k-w and k-epsilon equation as per Nasa site (which depends on the distance to wall) . Other boundary conditions are inlet , outlet etc I checked my code several times but not giving convergence Re ~= 1e6 , vortex shedding present Thanks

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
convergence problem when use pisoFoam, LES for wind tunnel case	Forrest_Lei	OpenFOAM	3	July 19, 2011 07:00
convergence problem	commonyue	Main CFD Forum	1	December 1, 2009 04:54
Convergence of CFX field in FSI analysis	nasdak	CFX	2	June 29, 2009 02:17
3D Fluid Flow Convergence problem	Emily	FLUENT	2	March 21, 2007 23:18
Non Convergence of 3D Heat transfer cfd problem	Balraj	Main CFD Forum	3	December 9, 2004 01:24

July 21, 2010, 18:01		#2
artemiss1984 Member Join Date: Apr 2009 Posts: 38 Rep Power: 17	Dear friend your statement is general. It can occur because of several reasons. What is is your initialization for k and omega? What is flow Reynolds number? Probably your simulation has vortex shedding process. The SST turbulence model uses k-w near the wall and k-e far from the wall.what is your methodology for matching this two turbulence models? be more specific for your problem. QUOTE=lalupp;268153]I am developing a 3d code with SST turbulence model (incompressible,unsteady) which is built based on OPENFVM. My problem is that my k and omega equations are giving convergence for the first time step but diverges in the next iteration onwards (segregated approach). Is it because of lack of proper updation procedure or if some modification in the linear equation especially in the source term is necessary for the implementation of wall boundary condition I rechecked my k and omega equation several times but not giving My problem is flow past a sphere but please help Thanks in advance[/QUOTE]

July 21, 2010, 18:02		#3
artemiss1984 Member Join Date: Apr 2009 Posts: 38 Rep Power: 17	Dear friend your statement is general. It can occurs because of several reasons. What is is your initialization for k and omega? What is flow Reynolds number? Probably your simulation has vortex shedding process. The SST turbulence model uses k-w near the wall and k-e far from the wall.what is your methodology for matching this two turbulence models? tell more detail about your problem. QUOTE=lalupp;268153]I am developing a 3d code with SST turbulence model (incompressible,unsteady) which is built based on OPENFVM. My problem is that my k and omega equations are giving convergence for the first time step but diverges in the next iteration onwards (segregated approach). Is it because of lack of proper updation procedure or if some modification in the linear equation especially in the source term is necessary for the implementation of wall boundary condition I rechecked my k and omega equation several times but not giving My problem is flow past a sphere but please help Thanks in advance[/QUOTE]

July 23, 2010, 07:33		#5
artemiss1984 Member Join Date: Apr 2009 Posts: 38 Rep Power: 17	Dear friend. my first suggestion is about your numerical simulation if your code is well prepared. 1.check the CFL condition (the relation between time step and grid cell sizes) 2. what is the properties of fluid? at given Re number use of air with density of 1kg/m3 help to better convergence rather than water with density of 1000 kg/m3. 3.if you are using iterative solution check for the under relaxation values.Don't select high values like usual commercial software(s).I doubt the reality of this great values in commercial software(s).