CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > CFX

why my result is 100% decided by initialization?

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   June 1, 2011, 23:53
Default why my result is 100% decided by initialization?
  #1
Member
 
Ji Guozhao
Join Date: Jan 2011
Posts: 37
Rep Power: 6
jiguozhao is on a distinguished road
Dear all:

i was trying to simulate a gas mixture(H2/Argon) going through a gas separation module. there is a change in H2 fraction(or Argon fraction) due to the selectivity of the membrane(more H2 is going through the membrane).
but the H2 fraction distribution result is 100% decided by my initial guess in initialization. if i set 0.1 as H2 fraction in initialization, i get 0.9925 at retentate outlet; if i set 0.3 as H2 fraction in initialiazation, i get 0.293072 at retentate; when i set 0.5 initially, i get 0.491635 at retentate; for setting 0.7 initially, i get 0.692509 finally. in all the cases above, the solving can converge to 10e-5. so which one is the really result?

(in experiment i think at the beginning it is transient state, but after 5 minutes or a little more, it reaches steady state. in the simulation i did steady state, one reason is that i just want to have a look at what is going on after 5 minutes, i do not care the first 5 minutes; another reason is my limited capability in cfx only allows me do steady state at the moment.the boundary at inlet is 1E-6kg/s and 0.7 H2 molar fraction, 0.3 molar fraction of Argon. retentate outlet is 500kpa pressure. the module is 1 meter long and about 0.3 meter in diameter.)
is there anyone can help me?

kind regards

Ji Guozhao
g.ji@uq.edu.au

Last edited by jiguozhao; June 3, 2011 at 02:45.
jiguozhao is offline   Reply With Quote

Old   June 2, 2011, 13:23
Default
  #2
New Member
 
A.R. Baserinia
Join Date: Jan 2010
Location: Canada
Posts: 24
Rep Power: 7
baserinia is on a distinguished road
You have not provided enough info; is the simulation steady-state or transient? What are the inlet boundary conditions?
baserinia is offline   Reply With Quote

Old   June 2, 2011, 17:57
Default
  #3
Member
 
Ji Guozhao
Join Date: Jan 2011
Posts: 37
Rep Power: 6
jiguozhao is on a distinguished road
Quote:
Originally Posted by baserinia View Post
You have not provided enough info; is the simulation steady-state or transient? What are the inlet boundary conditions?
thanks a lot for your care about my question!
to answer your first question, in experiment i think at the beginning it is transient state, but after 5 minutes or a little more, it reaches steady state. in the simulation i did steady state, one reason is that i just want to have a look at what is going on after 5 minutes, i do not care the first 5 minutes; another reason is my limited capability in cfx only allows me do steady state at the moment.

for the second question, the boundary at inlet is 1E-6kg/s and 0.7 H2 molar fraction, 0.3 molar fraction of Argon. retentate outlet is 500kpa pressure. the module is 1 meter long and about 0.3 meter in diameter.

thanks again!
jiguozhao is offline   Reply With Quote

Old   June 3, 2011, 09:30
Default
  #4
Senior Member
 
Join Date: Apr 2009
Posts: 511
Rep Power: 12
stumpy is on a distinguished road
Your timestep might be way too small.
stumpy is offline   Reply With Quote

Old   June 3, 2011, 12:06
Default
  #5
New Member
 
A.R. Baserinia
Join Date: Jan 2010
Location: Canada
Posts: 24
Rep Power: 7
baserinia is on a distinguished road
This is indeed interesting! You said it takes five minutes or longer for the actual device to reach steady-state conditions. That means the diffusion time scale is much larger than the momentum time scale. So if you want to get a reasonable steady-state solution, you should use a larger time scale for the mass fraction equation. If you use the same time scale for momentum and mass fraction, the changes would be so small that CFX thinks the solution has converged.

Last edited by baserinia; June 3, 2011 at 17:04.
baserinia is offline   Reply With Quote

Old   June 4, 2011, 07:18
Default
  #6
Member
 
Ji Guozhao
Join Date: Jan 2011
Posts: 37
Rep Power: 6
jiguozhao is on a distinguished road
Quote:
Originally Posted by baserinia View Post
This is indeed interesting! You said it takes five minutes or longer for the actual device to reach steady-state conditions. That means the diffusion time scale is much larger than the momentum time scale. So if you want to get a reasonable steady-state solution, you should use a larger time scale for the mass fraction equation. If you use the same time scale for momentum and mass fraction, the changes would be so small that CFX thinks the solution has converged.
Hi friend:

thank you for your answer!

i just set time scale in solver control, and i can just set one value.
can i set different time scale for different equations? if so, how can i set them separately?

thanks again!

Ji Guozhao
jiguozhao is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Comparison the airfoil 0012 experimental result and simulation result harrislcy FLUENT 30 August 29, 2013 10:27
Questions about Mass Sources jiguozhao CFX 10 May 19, 2012 07:11
Read Result Files or User Defined Result Aquilaris ANSYS 0 February 9, 2011 06:25
FMG initialization query Mohsin FLUENT 4 August 16, 2010 01:21
In a distance above the ground, my LES result is not good panda60 OpenFOAM 2 January 22, 2010 08:52


All times are GMT -4. The time now is 01:51.