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

Restarting case problem - transient introduced

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

Reply
 
LinkBack Thread Tools Display Modes
Old   March 14, 2012, 15:26
Default Restarting case problem - transient introduced
  #1
New Member
 
Patryk Radyjowski
Join Date: Feb 2012
Posts: 2
Rep Power: 0
Legwan is on a distinguished road
Hi,

I'm quite new to CFD community - joined the research group last month. Despite no previous experience I was allocated to CFD part of the project. I've inherited after previous PhD a quite complex reactor model in Fluent 6 using GRI-Mech 2.1 and couple UDF functions.
Reactor is a heat recirculating lean premixed superadiabatic cross-flow laminar design and half of single channel (out of several) is modeled in transient 2D planar with symmetry in the middle. UDF's are mainly used to impose specific temperature distribution at the wall (crucial for operation) and introduce radiation model. I'm still getting up to speed with everything so I don't know to much about the UDF details. So far I've managed to successfuly (?) recompile UDF and run the case in the newest Fluent 14.

I'm writing to ask for Your opinions on one disturbing problem - first part of the task is to run simulation on relatively coarse mesh (+ first order discretization etc) to let it calculate the following time steps quickly until a steady state is achieved. However every single time I'm restarting the case (load last .dat file and continue calculation from this time step) it introduces significant transients. In other words: if Fluent is calculating cycle of 1000 time steps after which I do start it for another 1000 t-s the resulting t-s #1001 shows much higher node temperature fluctuations compared to #1000 than between lets say #1000 and #999 (not only temperature). Furthermore if I decide to calculate 2000 time steps at once there is no such a difference between #1000 and #1001.
This causes a lot of hustle because once the steady state is reached I need to refine the mesh and start it again from current time step - if any transients are introduced then it can take ages for them to damp-up in refined model (note - i'm using full combustion mechanism with 50 species and 277 reactions :/) Surprisingly previous guy hasn't reported anything about this problem - he was working on rich mixture & Fluent 6 though.

Furthermore I'm getting very strange CH3 & C2H6 behavior upstream of the flame - anyone experienced anything like this? It is also connected with restart problem. After starting continuing calculations i get big burst of CH3 filling whole reaction channel upstream of fire interface.

Also any hints on reducing mechanism would be warmly welcomed (PM me)

Thank You very much for help.

PS. Not sure if "restart" is a good description.
Legwan is offline   Reply With Quote

Old   March 15, 2012, 16:30
Default
  #2
New Member
 
Patryk Radyjowski
Join Date: Feb 2012
Posts: 2
Rep Power: 0
Legwan is on a distinguished road
Today I've received help from Ansys and it solved the problem, therefore I feel in need to share solution:

"I assume that you are using stiff chemistry solver. The jump in residual and solution is normal behavior since it started from new ISAT table. But solution should come back to initial situation after some iteration. If this is not the case then your initial solution was not fully converged. There is some useful information at users guide about ISAT table at : // User's Guide :: 2 // 21. Using Chemistry Acceleration // 21.1. Using ISAT.
You can also save the ISAT file while saving the case and data files. While rereading the case and data you can also read the ISAT table."
Legwan is offline   Reply With Quote

Reply

Tags
combustion, continue, fluent, restart, udf

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
convergence problem when use pisoFoam, LES for wind tunnel case Forrest_Lei OpenFOAM 3 July 19, 2011 06:00
Star-CD transient simulation, problem of loss of mass kit STAR-CD 3 April 13, 2010 11:34
Transient case running with a super computer microfin FLUENT 0 March 31, 2009 11:20
Problem of restarting moving meshes with icoDyMFoam rolando OpenFOAM Bugs 8 March 23, 2007 07:18
Free surface boudary conditions with SOLA-VOF Fan Main CFD Forum 10 September 9, 2006 12:24


All times are GMT -4. The time now is 10:10.