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

DPM coupling steps

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

Reply
 
LinkBack Thread Tools Display Modes
Old   June 16, 2005, 22:43
Default DPM coupling steps
  #1
kiao
Guest
 
Posts: n/a
Hi,

Im confused about the steps undertaken for DPM coupling. As I am aware it is: 1. Solve the continuous phase (So I iterate without injections until it converges) 2. Define injections and setup in dpm panel. 3. I set up interaction with continuous phase. Some posts have suggested between 10-40, I usually use 3. Is this correct. And when I go to iterate, Fluent doesn't solve for the injections immediately, it iterates the coninuous phase to the point with which I set it at. When it finally gets to the injection, there is a spike in the residuals.(.. and yes I have problems with converging) So my questions are:

1. Are the steps I undertook correct, I mean I have to keep pressing the iterate button until the injections are introduced and sometimes it doesn't spike and converges immediatly? Is this possible?

2. How do you define the convergence. If you set the interactions with continuous phase high, then the continuous phase simply solves itself until it reaches its initial convergence regardless of where the residuals were at the point when the injections were introduced. Doesn't this mean that you can converge easily if you set the iterations per continuous phase to 1000, therefore wouldn't a lower value of iterations per continuous phase iterations be more accurate?

thanks for reading... i am so confused...

kiao
  Reply With Quote

Old   June 17, 2005, 00:33
Default Re: DPM coupling steps
  #2
Me
Guest
 
Posts: n/a
"1. Solve the continuous phase (So I iterate without injections until it converges) 2. Define injections and setup in dpm panel. 3. I set up interaction with continuous phase. Some posts have suggested between 10-40, I usually use 3. Is this correct. ""

It is correct. contineous phase iterations/DPM iterations cann't be very high. However, upto 10 should be fine. Lower the number, more computional resources. Actualy, 1 is best if you have enough computational resources with your given case and time.

""And when I go to iterate, Fluent doesn't solve for the injections immediately, it iterates the coninuous phase to the point with which I set it at. When it finally gets to the injection, there is a spike in the residuals.(.. and yes I have problems with converging) So my questions are:

1. Are the steps I undertook correct, I mean I have to keep pressing the iterate button until the injections are introduced and sometimes it doesn't spike and converges immediatly? Is this possible? ""

You set Conti/DPM iterations to 3 in DPM panel. It is fine. But I don't understand why you need to keep on pressing iteration button. Probably bcs untill first DPM iteration takes place, Fluent finds yr solution converged. In that case, lower the residue criteria and run your calculation with DPM ON. (Ideally, residue should be zero, however, not possible)

Residue jumps when it attempts first DPM iteration. But after that it should come down and get steady. Probably some fluctuations. Well, that's why you need to find if you hv enough number of particles to get correct solution. By the way, what residue value you can achieve?

""2. How do you define the convergence. If you set the interactions with continuous phase high, then the continuous phase simply solves itself until it reaches its initial convergence regardless of where the residuals were at the point when the injections were introduced. Doesn't this mean that you can converge easily if you set the iterations per continuous phase to 1000, therefore wouldn't a lower value of iterations per continuous phase iterations be more accurate?""

Define covergence is a tricky question to answer. Convergence should be judged based on multiple criterias and residue check should be just one of them. One must also check critical solution parameters if that gives logical values and if that is steady . Check mass balance in Report panel. That should be zero or atleast 0.001.

Interaction with contineous phase cannot be very high.What you understand is correct in this case. Lower value is better. If you set, Conti/DPM is 1000, we can't say the coupled solution is converged. It's just the gas phase.
  Reply With Quote

Old   June 17, 2005, 10:21
Default Re: DPM coupling steps
  #3
Me
Guest
 
Posts: n/a
A correction to my reply. Conti/DPM iteration should be sufficient to allow contineous phase to absorb the change in momentum through drag term due to Discrete phase. During the initial DPM iterations, the jump in residue is higher but eventually it should die down. Make sure you perform sufficient iterations untill coupled solution is selfconsistant.

In my reply, i wrote Conti/DPM should be mini and it should be best if you use 1. But i should be wronge there. Sorry about this.
  Reply With Quote

Old   June 19, 2005, 18:50
Default Re: DPM coupling steps
  #4
kiao
Guest
 
Posts: n/a
Thanks for clarifying for me. I usually set convergence at 0.00001 to solve gas phase. After converging I turn on DPM coupling with conti/dpm iteraition of 3. I have to press iterate 3-4 times as the solution is already converged and I need to introduce the injections. Maybe i could lower the residuals again after the initial convergence. Do you think that is ok?

thanks again
  Reply With Quote

Old   June 20, 2005, 13:34
Default Re: DPM coupling steps
  #5
Me
Guest
 
Posts: n/a
Yes, and keep running solution until you reach to steady residue solution.
  Reply With Quote

Old   June 23, 2005, 11:04
Default Re: DPM coupling steps
  #6
K. Jagus
Guest
 
Posts: n/a
I will reactivate this post since there are some things which are of interest to me. It is said here that iterations per DPM should be as low as possible. I am modelling fuel spray from pressure-swirl atomizer. I have many problems with obtaining results. However I read in fluent tutorail regarding DPM that for sprays number of iterations per DPM should be high, and they put 1000 there. This is rather contrary to what was said here. Now I am counting the case and I wait for those 1000 iterations, only then Fluent will inject particles. Also I am confused about number of particle streams. Since the spray is very fine shouldn,t I put there for example 1000 rather than 60 what tutorial says? Thank You very much for any suggestions.
  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
How to set up transient one way coupling? peiji1984 CFX 2 March 18, 2011 10:44
DPM Scalar Update at specific time steps sega Fluent UDF and Scheme Programming 1 April 27, 2010 23:15
What is weakstrong coupling in FSI problems hajo OpenFOAM Running, Solving & CFD 5 May 15, 2008 01:45
one/two way coupling of DPM Angela FLUENT 3 April 28, 2008 09:29
Two way coupling in DPM fpingqian FLUENT 3 November 28, 2004 23:47


All times are GMT -4. The time now is 04:12.