DPM model problem

wwa · April 21, 2010, 06:07

Hello all,

i have a problem with increasing number of particles tracked after each DPM iteration. Normally the number of particles is around 100000, but later it goes up to several millions and takes more time to calculate. i don't know why this happens

. anybody has experienced similar problem?

i run steady simulation of combustion in 3D, standard k-epsilon, species transport model - EDC concept (with mechanism of 22 species and 49 reactions)

As for DPM, in total there are 36 injections of combusting particles, released from 2 symmetrical surfaces, with rosin-rammler diameter distribution. Each injection has different particle size range and devolatilizing species.

i also use udf for species source terms in one part of the volume, which is porous.

thanks in advance for any comments.

Peter

alikami · May 3, 2010, 05:00

I think for staedy state , when you check particle tracks, It gives history of all particles tracked ...
meaning evry DPM iteration , it tracks particles from defined injection .
At end it gives results of all. so that if your dpm runs n times you have
n*100000 tracks , I hope i am correct.
Regards ali

sydney_tekam · May 3, 2010, 09:13

HI

Do you run couple or segregated solutions?
Are you simulations steady or unsteady for dpm?

wwa · May 3, 2010, 09:50

thanks for replies.
dpm is also steady, and the solver is pressure-based so i guess it's segregated (i use Fluent 6.3).
also, i have to mention that it does happen only in one specific case. there is no problem in other cases with slightly modified mesh and very similar boundary conditions. total number of particles tracked there is always the same.

peter

sydney_tekam · May 3, 2010, 11:16

ok
is the option unsteady tracking activated?

wwa · May 3, 2010, 11:18

no, unsteady tracking is disabled. i have never used it.

sydney_tekam · May 3, 2010, 16:26

You said that you were modeling combustion .
Have you activated droplet break up or coalescense?
Can you post a sample of your TUI report please?

Regards,

wwa · May 4, 2010, 10:26

no, i didn't use droplet break-up as it is related to unsteady tracking.

to be more precise, i'm modelling wood particles combustion.
in total there are 36 injections released from 2 surfaces (18 per each).

or to break it down further there are 3 sets of injections by size (small, normal and large) per 6 devolatilizing species (co, co2, h2, ch4, c2h4 and c2h6). all injections have rosin rammler diameter distribution.

i'm not sure which part of the tui report are you interested exactly.
here is the sample of DPM iteration:

DPM Iteration ....
number tracked = 158040, escaped = 104527, aborted = 0, trapped = 0, evaporated = 0, incomplete = 53513, incomplete_parallel = 0
Parallel particle tracking concurrency: 38%

regards,
peter

April 21, 2010, 06:07	DPM model problem	#1
wwa New Member Join Date: May 2009 Location: France Posts: 14 Rep Power: 17	Hello all, i have a problem with increasing number of particles tracked after each DPM iteration. Normally the number of particles is around 100000, but later it goes up to several millions and takes more time to calculate. i don't know why this happens . anybody has experienced similar problem? i run steady simulation of combustion in 3D, standard k-epsilon, species transport model - EDC concept (with mechanism of 22 species and 49 reactions) As for DPM, in total there are 36 injections of combusting particles, released from 2 symmetrical surfaces, with rosin-rammler diameter distribution. Each injection has different particle size range and devolatilizing species. i also use udf for species source terms in one part of the volume, which is porous. thanks in advance for any comments. Peter

May 3, 2010, 05:00	Particle tracks	#2
alikami Member ali Join Date: May 2009 Posts: 58 Rep Power: 17	I think for staedy state , when you check particle tracks, It gives history of all particles tracked ... meaning evry DPM iteration , it tracks particles from defined injection . At end it gives results of all. so that if your dpm runs n times you have n*100000 tracks , I hope i am correct. Regards ali

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
DPM mass problem	Henrik Ström	FLUENT	9	May 20, 2016 12:00
Problem on EDC model for coal combustion	lei	FLUENT	4	September 3, 2015 10:39
lenght and number of nodes for an orifice meter flow problem - k-eps model	fofo	FLUENT	0	November 15, 2009 04:49
problem about DPM of cyclone separator simulation	jason	FLUENT	4	November 12, 2007 04:15
how to do with the DPM problem?	x.p.lee	FLUENT	1	June 26, 2003 14:07

May 3, 2010, 09:13		#3
sydney_tekam New Member Sydney Miles TEKAM Join Date: Apr 2010 Posts: 7 Rep Power: 16	HI Do you run couple or segregated solutions? Are you simulations steady or unsteady for dpm?

May 3, 2010, 09:50		#4
wwa New Member Join Date: May 2009 Location: France Posts: 14 Rep Power: 17	thanks for replies. dpm is also steady, and the solver is pressure-based so i guess it's segregated (i use Fluent 6.3). also, i have to mention that it does happen only in one specific case. there is no problem in other cases with slightly modified mesh and very similar boundary conditions. total number of particles tracked there is always the same. peter

May 3, 2010, 11:16		#5
sydney_tekam New Member Sydney Miles TEKAM Join Date: Apr 2010 Posts: 7 Rep Power: 16	ok is the option unsteady tracking activated?

May 3, 2010, 11:18		#6
wwa New Member Join Date: May 2009 Location: France Posts: 14 Rep Power: 17	no, unsteady tracking is disabled. i have never used it.

May 3, 2010, 16:26		#7
sydney_tekam New Member Sydney Miles TEKAM Join Date: Apr 2010 Posts: 7 Rep Power: 16	You said that you were modeling combustion . Have you activated droplet break up or coalescense? Can you post a sample of your TUI report please? Regards,

May 4, 2010, 10:26		#8
wwa New Member Join Date: May 2009 Location: France Posts: 14 Rep Power: 17	no, i didn't use droplet break-up as it is related to unsteady tracking. to be more precise, i'm modelling wood particles combustion. in total there are 36 injections released from 2 surfaces (18 per each). or to break it down further there are 3 sets of injections by size (small, normal and large) per 6 devolatilizing species (co, co2, h2, ch4, c2h4 and c2h6). all injections have rosin rammler diameter distribution. i'm not sure which part of the tui report are you interested exactly. here is the sample of DPM iteration: DPM Iteration .... number tracked = 158040, escaped = 104527, aborted = 0, trapped = 0, evaporated = 0, incomplete = 53513, incomplete_parallel = 0 Parallel particle tracking concurrency: 38% regards, peter