Particle injection DPM incomplete?

ferayedh · November 16, 2014, 12:55

Hi,

I'm simulating a discrete phase flow in a cilindrical domain of 1m length and .3m diameter.
Air flows through the cilinder and at the beginning of the cilinder there should be a dispersion of solid particles.
However, if i put the origin of the injection cone at 0,0,0 all particles are incomplete. Same if it is at 0.1 0.2 or 0.3m into the tube. Only from 0.5m in the tube, can the injection be completed, with all particles escaping through the outlet.

I have a Hexa-dominant mesh w. resp. to the curves and pressure in-and-outlet, both at 0 gauge pressure.
Even if i specify a velocity inlet, the particles are incomplete.

Help anyone?

pakk · November 17, 2014, 04:24

Quote:

Originally Posted by ferayedh

Hi,

I'm simulating a discrete phase flow in a cilindrical domain of 1m length and .3m diameter.
Air flows through the cilinder and at the beginning of the cilinder there should be a dispersion of solid particles.
However, if i put the origin of the injection cone at 0,0,0 all particles are incomplete. Same if it is at 0.1 0.2 or 0.3m into the tube. Only from 0.5m in the tube, can the injection be completed, with all particles escaping through the outlet.

I have a Hexa-dominant mesh w. resp. to the curves and pressure in-and-outlet, both at 0 gauge pressure.
Even if i specify a velocity inlet, the particles are incomplete.

Help anyone?

There is a limit for particle-time steps in Fluent. By default it is 500. If you see many incomplete paths, it means that 500 time steps is not enough.

You can increase the number. Go to "models", "discrete phase", and change the number 500 to (for example) 5000. (This is from memory, it might be slightly different.)

ferayedh · November 17, 2014, 09:29

Thanks pakk!

I know I have tried that before, one time it worked and another time it didn't.
I'll try it again.
Could the problem also lie in the 0 gauge pressure? I mean, could it be that the pressure at the inlet is too low for the particles to move further in the flow?

pakk · November 17, 2014, 09:51

It could, but then it is not a computational problem but a physical problem.

You don't have to guess about this, you can just check. Plot a few particle trajectories, colored by velocity, and look what the velocity at the end of the trajectory is. Is it significantly larger than zero, than you should increase the number of iteration steps. If the final velocity is zero, then your particle is just stuck there. Increasing the number of iterations will not change anything; your simulation worked perfectly and the outcome is that particles do not leave the simulation box.

ferayedh · November 19, 2014, 08:10

Thanks, it seemed to work.

My model does, however, not show the expected type of pressure distribution
(contours). The same goes for temperature etc.
The velocity of the airflow is correct.
In the next step i tried turning on the energy equation and the radiation discrete ordinates method.
NO heat flux at all. No heat transfer. I don't understand, because an earlier model i made (lost in hard drive crash) showed acceptable heat transfer and pressure distribution!

In another, almost identical, model i made, i used a tetra-dominant mesh and here it seems to work fine.
The differences in pressure, velocity, temperature etc. can be distinguished easily.

November 16, 2014, 12:55	Particle injection DPM incomplete?	#1
ferayedh New Member Ferayed Hok Join Date: Nov 2014 Posts: 3 Rep Power: 12	Hi, I'm simulating a discrete phase flow in a cilindrical domain of 1m length and .3m diameter. Air flows through the cilinder and at the beginning of the cilinder there should be a dispersion of solid particles. However, if i put the origin of the injection cone at 0,0,0 all particles are incomplete. Same if it is at 0.1 0.2 or 0.3m into the tube. Only from 0.5m in the tube, can the injection be completed, with all particles escaping through the outlet. I have a Hexa-dominant mesh w. resp. to the curves and pressure in-and-outlet, both at 0 gauge pressure. Even if i specify a velocity inlet, the particles are incomplete. Help anyone?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[swak4Foam] GroovyBC the dynamic cousin of funkySetFields that lives on the suburb of the mesh	gschaider	OpenFOAM Community Contributions	300	October 29, 2014 19:00
injection problem	Mark New	FLUENT	0	August 4, 2013 02:30
how to get infos of DPM -- incomplete particle	tommyan	Fluent UDF and Scheme Programming	2	March 15, 2013 09:44
DPM UDF particle position using the macro P_POS(p)[i]	dm2747	FLUENT	0	April 17, 2009 02:29
Particle injection in DPM	Zenith	FLUENT	1	March 17, 2007 05:33

November 17, 2014, 09:29		#3
ferayedh New Member Ferayed Hok Join Date: Nov 2014 Posts: 3 Rep Power: 12	Thanks pakk! I know I have tried that before, one time it worked and another time it didn't. I'll try it again. Could the problem also lie in the 0 gauge pressure? I mean, could it be that the pressure at the inlet is too low for the particles to move further in the flow?

November 17, 2014, 09:51		#4
pakk Senior Member Join Date: Nov 2013 Posts: 1,965 Rep Power: 27	It could, but then it is not a computational problem but a physical problem. You don't have to guess about this, you can just check. Plot a few particle trajectories, colored by velocity, and look what the velocity at the end of the trajectory is. Is it significantly larger than zero, than you should increase the number of iteration steps. If the final velocity is zero, then your particle is just stuck there. Increasing the number of iterations will not change anything; your simulation worked perfectly and the outcome is that particles do not leave the simulation box.

November 19, 2014, 08:10		#5
ferayedh New Member Ferayed Hok Join Date: Nov 2014 Posts: 3 Rep Power: 12	Thanks, it seemed to work. My model does, however, not show the expected type of pressure distribution (contours). The same goes for temperature etc. The velocity of the airflow is correct. In the next step i tried turning on the energy equation and the radiation discrete ordinates method. NO heat flux at all. No heat transfer. I don't understand, because an earlier model i made (lost in hard drive crash) showed acceptable heat transfer and pressure distribution! In another, almost identical, model i made, i used a tetra-dominant mesh and here it seems to work fine. The differences in pressure, velocity, temperature etc. can be distinguished easily.