CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   steady or unsteady DPM ??? (

Marina July 3, 2006 09:50

steady or unsteady DPM ???
I am trying to simulate a boiler (25 kW) fuel: EL oil, non-premixed combustion, prePDF UNSTEADY DPM pressure swirl atomizer Start time: 0 s Stop time: 1000 s Particle Time Step Size: 0.001 s Number of Time Steps: 1 Stohastic tracking usw. und here is what I get... start time:0 s, stop time: 1000 s, time: 3.0e-03 s, time step: 1.00e-03s number tracked=6, escaped=0, aborted=0, trapped=0, evaporated=0, incomplete=0, shed =0

The dispersed particles should evaporate, but there is no evaporation ! (satuation vapor pressure is defined correctly as a function of temperature)

Please, what am I doing wrong ?? Is it better to use steady DPM for my problem ? steady or unsteady DPM ???

Can any body help me please ?? Very much thanx in advance for any sugestion !

Mathias July 4, 2006 04:43

Re: steady or unsteady DPM ???
My experience with Fluent is limited, but I remember a similar question before. IIRC the evaporated=0 does not say that there is no evaporation. Fluent doesnt track single particles because this would be to expensive. Instead it works with parcels, that are containing a (high number) of particles, each with the same properties I think. Fluent computes evaporating of "particles" in a parcel, but this leads to a shrinking "parcel" mass. As long as one particle remains in a parcel, this specific parcel will not be count as evaporated.

I think you could have evaporation in your case, but you have to use a different approach for verifying it.

Other comments? ;)

Marina July 4, 2006 08:03

Re: steady or unsteady DPM ???
Thank you very much for your reply Mathias ! I know that Fluent works with parcels, and when I use steady DPM model, I have evaporation, although there is always evaporated=0. I wanted to start unsteady DPM model because than I can use BREAKUP Model Wave (We>100) und dynamic drag law, but it didn't work. There is so much parameter that I could change und for which I still don't have feeling.

Particle Time Step size = 0.001 s ??? Number of time steps= 1 ???

Pressure Swirl Atomizer Start time=0s Stop time = 1000 s ???

Also I am not sure about DPM-urf factor 0.25. I am trying to change those parameters, but it still doesn't work !! Obviosly I did some basic mistake in my set parameters and ??? oder ??? My question is: Do I make a big mistake, or what is the effect on my results if I do my simualtion without wave breakup model und dynamic drag model.


Mathias July 4, 2006 08:41

Re: steady or unsteady DPM ???
Well, basicly Im in a similar situation. Im trying to simulate a water spray splashing against a hot plate. Goal is a working transient simulation of the heat transfer that can be compared to experimental data.

Steady with unsteady DPM is working more or less now, at least with very relaxed URFs, a well(?)-refined mesh, long iteration cycles and a well-convergend flow solution for startup. I used k-e-realizable with enhanced wall treatment and dynamic drag/breakup/collision etc. for this, some time ago.

The transient calculation isnt working at all and Im becoming clueless, because I dont know whether Im using wrong parameters (timestep, urfs, ...) or if there is a deeper problem with the whole setup. In my case the spray-wall-interaction does make it more critical, I believe. IIRC my unsteady-DPM runs crashed often when the main part of the spray had reached the wall.

You could try first of all the usual methods, smaller timestep and/or more continuos iterations per DPM iteration and further URF relaxing. A look on the grid may help too, maybe there is need for a refinement in some areas. IIRC a mesh refinement helped in my case at least a bit.

Marina July 4, 2006 10:32

Re: steady or unsteady DPM ???
Unfortunately I can't even start steady with unsteady DPM simulation. Thanks for your advices, I'll keep on trying.

Please tell me, which discret phase boundary condition types did you set for surrounding walls ??

Mathias July 5, 2006 04:52

Re: steady or unsteady DPM ???
In my case I only have a wall boundary at the "target" plate, that is hit by the spray. The other sides are either massflow-inlets or pressure-outlets. They are far away from the spray cone, so I guess that occuring backflows doesnt have to much influence. I used wall-jet for the wall boundary at the plate, but later I will also try wall-film.

All times are GMT -4. The time now is 11:27.