DPM model, droplet size distribution in atomizer model

davidx · November 12, 2010, 00:36

Dear All,

I am new in DPM modeling.
Currently I try to use the pressure swirl atomizer model to simulate some spray characteristics results to compare with the experiment results of a particular nozzle, such as the droplet size distribution, etc.

I used the same conditions of the experiments in the simulation and found the trends of the droplet size distribution is very similar for the experiments and simulation results, however, the droplet size in simulation is much smaller than our experiment results.

One thing I am really confused is that, Can this model be used for my case? I mean to roughly predict the spray characteristics of a certain nozzle. I read both the User and Theory guide, seems there is no such comments on this part, but there are people use this model to simulate the droplet size in the references of Fluent.

Can anyone help to clarify this part? Thanks.

Regards.

Allan Walsh · November 12, 2010, 13:34

What do you mean by the "trends" predicted by simulation are similar to experimentation? Things like larger nozzle, more coarse spray? Larger velocity, finer spray?
I would have to believe the actual drop diameters will be an important part of your model. If the initial diameters are not representative, then the modeling results will be in question. In creating sprays for input to Fluent, I have always created my own drop velocity and diameter distribtution information and read that into the injection set-up as a separate file. This is quite time consuming but gives more control.
In your case, you could always "fool" the nozzle set-up to give an appropriate diameter distibution by arbitrarily changing the fluid viscosity or nozzle diameter or whatever is needed. Good luck.

davidx · November 13, 2010, 03:43

Allan,Thank you very much.

I also read the DPM chapter from the ANSYS FLUENT 12.0, Theory Guide and User Guider, which mention that "Once the nozzle state is determined,the exit velocity is calculated, and appropriate correlations for spray angle and initial droplet size distribution are determined." From this point, can I say that if I determine all the so-call global parameters in the atomizer model (let's say the flow rate, pressure, spray angle), this model will generate the initial velocity, droplet distribution according to these global parameters by some empirical correlations. After that, you turn on the spray model (breakup and collision model), then go to simulate the atomization process?
From the references of DPM model I read, people do use this model to predict the droplet size and compare with their experimental results, which agree very well.

I believe it is really good idea as you suggested. I will give it a try when I am more familiar with DPM model. Have you got the reasonable results when you create your own information?

Thanks once again.

Sankaranand · May 5, 2017, 04:18

Hi David,
I'm trying my luck here. It is 7 years after your post. This is my question..

I am working on pressure swirl atomizers now and I get Overall mean D32 values larger than the experimental ones on a research paper. Could you tell me how you modeled the spray ? The pressure swirl atomizer model in Fluent doesn't ask you anything about droplet distribution or initial droplet diameter. Just the injection liquid parameters. I following the ANSYS fluent evaporating liquid tutorial now.

November 12, 2010, 00:36	DPM model, droplet size distribution in atomizer model	#1
davidx New Member Jim Join Date: Oct 2010 Posts: 2 Rep Power: 0	Dear All, I am new in DPM modeling. Currently I try to use the pressure swirl atomizer model to simulate some spray characteristics results to compare with the experiment results of a particular nozzle, such as the droplet size distribution, etc. I used the same conditions of the experiments in the simulation and found the trends of the droplet size distribution is very similar for the experiments and simulation results, however, the droplet size in simulation is much smaller than our experiment results. One thing I am really confused is that, Can this model be used for my case? I mean to roughly predict the spray characteristics of a certain nozzle. I read both the User and Theory guide, seems there is no such comments on this part, but there are people use this model to simulate the droplet size in the references of Fluent. Can anyone help to clarify this part? Thanks. Regards.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
critical error during installation of openfoam	Fabio88	OpenFOAM Installation	21	June 2, 2010 04:01
DPM particle size distribution in Fluent	Pete	FLUENT	5	July 7, 2008 13:22
DPM injection with RR distribution	dan	FLUENT	2	November 1, 2005 10:29
Particle Size Distribution with R-R fails?	HP	FLUENT	6	May 19, 2005 07:35
CFD model to be used for feed distributor	R.Sripriya	Main CFD Forum	0	November 19, 2002 02:04

November 12, 2010, 13:34		#2
Allan Walsh Member Allan Walsh Join Date: Mar 2009 Posts: 56 Rep Power: 17	What do you mean by the "trends" predicted by simulation are similar to experimentation? Things like larger nozzle, more coarse spray? Larger velocity, finer spray? I would have to believe the actual drop diameters will be an important part of your model. If the initial diameters are not representative, then the modeling results will be in question. In creating sprays for input to Fluent, I have always created my own drop velocity and diameter distribtution information and read that into the injection set-up as a separate file. This is quite time consuming but gives more control. In your case, you could always "fool" the nozzle set-up to give an appropriate diameter distibution by arbitrarily changing the fluid viscosity or nozzle diameter or whatever is needed. Good luck.

November 13, 2010, 03:43		#3
davidx New Member Jim Join Date: Oct 2010 Posts: 2 Rep Power: 0	Allan,Thank you very much. I also read the DPM chapter from the ANSYS FLUENT 12.0, Theory Guide and User Guider, which mention that "Once the nozzle state is determined,the exit velocity is calculated, and appropriate correlations for spray angle and initial droplet size distribution are determined." From this point, can I say that if I determine all the so-call global parameters in the atomizer model (let's say the flow rate, pressure, spray angle), this model will generate the initial velocity, droplet distribution according to these global parameters by some empirical correlations. After that, you turn on the spray model (breakup and collision model), then go to simulate the atomization process? From the references of DPM model I read, people do use this model to predict the droplet size and compare with their experimental results, which agree very well. I believe it is really good idea as you suggested. I will give it a try when I am more familiar with DPM model. Have you got the reasonable results when you create your own information? Thanks once again.

May 5, 2017, 04:18		#4
Sankaranand New Member Join Date: Oct 2016 Posts: 6 Rep Power: 10	Hi David, I'm trying my luck here. It is 7 years after your post. This is my question.. I am working on pressure swirl atomizers now and I get Overall mean D32 values larger than the experimental ones on a research paper. Could you tell me how you modeled the spray ? The pressure swirl atomizer model in Fluent doesn't ask you anything about droplet distribution or initial droplet diameter. Just the injection liquid parameters. I following the ANSYS fluent evaporating liquid tutorial now.