CFD Online Discussion Forums - Something about raindrops simulation by DPM

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- FLUENT (https://www.cfd-online.com/Forums/fluent/)

- - Something about raindrops simulation by DPM (https://www.cfd-online.com/Forums/fluent/66823-something-about-raindrops-simulation-dpm.html)

Something about raindrops simulation by DPM

hi ,everybody.
I meet some troubles about raindrops simulation by DPM.
I want to simulate full-scale raindrops by DPM.I use the surface injection.

the computional domain is 3D and is very large,900*900*3000(height).
So, the number of raidrops in domain is very very large with exceeding 1E+12.

The bounded face was adopted as surface injection.

Now, if I want to simulate this project, the grid faces in bounded face (surface injection) were very very large numbers .And , it must exceeds the capitcy of the Computer .

Is any other way which I can choose to avaoid this problem?
please.

As I suppose, you want to decrease the number of raindrops tracked by DPM analyse.

The number of tracked particles in surface injections depend of the number of cells used to create this surface... something like that ;)
And there is no option to change the numbers.

There are two options...

You can use UDF (I have readen somewhere here about it, but never used it)

or you can use injections from file...
just create a file in notepad. Rename it to blabla.inj
And inside write some lines (you can use Excell to compute this) like:

Code:

(( x y z u v w diameter temperature mass-flow) name )

x,y,z - location

u,v,w - velocities

each in SI units



eg:

((0.01 0.03 -2.1 0 0 -4.5 0.00004 123 0.001) 1)

((0.02 0.05 -1.9 0 0 -4.5 0.00004 123 0.001) 2)

...

...

((0.121 0.123 -67.1 0 0 -4.5 0.00004 123 0.001) 1131)

Check User Guide for more information.

Hope it will be works.
Regards

Thanks for Geisel

your reply give me a lightpoint,especially the second option would fit my problem well.I will try it immediately.

Lastnight,I also thought another way that I will divide the full-scale computational domain into 12 parts by 30 degrees each part ,because the domain is axial symmetrical. I will carry out calculation only on one part.
Then ,the domain is redued and the total drops is also reduced spontaneously.