CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   how to inject many particles in fluent? (http://www.cfd-online.com/Forums/fluent/85338-how-inject-many-particles-fluent.html)

elah599 February 23, 2011 11:38

how to inject many particles in fluent?
 
I want to inject nano particles in flow and because of small diameter,I have to inject near 2e+15 particles, but fluent doesnt accept more than 2e+9 particles. does anybody know what should I do?

ral007 February 23, 2011 11:43

Are you looking at interaction between each individual nano-particle?...
You could try to aggregate physical particles into 'numerical' particles and examine the solution that way. e.g. each 'numerical' particle represents, say, 1000 or even 1.e+6 physical particles.

elah599 February 23, 2011 12:03

no,I just want to investigate the effect of presence of nano particles into heat transfer and flow field. how should I change the physical particles into numerical particles? and it doesnt affect the flow properties?
thanks

ral007 February 23, 2011 17:07

I think it all depends on how you interpret the injected particles.
For a given particle mass flux, if I decide to inject 10 particles/sec, this determined how any physical particles it represents, provided that all particles are of the same size. If you have different particle sizes, you'd have to know the distribution (i.e. particles as a function of size).
I am also assuming that gasdynamic flowfield is not affected by the presence of particles, and the drag & heat transfer source terms provide means of particles following gas streams. If two-way coupling is needed, there are interaction source terms that you can model.

elah599 February 25, 2011 05:10

first of all, thanks for your help. all of my particles have a same size. as I found, I have to control the number of particles in flow rate on point properties tab. but I dont know how to implement it. for instance, if I have 1e+9 particles with mass fraction 1% and I want to inject 1000 particles, I have to insert total mass fraction (0.01) in flow rate or I have to divide the number of total particles into injection particles and calculate the mass fraction of them.

xrs333 February 27, 2011 03:12

Hi,
As stated by Ral007, instead of calculating the kinatic behavior of every particle in the in real world, DPM simulates them with numerical particles, and the number of the numerical particles have to be restricted by the computing resource available.
So, when you define the injection, just specify the particle stream, which is the number of particles every time the injection occurs; and then you should give the mass flow (in the point properties tab of FLUENT DPM panel), which should equals to real flow rate of the particles. In such a way, one numerical particle in the model may represent thousands of physical particles, or just a fraction of it.

Amir February 27, 2011 12:12

Quote:

Originally Posted by xrs333 (Post 297156)
Hi,
As stated by Ral007, instead of calculating the kinatic behavior of every particle in the in real world, DPM simulates them with numerical particles, and the number of the numerical particles have to be restricted by the computing resource available.
So, when you define the injection, just specify the particle stream, which is the number of particles every time the injection occurs; and then you should give the mass flow (in the point properties tab of FLUENT DPM panel), which should equals to real flow rate of the particles. In such a way, one numerical particle in the model may represent thousands of physical particles, or just a fraction of it.

hi,
it's true if and only if you had done particle study before. i.e. for example penetration coefficient should be independent of particle streams.

xrs333 February 28, 2011 06:57

Quote:

Originally Posted by Amir (Post 297187)
hi,
it's true if and only if you had done particle study before. i.e. for example penetration coefficient should be independent of particle streams.

Hi,
In DPM, the interaction of a numerical particle with the continuous phase depend on the size, shape, mass and velocity. So you have to give right value for these properties. Remember, a particle in DPM is a representative of physical particles with the same properties.

Amir February 28, 2011 10:07

Quote:

Originally Posted by xrs333 (Post 297264)
Hi,
In DPM, the interaction of a numerical particle with the continuous phase depend on the size, shape, mass and velocity. So you have to give right value for these properties. Remember, a particle in DPM is a representative of physical particles with the same properties.

Hi,
I didn't talk about size,shape,mass or velocity !!!
particle study is a very important stage in DPM, here we obtain No. of particle streams that particle properties don't change with increasing streams any more. by this method you can link between realistic and numeric particle streams.;)

elah599 February 28, 2011 11:33

as I found, we can have 1000 particles in stream that each of them represents for example 1000 other particles. so I have to consider the mass flow of 1000 particles for each particle in the stream.

Amir February 28, 2011 16:23

Quote:

Originally Posted by elah599 (Post 297318)
as I found, we can have 1000 particles in stream that each of them represents for example 1000 other particles. so I have to consider the mass flow of 1000 particles for each particle in the stream.

Hi Iman,
I meant that you should specify No. of streams wisely. e.g. consider a parameter that relate to the particle cloud like penetration coefficient, increase No. of streams until this parameter doesn't change.

elah599 March 1, 2011 05:32

at first, thanks to all my friends. another question. when I use numerical particle, changing the mass flow is enough or I have to change all properties of starting and end point, such as diameter?

Amir March 1, 2011 08:42

you should keep diameter,velocity,... constant. the only parameters you can change are flow rate and stream numbers.

abhijeet bhagat October 8, 2013 08:43

injection
 
can we inject species from 2 point

Goutam December 16, 2013 13:40

I am considering nanoparticle of size diameter = 100 nm = 1e-09 m.
To calculate the total flow rate, I used the formula = density*velocity*area. But when I did the simulation, I didn't find any effect of nanoparticles using DPM.

How to use nanoparticle for DPM?


All times are GMT -4. The time now is 02:58.