CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   I confused between DPM and Multiphase model. (

Wonjin November 14, 2008 06:10

I confused between DPM and Multiphase model.
Hello, everyone.

I'm a graduate student using the Fluent6.1 to simulate mixing performance in micro-scaled PDMS chips.

Two inlets and one outlet exist in the chips. One inlet is composed of water+nano gold particles and the other is water+proteins. I'm gonna assume the proteins as spherical nano particles.

Which modeling method is proper to simulate that conditions. I wanna know about particles tracking and mass fraction(or phase fraction) of each solid component.

Thanks for your concern.

KJ Larsen November 14, 2008 12:21

Re: I confused between DPM and Multiphase model.
In my understanding the difference between multiphase and discrete phase models is that the multiphase models (VOF, mixture and Eulerian) are intended for multiple continuous phases, solid phases can be treated by the solver as a liquid phase inside another continous phase(not by VOF), most calculations are solved for the mixture and individual phases interpolated. Generally the Eulerian handles granular solid phases the best.

DPM models particles as a point of mass (with associated negligible volume compared to continous phase) requiring nominally a particle volume fraction of less than 12%. Individual particles are accelerated by forces based on interaction with the continous phase, such as bouyancy and drag. If you want it can also exert forces back on to the continous phase. DPM particle concentration (mixture density ignoring fluid mass) can be obtained anywhere in the profile in the same way as pressure etc.

DPM is probably the most accurate if appropriate and used properly but also take the greatest time.

Multiphase gives pathlines, PDM particles tracks. I think I'm right in saying pathlines are like stream lines that massless particles have followed, whereas particle tracks are the actual position histories as calculated for individual particles with inertia.

However, I'm no nanotechnologist but a quick web search suggested atomic diameters are 0.1-0.5nm and I believe all the models employed by fluent require the assumption that the discrete particles are sufficiently larger than the fluid molecules for the fluid to be considered continuous. So I don't know if any of the models will be appropriate for nano scale / protein particles.

Are these not effectively dissolved in water?

The User Guide gives good summary of applications of each model.

Hope this was helpful. Kris

Wonjin November 15, 2008 03:20

Re: I confused between DPM and Multiphase model.
Thanks for your kind response.

The particles' diameters of bio-sample and Au NPs are 3~5 nano-meter and 20~50 nano-meter, respectively. As you told me, those values are negligible compared to continous phase(water or buffer). The particles are physically mixed and float through the micro-fluid stream in micro-channel.

In that case, is it clear to use DPM model?? I already tested with using two models, DPM and Multiphse.

In the Multiphase model, each phase fraction could be expressed by contour plot and pathline also could be discribed. However, solid materials(Au particles) can be selected as a second phase ?

In the DPM, the solid particles could be selected as an inert. And I could get the particle tracking results and pathline stuff. However, I could not get the phase fracton distribution of each phase. DPM cannot originally discribe the distribution with contour plot ??

Ultimately, I wanna express the mixing performance quantitatively with numerical plot.

I think my question makes you confused, cause I'm a beginner in Fluent. Please understand and help me.

KJ Larsen November 15, 2008 12:58

Re: I confused between DPM and Multiphase model.
Hello Wonjin

I'm afraid I'm also fairly new and don't know about the physics at the scale you are working on.

Just to clarify: The total volume of all the particles needs to be below 12% of the domain volume. My worry about the individual particle sizes was that they may be too close to that of water molecules (3nm ~=30 times smallest atomic diameter and water has 3 atoms anyway) so I don't know whether water could then be considered a continuum relative to such small particles.

Do you treat the protein and water as a liquid solution then? I think you could set Au as a granular phase (but you have to define the material as a fluid to use it in the mixture/Eulerian models) or perhaps gold laden water would be an acceptable phase. You should be able to choose mixture or an individual phase when selecting most display options.

In the DPM individual particles are not influenced by other particles (except indirectly through particle intraction with the continuous phase), so separate discrete phases are oblivious to each other. Therefore it would make no difference to your model (with two-way coupling disabled) if only one of your discrete phases were injected at a time, rather than simultaneously. I think DPM concentration is calculated from the mass of particles from all discrete phases so no simple way to get mixing information about a ternary system.

Have you tried tutorials 19&20 for multiphase? The prior shows fluid mixing and the latter the granular phase.

The applicability of the standard models to nanoscale I can't be sure of however.

Hope that was helpful, might want to wait for a second opinion. Kris

Wonjin November 15, 2008 16:12

Re: I confused between DPM and Multiphase model.
Thank you very much, KJ Larsen !

Your valuable answers help me progessing my research.

Arrangement of my questions,

- The total volume of all the particles are below 12% of the domain volume. ( I agree to DPM case )

- If the size of solid particles is extremely small as water molecules, there is no problem to consider the solid particles as the liquid phase ? ( Multiphase model is right?)

- Above two questions, creteria influencing model selection are overlapping. Are there lots of differences btw DPM results and Multiphase modelresults about same case as my case.

It's 5 AM in my country, South Korea. In my questions, many of wrong sentence structures may exist due to my sleepy state. Please understand me ^^;. A few hours later, I'll check and modify my model with your worthy answers.

Thank you and Thank you.

KJ Larsen November 17, 2008 07:43

Re: I confused between DPM and Multiphase model.
To be honest, I can't answer.

You need to find out how the particles behave in water to know if the models are appropriate, or have some experimental results with which to compare your model results, otherwise you cannot know whether you're predicting a real problem. The standard spherical drag law may not even be realistic as the discrete water molecules may be unable to exert a continuous distribution of force on the particles.

You can get significant disparities in qualitative results from different sub-models (random walk etc.) so I imagine Multiphase and DPM will be different.

Goodluck. Kris

Amir1 April 17, 2014 09:36

Hi Guys,

I am working on a study for particle-surface entrapment. I would like to know more about DPM vs DEM.
As you said in DPM, partciles are considered as dots (no volume) and then they interact with the continious phase. What about DEM?

I know both of them use Lagrangian method. DPM for whole particle but DEM for evry single particle. What is the main difference between DPM abd DEM in terms of considering particle VOLUME ?

sarp February 26, 2015 16:32

Fluent Multiphase Flows Euler Lagrange & Euler Euler Methods
A gas-solid regime multiphase flow and both euler-euler and euler-langragian methods could be used for this kind of analysis.
Partcile trajectories can be calculated and DPM boundary conditions (trap, reflect, escape, etc) can be used by using DPM method.

My question is using other multiphase methods is it possible to track particles and use dpm Wall boundary conditions (trap, reflect etc:) as disperded pahse model in roder to provide same outputs.

in my opinion, methods except DPM method (lagrange), assume that particle phase is a continua and particle can not be trapped or reflect in order to maintain conservation.

All times are GMT -4. The time now is 12:45.