CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   2 phase flow (http://www.cfd-online.com/Forums/main/106798-2-phase-flow.html)

 mehdi kamyabi September 9, 2012 03:45

2 phase flow

Dear all
I'm looking for a method to simulate flow of two phase fluids which one phase is dispersed in to another phase (such as fluids filled with nano particles)
I'm looking for flow patterns not fluid equilibrium properties.
among my researchs I found maybe particle methods are useful. if so which one is
better?(theoretically and technically) . I was looking for SPH but now I think
Molecular Dynamic is better, Am I right? is there better choice also?
I will be glad to know any idea from experts
Thanks

 PGodon September 11, 2012 14:04

two-phase flow

You said you are interested in flow patterns, I understand by that you mean you are looking at the particles as "tracers" to follow the motion of the fluid elements ("flow pattern"). SPH is rather coarse and does not very well at low density and low viscosity. Why not solve for a two-phase flow directly? You can use the limit that the particle size and density is such that the particles act like tracers then you just need to follow them. There are really like 3 regimes in two-phase flows (if one looks at this simply), (1) the regime where the particles just follow the fluid elments (tracers), (2) the regime where the particles follow their own path (subject to say gravity or other forces) and are barely affected by the fluid, then there is the intermediate regime (3) where the motion of the particles depart significantly from that of the fluid elements and from that of particles in vacuum. So is your problem dealing with (1)? and if so why not just solve the hydrodynamical equations and use two-phase flow in that limit (tracers) ?

 mehdi kamyabi September 12, 2012 04:26

Thank you PGodon
I should appreciate for this nice information about two phase flows, I agree that SPH can not follow particles to predict flow patterns (particles are just computational dots not as tracers and they are coarse as you told) but it is not my problem because I just want to model two phase flows in order to find their behavior not to use second phase as tracers to detect behavior of first phase.I'm looking for nanofluids which have totally different properties in comparison with fluids withoud nano particles.
maybe it was better if I explain about why I'm looking for mesh-less methods not common two phase simulation methods (such as VOF or etc); there is a reason and that is I'm trying to simulate sedimentation of macro scale solids in such this fluids, so I am facing with solid-liquid interface problem also which can not be handled with common two phase models according to my present knowledge. am I mistaken? therefore I have been encouraged to use mesh-less methods which can handle this problem better.

 PGodon September 12, 2012 09:27

Sedimentation

you say you are trying to simulate the sedimentation of solids in "this fluids". Do you mean you want to simulate the sedimentation of large solids in that two-phase flow made with the nanoparticles?

Am I understanding you correctly? You are looking at large solids submerged in that two phase flow? so that you have actually large particles (or "solids") together with nano particles in the fluid?
Or are you looking into the sedimentation of the nano particles agglomerating into larger solids, something like aglomeration, aggregation of particles into larger particles??

There are plenty of similar problems being solved in Astrophysics in the modeling of the formation of the planets, where microparticles agglomerate into chondrules and then into cm-size particles and into m-size particles to eventually be subjet to a gravitational instability forming km-size objects. The early modeling (of smal particles) involves the sedimentation of these particles due to gravity (gravitational sedimentation, also named "differentiation" especially when referring to two gases like H + He separating under the influence of gravity) to form larger particles; which sounds very similar to your problem (that is, if I understood you corrrectly).

Could be more explicit as to what problem exactly you are trying to solve? What is the setting and so ? :confused:

 mehdi kamyabi September 15, 2012 05:01

multi scale

first I should appreciate again for your kind endeavoring to understanding and helping me, I should take it clear that you are right in your first guess; I want to simulate nano-fluids which contain solid particles also (which this particles are different from nano particles agglomerates). yes actually I have large solids together with nano particles in the fluid which are different in size scale and material.
suppose I am looking for simulating one solid ball (micrometer diameter to centimeter diameter) falling in a nanofluid medium. in the first stage I am not concerning nano particle agglomeration formation. I think my major problem is multi scaling, I mean I think I am facing with nano scale (which calls MD or else) with macro scale plus solid-liquid interface problem(which calls SPH or esle) together. another concern is technical support of this two methods (I mean softwares and their applicability) .

Thank you

 All times are GMT -4. The time now is 01:02.