CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Liquid-Solid Modeling (

safikhani_hamed August 11, 2011 05:54

Liquid-Solid Modeling
Dear friends,
I am going to modeling a simple tube containing water and solid particles using eulerian-eulerian model but unfortunately I do not get good results and I get very bad results , I mean the particles at the end of the tube collect at the top of the tube , how ever that the density of particles are much more than that of water and the particles should collect at the bottom of the tube, I do not know what is the problem, can any body help me ? is there any tutorial about liquid-soid modeling ? I have read the examples of tutorial.

ghorrocks August 11, 2011 07:55

I think there are some tutorials on eular-eular modelling which come with CFX.

If the particles go up not down are you sure you have the gravity vector in the correct direction?

safikhani_hamed August 11, 2011 08:11

yes I have defined the gravity direction correctly.

safikhani_hamed August 11, 2011 11:43

Dear Glenn,

I checked the problem, I found the reason in fact I do not know why the software automatically change the particle material to air, do you know what is the problem ?

ghorrocks August 11, 2011 19:19

You have not specified your materials correctly. Have a look at the CFX tutorials on how to specify materials.

juliom August 16, 2011 08:48

Colleague work with the discret model (lagrange).
good Luck

ghorrocks August 16, 2011 18:27

Why use particle tracking? The eularian approach is a valid way of modelling particle laden flows and for densely packed flows the lagragian approach is not suitable.

Hamed has not said anything which suggests the eularian approach is not suitable.

juliom August 16, 2011 19:17

You are right, but when you have small quantities of a dispresed phase, you should use the disscrete model.
Indeed both approach are valid, but Lagrange model will work better, you can read the FLUENT tutorial, where they explain very well about the stokes number and the other propierties which you have to evaluate before selecting the suitable model. In theory Every model would work, but in the reality just one will give you the results you need.

All times are GMT -4. The time now is 06:42.