CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Moving Solid particle in a flow (http://www.cfd-online.com/Forums/cfx/66099-moving-solid-particle-flow.html)

yigity83 July 6, 2009 08:18

Moving Solid particle in a flow
 
Hi everybody

I m trying to model and analyse the behaviour of a solid particle (in a a7 paper size) in a wind tunnel with inlet velocity 180 km/h.

does anybody make such an analyse? I want to see how solid material move and velocity behaviour of this material..

It would be nice if u have send me any information page or document...

ckleanth July 6, 2009 10:54

check the smoke tutorial

yigity83 July 6, 2009 15:36

hi George

thanks but how can i control the geometry and size, I mean my solid material is just like a coin (d:30mm t:1mm)

What i want to do i place this coin in to the windtunnel so that its big lateral surface be parallel to the inlet flow and then in the second analyse perpendicular to flow to see how it moves under 180kmh inlet velocity..



ckleanth July 6, 2009 16:47

3 ways that i can think of all have its pros and cons
a) it could be possible to use Lagrangian–Eulerian multiphase model and define the particle using shape factors. you need to check whenether you can model the coin shape using shape factors
b) moving mesh... very very messy
c) use the immersed solids in v12 however you will use a beta function and turbulence effects due are not calculated

ghorrocks July 6, 2009 22:23

Hi,

Immersed solids is a fully released feature, not a beta feature. Also it does include turbulence effects but cannot accurately capture detailed boundary layer effects as it is very difficult to use a fine enough mesh to capture these effects.

My first go would be to use the lagrangian particle tracking but that will not be very accurate as you won't be able to take into effect the fact the coin has very different drag forces depending on its orientation.

The immersed body feature will be able to capture these effects but to get an accurate overall aerodynamic force on the coin will require a very fine mesh.

A tricky simulation.

Glenn Horrocks

yigity83 July 7, 2009 06:53

In our university we use Ansys 11 so I cant concentrate on immersed body feature..
I will give a try to lagrangian particle tracking.. But can u tell me how to implement this feature? or any example just to understand the idea

I never used it

yigity83 July 7, 2009 07:50

Ok i found it... Thanks for the help...

michael_owen July 8, 2009 22:21

The relevant feature of R12 is not immersed solids but rather the 6DOF solver, which is in fact in beta.

This problem can be attacked in v11 without resorting to FORTRAN. However, the techniques I developed for this are advanced, and become more complex the more degrees of freedom are allowed. Here is an example that illustrates some of the functionality that you would need:

http://www.youtube.com/watch?v=md8vXAUn7g0

This could be extended to allow translations in addition to rotation about a single axis, but allowing arbitrary rotations is probably not feasible (although I haven't tried to develop a full 6DOF technique for v11; so maybe it can be done).

ckleanth July 9, 2009 02:54

Quote:

Originally Posted by michael_owen (Post 221981)
The relevant feature of R12 is not immersed solids but rather the 6DOF solver, which is in fact in beta.

This problem can be attacked in v11 without resorting to FORTRAN. However, the techniques I developed for this are advanced, and become more complex the more degrees of freedom are allowed. Here is an example that illustrates some of the functionality that you would need:

http://www.youtube.com/watch?v=md8vXAUn7g0

This could be extended to allow translations in addition to rotation about a single axis, but allowing arbitrary rotations is probably not feasible (although I haven't tried to develop a full 6DOF technique for v11; so maybe it can be done).

cool videos mate :D

ghorrocks July 9, 2009 07:04

Hi,

Yes, you are right. The immersed solids is a full release feature in V12 but linking it to a 6DOF solver is beta.

Some nice videos on your youtube site Micheal. The bullet in the tube is especially cool.

Glenn Horrocks

michael_owen July 9, 2009 08:58

RE: that video, with all the threads/questions about the Riemann problem/shock tubes, I'm surprised I haven't seen anyone mention PPM.

ghorrocks July 9, 2009 18:36

PPM? What's that?

mechovator July 14, 2009 08:30

Mechovator
 
PPM means part per million. like sand suspended in water is measured with PPM.

Timon July 14, 2009 08:48

Quote:

Originally Posted by mechovator (Post 222613)
PPM means part per million. like sand suspended in water is measured with PPM.

Not untrue, but somewhat out of context. Piecewise parabolic method might be a better guess here.


All times are GMT -4. The time now is 19:31.