CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   ASM-bottom variation (http://www.cfd-online.com/Forums/cfx/95470-asm-bottom-variation.html)

 antonio December 19, 2011 12:50

ASM-bottom variation

Dear all,
I am modeling a density current composed of a mixture of water plus sediments using the algebraic slip model (ASM).
I would like to compute the bottom variation (due to the settling of the particles). Is there any way of doing this using ASM (or at least to infer this)?
Regards

 ghorrocks December 19, 2011 18:05

Just for clarification - due you mean the ASM turbulence model or a particle slip model?

 antonio December 19, 2011 18:09

I mean algebraic slip model.

 ghorrocks December 19, 2011 18:17

OK, thanks. What do you mean by the "bottom variation"? Are you trying to work out how the particle settle at the bottom of the thing?

What type of particles are they?

 antonio December 19, 2011 19:02

Please imagine a mixture of water and sediments (by definition the algebraic slip model uses spherical particles) entering in a smooth flume with "clean water". You will have a wave of sediments that will "travel" along the flume and some of these sediments will settle in the bottom. This rate of deposition will be different along the flume.My main purpose is to see how the bottom longitudinal configuration has changed along the flume from time t=0 (where the bottom was smooth) to a certain time (where some sediments have settled in the bottom).

 ghorrocks December 19, 2011 19:33

CFX does not have good models for what you want to do.

If you consider the lagrangian particle approach - CFX can model particles falling out of the flow, but does not have deposition models or scouring models. So the particles will fall out of the flow and hit the bottom wall, but will disappear and not build up a bed.

If you consider the eularian approach (which appears to be what you are looking at) - you can use maximum packing fractions to build up something which looks like a "bed", but it does not have the proper physics of particle bedding and scouring so would not be very accurate in many applications.

I think Fluent has some recent stuff on this area in the V14 release. Also other codes such as Flow3D have some models here. But if you want to use CFX then I would consider linking it to a DEM code (eg EDEM) so the DEM code handles the bedding and scouring models and CFX does the fluid flow.

 antonio December 19, 2011 19:46

Please can you explain in more detail what you mean by "you can use maximum packing fractions to build up something which looks like a "bed""?Are you saying to consider a certain upper limit of the sediments mass fraction in order to "predict" the "new" bed configuration?

 ghorrocks December 19, 2011 19:50

It depends on what you are trying to do. If you only want to see where the particles drop out of the flow then do not worry about it and what you have described is fine. But if you want to form a bed or a bed which then can be scoured then it is more complex.

The issue is that there is no limit to the volume fraction of particles by default. The particles will keep building up assuming they take no space until the volume fraction reaches 1. But in reality there is a maximum packing fraction where all the particles are packed together, but the fluid exists in the gaps between the particles. This maximum packing fraction is a volume fraction less than 1.

 jjansen October 9, 2012 09:34

I am currently trying to tackle a similar problem using cfx 14.
I decided to try a mesh displacement method using ASM to simulate the settled sand as a bed. Simultaneously I remove the mass of the settled sand using a sink term in the bottom boundary. This seems to work for simple settling colum cases, however when I swith to a case with a "flow component" (for instance a bucket with inlet (0.1 volconc. particle mixture) and an outlet (stat.pr. boundary) the model crashes after a certain amount of time. Did anybody gain any more knowledge on this topic so far, with perhaps related tips and tricks?