hi there,

I am new user of CFX. Now I am trying to use cfx to simulate the flows of mixture of gas and solid in horizontal pipe. can anyone whoever use cfx give me some advices on the application of CFX in this field?

RGDS,

Daninwin

Hi Daninwin, Welcome as a new CFX user.

Based on your question it seems like you are also new to gas-solid flow, you need to be more specific

First of all you have to consider what kind of problem you have, is it dense or dilute? This would give you an indication of which model is most suited for your problem; either Lagrangian where solids are treated as single particle, or an Eulerian model where the solids are regarded as a kind of continiuum.

When you have decided which model you are going to use, new questions will arise. Do look in the CFX examples, find some basic information in your library about modeling gas-solid flows, post your new question here, and I am sure you will get some help

Regards Jan

DEAR JAN, Is it possible to consider the k-e-kp or any other new turbulent-particle interaction model in CFX?if yes,how?please explain. Thanks, sanjib.

Dear Sanjib

Note my comments are only for CFX4.3 When using a Lagrangian approach for the particles it is possible to include the effects of turbulence on the particles with the standard eddy-life time model. It is easy to include, only a keyword under model characteristics. (note when including particle dispersion to have many particle trajectories in ordrer to obtain a statistical proper solution)

If using an Eulerian approach it is possible to include two-way dispersion effects, gas turbulence on particles and particles influence on gas turbulence. It is all desribed in the manual

It is also possible to include the two-way effect in the Lagrangian model, but you have to do some user fortran then.

Hope this helps,

Regards Jan

Dear Jan,

Thank you for your comments.But I think you missed my question.I asked,whether I can write a separate user-defined function for a turbulence model for gas-solid flow that is not there in CFX.I'll be grateful to you for the answer.

Best regards,

Sanjib.

Sorry if I missed your point.

I am still not sure what you mean, (it's early on the day here, I have not got my coffee yet).

I you want to write your own dispersion model,- for turbulent dispersion of the particles , instead of the standard eddy-lifetime in the CFX. The answer is yes, it can be included via the user fortran routine usrdrg. I have been looking at it once, if that is what you want let me know, I am very interested in that.

If your question is if you can write your own turbulence model, please explain a bit more.

Regards Jan

Dear Jan, Thanks for the reply.Let me explain my point to you.I have solved dilute-particulate flow using the Lagrangian technique.But the limitation is that I have till now used only those turbulence models that are there by default,eg.k-e model,k-w model.But there are many other new advanced models such as the k-e-kp model where in addition to the k,e/w, there is another similar equation involving the kp term.It is suggested that the turbulent-particle interaction is better simulated using this model.If I want to use this or any other new advanced model that is not there in CFX4.3 by default,can I do that?I think I have to write a separate User-Defined Function for that.Am I right? Another important modification is to include the particle-particle interaction.Many new models have come up.CFX4.3 does have the provision to include the Kinetic Theory.However,if I want to include a new equation then the same problem as I have stated above will come. How do I solve these?

I think I have explained my point. I'll be grateful to you for the answers.

Regards, Sanjib.

Dear Sanjib,

Interesting work you are doing.

You should be able to define your own transport equations with the use of user scalars and invoking the appropriate source terms via user fortran. Have you tried other models that include the impact of particles on the turbulence? They should be at bit easier to include- only extra source term in the k and epsilon equations. Regarding the Kinetic Theory model for inclusion of particle-particle interaction, is it really necessary with and additional transport equation, is it not more a question about adjusting the turbulent viscosity? which can be done in usrvis.

Regards Jan

Hi Sanjib,

I was reading through this thread and was interested in this new turbulence model that you discuss (k-e-kp). Could you direct me to some references from the open literature that describe this model in greater detail.

Thanks,

Jorge

you will find this in a recent issue of CHEMICAL ENGINEERING SCIENCE and INTERNATIONAL JOURNAL OF MULTIPHASE FLOW.also look for POWDER TECHNOLOGY. best of luck.