hi, dear:

is it possible for me to use Fluent to solve the following problems: using lagrangian method to track each particle moving in fluidized bed, particle-particle collision must be considered and using LES model to calculate gas hydrodynamics?

thanks for your help

zhou

No. Lagrangian method can not be applicable to the fludized bed because,

1) Mass loading ratio(solid mass / gas mass) in the fludized bed exceeds the CFD limit of dilute phase assumption. According to my experience and Fluent's manual, the limit of mass loading ratio is about 1.0. But, in the fludized bed reactor, the ratio is order of 10, or sometimes higher.

2) Particle size in the fludized bed reactor is too large for Lagrangin method.

You should use another method, full Eulerian-Eulerian approach. Fluid flow can be solved by using ASM(approximate solution) with current version, and I heard that E-E can be used with Fluent's next version(6.x). Unfortunately, with Fluent 5.x(maybe with Fluent 6.x), reaction in the fludized bed reactor can not be solved.

Sincerely, Jinwook

hi, dear LEE

thanks very much.

however, Eulerian-Lagrangian approach was widely used to simulate hydrodynamics in fluidized bed (for example, the groups are Tsuji, kuiper, tadrist etc.) this years. for us, we are very interested in particle of different size and density. as you know, it is too complicated and difficult to use E-E model under the situations although kinetic theory was used.

anyway, thanks again.

zhou H.S

Are you sure they use Lagrangian or the recently popular discrete element method (DEM) which accounts for particle-particle interactions?

This is probably a reasonable approach if you have the memory/speed to simulate all those particles. I know some researches have written a DEM code (they are quite small I undertstand) and then linked it through source tersm into CFX. I imagine you could do something similar with Fluent, but it will take a while to develop this...

I don't think they consider reactions at this stage just the particle and hydrodynamics.

Greg

A) There is no p-p intercation model in fluent so far. B) Other than this, you can do all of the above provided that you have enough computing facilities C) U can do p-p intercation in fluent provided you use UDF to implement any particular model.

DC

I do not think that you can obtain converged solution by E-L approach for fludized bed considering real operating condition. I think that It is the limit of CFD. Would you check the solid/gas mass loading ratio for your system ?

Sometimes, many research group, including me, do E-L approach for fludized bed by one-way coupling to see approximate behavior. But as far as my experience is concerned, it is impossible to obtain converged solution by two-way-coupling with actual solid/gas mass loading ratio of fludized bed. For Fluent, you may refer the manual.

Sincerely, Jinwook