[Many]

Hello Multiphase friends!

I am beginning with my research in multiphase flow simulation of a fixed bed reactor. I have a tube where flow a liquid and a solid particles dilute on it will be passing trough. In the tube I have a bed of solid particles as well, with a fixed fraction. I would like to study hydrodynamics of these system, mainly pressure drop and liquid holdup.

I have read several scientific literature about packed/fixed/trickle beds. There exist several approachs that seems to provide accurate results regarding aerodynamics behaviour.

I have tried to model my problem with several strategies:

1. Triphase flow using Eulerian approach, assuming that solid particles are fixed inside the tube (granular phase with packed bed option enabled and also fixed velocities for the solid granular phase)

2. Biphase flow, using the same approach but in this case granular phase is modeled using porous media approach (I have computed my inertial and viscous coefficients based on concentration of solid particles and diameter as well, using Blake-Kozeny equations)

3. Single phase flow, using species transports (liquid phase, with two species, liquid and dilute solid particles) and porous media in order to model the fixed bed.

I am quite surprised about the results. I realised that:

a. Presure drop varies a lot between models 1 and 3 and model 2 (by an order of magnitude). Differences between models 1 and 3 are lower, but also noticiable (pressure drop from 3 Pa to almost 6 Pa)

b. Velocity profile is almost the same in the three models.

c. Contours of solid dilute shows and evolution in triphase simulations with time. In the other two models, both fluids seems to be much mixed and evolution is not easy to distinguish.

So, please, let me know your thoughts about:

-Which model better represents my problem (1,2 or 3)?
-Are the differences between models reasonable?
-I have used fixed velocities for all zones for fixed bed particles. Is this the better strategy to maintain fixed the solid particles in a zone?

Thanks to all the experts that spend time helping beginners!

Cheers,

Many

[mila.d.b]

Hi,

I am starting some simulations in a packed bed (liquid-solid), Eulerian model. I am having problems with convergence when I try to increase the time step (I need to simulate around 3.000 s of flow. So using a time step of 0.01 s is not feasible. Do you have any tips and tricks to get that done?

Thanks

Camila

[mehul020892]

Take adaptive time steps and reduced convergence criteria . Also you can use quick scheme.