Learning Forum

[lucasmassarosousa]

Dear all,

I'm simulating a multiphase flow using Fluent (version 17.1) with air and a granular phase (TFM approach). I already set the physical parameters of the phases, as well as the boundary conditions, and the simulations are converging well.

Now, I want to evaluate the difference in the results when changing the value of the Friction Packing Limit (αs,min) of the granular phase, using the frictional pressure calculated based on 1) the KTGF, and on 2) the Johnson and Jackson model. 

From the Fluent Guide, I can clearly see the expressions for Pfriction and μs,fric that are used in the Johnson and Jackson model to account for the frictional stress (shown bellow).

 

Unfortunately, I can not understand which expressions are being used to calculate the frictional stress using the KTGF, as the only information given in the guide is "based-ktgf, where the frictional pressure is defined by the kinetic theory.", which is too vague for my understanding.

 

* Could you please explain which expressions are used to calculate Pfriction and μfric when using the KTGF approach?

 

* Are the KTGF expresions dependent on the Friction Packing Limit (αs,min) in the same way as in Johnson and Jackson model? 

 

Thank you in advance!

 

 

 

[Amine Ben Hadj Ali]

Same as what you select for the granular viscosity expression.

[lucasmassarosousa]

Dear Amine,

Thank you for responding promptly. Based on your answer, I have two follow-up questions if you allow me.

I choose the Gidaspow expression for the granular viscosity (as shown bellow). Could you confirm then that, when I choose the friction pressure to be calculated based on the KTGF, is this expression (17-309) that is used to calculate μs,fric?

 

The second question concerns some problems of convergence that I'm observing when using the Johnson and Jackson model. To account for the frictional stress I set the frictional viscosity, and the frictional pressure as Johnson and Jackson model in the granular phase properties section.

However, the simulation is running very slow, and I'm receiving messages as "turbulent viscosity limited to viscosity ratio of 1.000000e+05 in 18846 cells" which do not seems to be right for me. I'm using a time step of 10-6. 

Do I need to change other parameter/solver settings to use the Johnson and Jackson model? 

Thank you!

[Amine Ben Hadj Ali]

1/Yes

2/It would require baby-sitting. You at first try transient and/or reduce URF's. I cannot provide any other suggestions without looking into the case but as ANSYS Stuff I cannot do that.

[lucasmassarosousa]

Alright, thank you for the help!

[Amine Ben Hadj Ali]

Mark this as Is Solved.

[lucasmassarosousa]

Dear Amine, 

Some details are still confusing me on the friction stress calculation in Fluent. Please, find bellow the properties of the granular phase.

  

In this case, is the frictional viscosity (μs,fric) being calculated by Eq. 17-309 (Gidaspow's granular viscosity, based ktgf) or by the Schaeffer's equation (Eq. 17-315, as shown bellow)?

  

If the answer is Schaeffer's equation, what is the expression for Pfriction? Is it equal to the one shown bellow?

Thank you in advance!