[blerli_91]

Dear Forum,


For future projects I want to implement alternative Drag laws for the DDPM model. For getting experience in UDF programming (I am quite new to the topic) I am trying to recreate the Fluent Build-In Drag law by Wen and Yu. I have a 1D test case for comparison of the original drag law and my UDF. For comparison I am using a DPM-Sample at the Outlet for a patched Flow-Field.
According to UDF Manual the Define_DPM_Drag should return an expression equal to: . First I tried to directly insert the Wen-Yu Cd given in the Theory Guide into the expression:




However, I was not able to recreate the results of the original Drag Law. So I guess I will need another formulation for Cd? Maybe someone has experience with that? I am not sure which kind of expression I should use for : in case of Wen-Yu.
The code I was using.:

#include "udf.h" 

#include "dpm.h" 

DEFINE_DPM_DRAG(particle_drag_force,Re,tp) 

{ real drag_force, void_s,void_g; 

 cell_t cell; Thread *thread_s, *mix_thread;  

cell=P_CELL(tp); 
 
 mix_thread=P_CELL_THREAD(tp);  

thread_s = THREAD_SUB_THREAD(mix_thread,1); 

void_s = C_VOF(cell, thread_s);  void_g =1-void_s ; 
 //drag_force = 18.0 * pow(void_g,-2.65) * (1 + 0.15 * pow(void_g * Re,0.687))/void_g ;
 
drag_force = 18.0 * (1 + 0.15 * pow(void_g * Re,0.687))/void_g ; 
return (drag_force); }

Thank You and Best Regards

[blerli_91]

Looks like I found the problem here,


The Formulation ANSYS offers for the WenYu model in the Theory Guide is not complete.

Looks like Fluent is setting C_D = 0.43245 for particle Reynolds numbers > 1000 (Similar like for the Gidaspow model with C_D = 0.44 for Rep > 1000).

As far as I know this is not included in the original WenYu Formulation?



best regards

[filippo]

Hi Blerli,

I have your same problem. Have you managed to solve it? Even if I set a constant CD, I still don't get the same results of the Fluent Wen-Yu model.

Thanks and best regards

[filippo]

Hi,

I'm writing this in case someone is dealing with the same problem.

We were getting the wrong result because we were missing an alpha_g. The correct formula to get the drag force from the gas-solid exchange coefficient and viceversa is:

alpha_g*alpha_g*ro_p*F_D=K_pg*(Ug_Up) (alpha_g*ro_p comes from the assumption that n_p*m_p=alpha_g*ro_p in each cell)

So that finally the value that must be returned by the UDF is:

drag_force = 18.0 * pow(void_g,-3.65) * (1 + 0.15 * pow(void_g * Re,0.687))

I still get vaguely different results, but at least now they are comparable to the Fluent Wen-Yu model.

Best regards,
Filippo

[taha91_1@hotmail.com]

Originally Posted by filippo

alpha_g*alpha_g*ro_p*F_D=K_pg*(Ug_Up) (alpha_g*ro_p comes from the assumption that n_p*m_p=alpha_g*ro_p in each cell)

I do not understand how n_p*m_p=alpha_g*ro_p, left side has units of mass while right side has units of density.

[filippo]

Originally Posted by taha91_1@hotmail.com

I do not understand how n_p*m_p=alpha_g*ro_p, left side has units of mass while right side has units of density.

n_p is the number of particles per unit volume, so 1/m^3