CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Divergence detected in AMG solver:k (http://www.cfd-online.com/Forums/fluent/102769-divergence-detected-amg-solver-k.html)

thiyam June 1, 2012 13:27

Divergence detected in AMG solver:k
 
Hi all,

I am using Eulerian multiphase model with standard k-epsilon turbulence model in gas-liquid phase, udf for drag co-efficient is interpreted successfully but when i start iteration, exactly after 11th iteration it stops showing error as

Error: divergence detected in AMG solver:k

please help

Sixkillers June 1, 2012 14:16

Try to reduce Courant number (I am assuming that you are using coupled solver) and also under relaxation factors (for multiphase flow it is really necessary).

Jabba June 12, 2012 14:15

Quote:

Originally Posted by Sixkillers (Post 364350)
Try to reduce Courant number (I am assuming that you are using coupled solver) and also under relaxation factors (for multiphase flow it is really necessary).

How are you initiating your solution?

Maybe you can try to run a previous calculation without resolving the volume fraction (uncheck it on your Solution Controls) to get an initial field for the multiphase run then.

thiyam June 13, 2012 07:47

Thanks for your reply

Its running now but its not converging at all especially the gas phase residuals. Mine is pressure based solver and simulating in steady state condition and the following are the boundary conditions:

Velocity inlet -velocity of liquid=0 & volume fraction=0
-velocity of gas=0.0025m/sec & volume fraction=1
-diameter of bubble=0.002m
please check these conditions as i am little surprised when i mistakenly put gas volume fraction as zero, then the solution was run fine but when i come to know this mistake i changed the value from zero to one, it starts showing problem in convergence.

Help is needed

karan malik November 8, 2013 12:28

drag UDF divergence in eulerian model
 
Hi there fellow fluent users....this is the first time I've been using UDF's and I'm not a used user of Fluent. I am also getting the similar error.I have interpret the UDF for drag closure in eulerian model for air- water system.After 20 iterations I get an error message. The following message appears: divergence detected in AMG solver: k, at this instance the x,y momentum residual values are of the order of 10^-12
I am using implicit method

the code is:

#include "udf.h"
#define pi 4.*atan(1.)
#define diam2 0.005
DEFINE_EXCHANGE_PROPERTY(clift_new_drag,cell,mix_t hread,w_col,a_col)
{
Thread *thread_w, *thread_a;
real x_vel_w,x_vel_a,y_vel_w,y_vel_a,abs_v,slip_x,slip_ y,rho_w,rho_a, mu_w,reyb,drag_coeff;

/* find the threads for the water (primary) */
/*and air (secondary phases) */

thread_a = THREAD_SUB_THREAD(mix_thread, a_col);/* primary phase */
thread_w = THREAD_SUB_THREAD(mix_thread, w_col);/* secondary phase*/

/* find phase velocities and properties*/
x_vel_w = C_U(cell, thread_w);
y_vel_w = C_V(cell, thread_w);

x_vel_a = C_U(cell, thread_a);
y_vel_a = C_V(cell, thread_a);

slip_x = x_vel_w - x_vel_a;
slip_y = y_vel_w - y_vel_a;

rho_w = C_R(cell, thread_w);
rho_a = C_R(cell, thread_a);

mu_w = C_MU_L(cell, thread_w);

/*compute slip*/
abs_v = sqrt(slip_x*slip_x + slip_y*slip_y);
/*compute Reynold's number*/

reyb = rho_w*abs_v*diam2/mu_w;
if(1.0<=reyb<=10.0)

drag_coeff=(29.1667/(reyb))-(3.8889/(pow((reyb),2.0)))+1.222;
else
drag_coeff=(24.0/(reyb))*(1.0+0.15*pow((reyb),0.687));

return drag_coeff;
}

Thanks in advance.


All times are GMT -4. The time now is 15:05.