[victorueda7]

Good morning, I'm trying to program an UDF for viscosity. I have this code:

#include "udf.h"

DEFINE_PROPERTY(custom_viscosity, c, t)
{
double visco;
double gamma=C_STRAIN_RATE_MAG(c,t);
visco=75/gamma+0.042;
return visco;
}

But, starting to run the calculation, Fluent shows the next message:

"Divergence detected in AMG solver: x-momentum"

Does anybody knows which is the problem with the code?

Thank u very much.

[moh_zain]

Hi if the calculation start and then stop it may be from the mesh quality or solution method you use( focus on mesh quality first )

[pakk]

Hint: what happens if the strain rate is zero?

[victorueda7]

Thank u moh_zain, but the problem is that I'm simulating a function that is already in the database as "Herschel Bulkley" model, so, if the simulation works with the default "Herschel Bulkley" function, it should be the same with my UDF

[victorueda7]

Originally Posted by pakk

Hint: what happens if the strain rate is zero?

Thank u pakk, the problem is that this UDF function is supposed to return the same values as the "Herschel Bulkley" default function, that I have already tried with the same configuration (mesh...). So, I guess the problem is I don't know how to use the "Shear rate" variable, could u help me?

Thank u very much again.

[pakk]

No, your problem is in the physics.
The equation that you give is only valid if the strain rate is larger than the strain rate threshold. For low strain rates (and strain rate = 0 is low), a different equation holds.