[jerome_]

hi everyone,

I really need your help.
Well, i've got this equation :
F = Fx*10000*exp(-((x - x0 - y + y0)/y)² − betax*(y − y0)²) i + Fy*10000*exp(-((x − x0)/y)² - betay*(y − y0)²) j

My geometry is a simple rectangle : (-10,0);(-10,10);(10,10);(10,0) dimensions in mm. All the edge are wall.

I try to create a 2D jet centre in (0,0). So i've created 2 forces which are symetric.
The distance between the 2 forces is 12mm.

That's my UDF :

#include "udf.h"

DEFINE_SOURCE(force_1_x,c,t,dS,eqn)
{
real x[ND_ND];
real source;
real fx0,x0,y0,betax;
fx0=2.6;
x0=-0.004;
y0=0.001;
betax=800000;
C_CENTROID(x,c,t);
source =fx0*10000*exp(-((x[0]-x0-x[1]+y0)/x[1])*((x[0]-x0-x[1]+y0)/x[1])-betax*(x[1]-y0)*(x[1]-y0));
dS[eqn]=0;
return source;
}

DEFINE_SOURCE(force_1_y,c,t,dS,eqn)
{
real x[ND_ND];
real source;
real fy0,x0,y0,betay;
fy0=2;
x0=-0.004;
y0=0.001;
betay=10000000;
C_CENTROID(x,c,t);
source =fy0*10000*exp(-((x[0]-x0)/x[1])*((x[0]-x0)/x[1])-betay*(x[1]-y0)*(x[1]-y0));
dS[eqn]=0;
return source;
}

DEFINE_SOURCE(force_2_x,c,t,dS,eqn)
{
real x[ND_ND];
real source;
real fx0,phi,x0,y0,betax;
fx0=2.6;
x0=0.006;
y0=0.001;
betax=800000;
C_CENTROID(x,c,t);
source =-fx0*10000*exp(-((x[0]-x0-x[1]+y0)/x[1])*((x[0]-x0-x[1]+y0)/x[1])-betax*(x[1]-y0)*(x[1]-y0));
dS[eqn]=0;
return source;
}

DEFINE_SOURCE(force_2_y,c,t,dS,eqn)
{
real x[ND_ND];
real source;
real fy0,phi,x0,y0,betay;
fy0=2;
x0=0.006;
y0=0.001;
betay=10000000;
C_CENTROID(x,c,t);
source =fy0*10000*exp(-((x[0]-x0)/x[1])*((x[0]-x0)/x[1])-betay*(x[1]-y0)*(x[1]-y0));
dS[eqn]=0;
return source;
}

So as you can see, the only difference is a minus in the force_2_x.
The problem is that i don t observe a symetric jet.

So to try to understand the problem, i've observe the result give by fluent when there is just the force 1 and when there is just the force 2.
The result are not symetric.
For example, with the force 1 i've a max velocity mag of 5.55m/s with the force 2 this value is about 5.75m/s

Anyone have an idea to resolve this problem or just to explain me why i've obtain such result.

Thanks,

Jerome