[kdot]

Hello,

I have UDF's for mass and momentum that are used to model jet blowing near a surface (acts perpendicular to wall). A gaussian is used to distribute the mass and momentum sources. My mass source works as intended, the sum of the user mass source equals the input value of mass in the UDF and the structure of the gaussian is present. My code for the mass UDF is the following:

#include "udf.h"
DEFINE_SOURCE(SJ_Massb2, cell, thread, dS, eqn)
{
real mass;
real massper;
real C, h;
real source;
real xcc, ycc;
real xc[ND_ND];
real sigmax, sigmay;
real distribution_func;
real r1, r2, A0;
real vol;
{
C_CENTROID(xc,cell,thread);
mass = 2.8194 ; // desired mass flow rate
xcc = 0.2; // x mean location of jet
sigmax = 1e-3;//%sigma value where 6*sigma = jet injector width
sigmay = 1e-3;//% sigma value that acounts for jet penetration
ycc = 3e-3 ; // y mean location of jet
r1 = (xc[0]-xcc)*(xc[0]-xcc);// (x-xc)^2 for numerator in gaussian distribution
r2 = (xc[1]-ycc)*(xc[1]-ycc);// (y-yc)^2 for numerator in gaussian distribution
A0 = 1.0; // momentum coeff (amplitude)n % 1 by default
distribution_func = pow(2.0*M_PI*(sigmax*sigmay),-1.0);// leading denominator in gaussian equation
distribution_func = A0*distribution_func*exp(-r1/(2.0*(sigmax*sigmax))-r2/(2.0*(sigmay*sigmay)));// completed gaussian equation

source = distribution_func*mass; // multiplying the gaussian by the source term
dS[eqn] = 0.0;
}
return source;
}

The method I am using to model the momentum source has been used in several papers (if references are desired I can dig them up). The momentum source is applied as mass flow rate (of jet) * velocity (of jet) and then distributed by the gaussian. My issue is that the momentum source appears to be weak and has little to no effect on the flow. The mass source appears to be the main contributor. My momentum source only appears to have an effect (similar to that of modeling the jet orifice and inlet) if I increase the magnitude by a factor of ~10. This worries me because a factor of 10 in the momentum source (mass flow rate * velocity) would require an unreasonable mass or velocity.

#include "udf.h"

DEFINE_SOURCE(SJ_Ysourceb2, cell, thread, dS, eqn)
{
real source;
real r1, r2;
real distribution_func,sigmax,sigmay;
real Upart,mass;
real Uinf;
real yc[ND_ND];
real xcc, ycc;
real C, h;
real angle;
real cs, sn;
real A0;
real vol;
{
/* source term */
C_CENTROID(yc,cell,thread);
sigmax = 1e-3;//sigma x to encapsulate all jets on surface
sigmay = 1e-3;//jet penetration
Upart = 1.4097; // velocity of jet
mass = 2.8194;// mass flow rate of jet
xcc = 0.2; // x mean location
ycc = 3e-3; // y mean locaiton
r1 = (yc[0]-xcc)*(yc[0]-xcc);
r2 = (yc[1]-ycc)*(yc[1]-ycc);
A0 = 1.0; // momentum coeff (amplitude)
distribution_func = pow(2.0*M_PI*(sigmax*sigmay),-1.0);
distribution_func = A0*distribution_func*exp(-r1/(2.0*(sigmax*sigmax))-r2/(2.0*(sigmay*sigmay)));

source = (distribution_func*mass*Upart)*1;//mdot * U * gaussian

dS[eqn] = 0.0;
}

return source;
}

I am wondering if I am somehow formulating the momentum source term incorrectly when I am using a gaussian to distribute the momentum. Is my momentum source formulation or application incorrect?