[svenki7]

hi,

I am doing "Residence Time Distribution" by pulse input method in fluent.But i am getting tracer (area weighted average) concentration greater than 1. I did experiments also. I compared with simulation results with experimental. The simulation results are less than 5 time the experimental results. In simulations, sum of tracer concentration is greater than 1. The simulation results are (mean residence time) is less than theoretical residence time.


Can anybody help me please.

Thanks in advance

[billwangard]

Easy way to do residence time in fluent (without particles):

1. Define a User-scalar
2. Set its diffusivity to very small: 1E-11 or so should work
3. Use default flux vector and unsteady term

Compille the following UDF and hook it to the SOURCES for the UDS:

#include "udf.h"
DEFINE_SOURCE(uds_source,c,t,dS,eqn)
{
dS[eqn] = 0;
return C_R(c,t);
}

Set the UDS to 0 at the inlet.

Then, when you solve, the UDS will have the value of residence time.

Regards,
Bill Wangard
Engrana LLC

[svenki7]

Thank for your reply Mr.Billwangard,
I am newer to UDF that why i choose the pulse method. I don't much about UDF.I attached my geometry.The above one is my geometry. Its converging-converging channel.one end is inlet and one end is outlet. Total length of my channel is 68 mm.The Inlet dia 1 mm

[Many]

Hi CFD friends!

I am trying to compute residence time contours using this kind of UDF.

My questions are the following:

1. Changing the mesh, then the residence time contours change (at least the maximum of residence time). I guess this is not desirable. Anybody has an idea about why this is happening? I realized that mesh cells appear in the function...

I am wondering if achieving mesh independent results will fix this, it's to said, max_residence_time tends to real_max_residence_time as cell_size tends to zero

If not, I am not understanding the value of this UDF as residence time will be mesh-dependent...

2. Which values should I use for Schmidt number and mass diffusivity for water?

3. When I got convergent results, I just solve for the defined scalar in steady state. Is unsteady solution absolutely needed? In other words, have I switch to unsteady model just for solving the scalar transport?

Thanks!