[Car]

Hello everyone,

I have some trouble simulating a fouling process in Ansys for my bachelor thesis. I have almost no experiences with the software, I'm learning it for about three months. My task is to simulate lime fouling on a heat exchanger on basis of an article that was handed to me. According to the article I have to write a UDF to define the deposition rate. I'm working on that since a few weeks and tried several options to solve the problem. Unfortunately none of them worked out so far and I don't know if it's depended on the UDF or the set up in general or maybe both. In the following I explain a bit what I did:

1. My first try was to simulate the chemical reaction as a wall surface reaction and define a source term with the define_source macro. It is possible for me to hook the function and to start the calculation successfully but I received some errors. For example a backflow in xxxx cells or some kind floating-point error.

I googled the errors and found that there might be a problem with the mesh but I think that should be fine. Another hint was that I might have tried to access some data (with the UDF) to which I have no access and I don't know how I can verify this or how I could gain the needed access if this is the problem. I was thinking that might be caused by the header files and checked them, and they seemed to be OK as well.

To exclude the possibility that the setup in general is bad I hooked the Ansys example of define_source and it worked perfectly.

Has anyone an idea what I can try in reference to the stated issue?

2. I tried to write a define_mass_transfer code but wasn't even able to run the example given by Ansys. If I got it right, this option includes in the setup the mixture model and not (only) the species transport model. But with this setup I am not sure how I can restrict the fouling to the surface. But that might not be necessary anyway (because the wall of interest is heated and the saturation concentration leads to more fouling in warmer areas which is defined in the UDF).

The problems now stated should explain my thoughts and I hope they are precise enough for you to have an understanding about what the purpose is and maybe a good advice respectively my starting point or a different way to solve the problem.

Thanks for your efforts

[`e`]

What are the flow conditions, setup, UDF, the article you are referring to and the specific errors? "Backflow" could be reasonable (consider recirculation at an outlet), whereas "floating-point error" indicates a mistake in the code. What models are you using, the species transport, or others?

[Car]

It is a turbulent flow through a straight pipe. It’s modelled in 2D. One of the walls is heated. The article I referred to is called ‘CFD modelling of CaCO3 crystallization fouling on heat transfer surfaces’ by T. Pääkkönen et al.
The process fluid is water. I used the species model to model the reaction ca^2+ + co3^2- à caco3. Therefor I created a mixture out of the different components which wasn’t easy because ions have not properties that can be set in Ansys. I created the ions in dependence on water because they are solved. I modelled it as a surface reaction.
Further I use the RNG k-e-model, energy equation is on.
Boundary conditions:
- Inlet: velocity inlet, 0,4 m/s
- Outlet: pressure outlet
- Walls: no slip, 300 K

[Car]

The UDF looks like this
#include "udf.h"
#include "mem.h"

#define con_bulk 0.000418
#define k_0 1.62*pow(10.,20.)
#define E_a 147700

DEFINE_SOURCE(massdeposition,c,t,dS,eqn)
{
real con_sat;
real k_r;
real j_def;
real U_tau;
real tau_wall;

con_sat = -3.97*pow(10., -10.)*pow((C_T(c,t)-273.15),3.)+1.28*pow(10.,-7.)*pow((C_T(c,t)-273.15),2.)-1.67*pow(10.,-5.)*(C_T(c,t)-273.15)+1.22*pow(10.,-3.);
k_r = k_0*exp(-E_a/(C_RGAS(c,t)*C_T(c,t)));
tau_wall = C_MU_T(c,t)*C_DVDY(c,t);
U_tau = sqrt(tau_wall/C_R(c,t));

j_def = k_r*pow((con_bulk-con_sat),2.)*(C_MU_T(c,t)/(C_R(c,t)*pow(U_tau,2.)));

return j_def;
}

[Car]

The UDF is hooked under cell zone conditions -> surface body -> source term -> mass



and after running I receive the message #QNBe+00 for all parameters and in the end

Error: WorkBench Error: Could not handle event: SolutionStatusUpdate Error Object: #f