[rjm1982]

Hi all,

I am having some difficulty working out a satisfactory way of tracking the time-varying height of the water surface at a particular (x,y) location in a simple 2 phase (water and air) channel simulation with waves. I would like to get a time series of z_ws(t), where z_ws(t) is the vertical elevation of the water surface at a specified location. I suspect the only way to do it might be with an UDF. Does anyone have any experience with this please?

Many thanks,

R

[qin chunqiu]

Hello,
I have the same problem as you posted. Have you solved it?

[DEd]

I think I have done this at some point. Don't have fluent available at the moment but if I remember it correctly I made a line and took the integral of VOF along this line.

[rjm1982]

Thanks for your reply DEd. In the end I did this by making a line (Surface-->Line), as you say. Then, with a text file for the VOF values along the line at each time step, I wrote my own post-processing code in Fortran to read the text files and then interpolate to get the position of the free surface at each time step. It does the job, and would be easy to do in Matlab or VBA too.
Thanks again,
Richard

[annan]

Hey,

Hope I'm not a bit late to answer your question, but another way would be to write a UDF to track the secondary phase volume fraction. when the value changes to let's say 0.5, you can consider that you reached the height of the interface. Then you can store those values in a UDM and exploite them afterwards.

Hope this will help.
Good luck

[elvislf]

Originally Posted by annan

Hey,

Hope I'm not a bit late to answer your question, but another way would be to write a UDF to track the secondary phase volume fraction. when the value changes to let's say 0.5, you can consider that you reached the height of the interface. Then you can store those values in a UDM and exploite them afterwards.

Hope this will help.
Good luck

Hi annan, I'm doing a project about two phase flow(water and vapor water) in a rotating pipe, currently I am confused about how to tract and keep the interface between those two phases, the interface information is the key for further simulation. Can U help me with that? thanks a lot

[annan]

Hi elvislf,

Could you please be more specific about what you want to do ? tracking the interface or finding its position can be done as I explained, by putting a condition on the volume fraction of one of the phases in a UDF, if its value reaches 0.5 than you are positionned at the interface.

Waiting for more details
Good luck

Annan

[elvislf]

Originally Posted by annan

Hi elvislf,

Could you please be more specific about what you want to do ? tracking the interface or finding its position can be done as I explained, by putting a condition on the volume fraction of one of the phases in a UDF, if its value reaches 0.5 than you are positionned at the interface.

Waiting for more details
Good luck

Annan

Thank you for your reply, I am doing the two phase simulation inside a rotating heat pipe with Fluent and UDFs, the VOF model was used here. The 2-D symmetric model was build and the working fluid are water and water vapor. I already know that tracking the interface between vapor and liquid is the key mission in evaporation simulation, I want to realize it with UDFs, but I don't have much experience in it. So can U help me dealing with this UDF or give me some idea about it?

[annan]

Hello elvislf,
I can give you a sketch UDF as I don't have time to test it, but the idea is the same. At first, I'd suggest you go for a DEFINE_ON_DEMAND UDF and use it for the post-treatment of one of your converged cases, it won't stop the simulation if there is any error and won't impact your final result.

DEFINE_ON_DEMAN(track_interface)
{
Domain d = Get_Domain(1);
face_t f;
cell_t c0,c1;
Thread *tf,*t0,*t1;

thread_loop_f(tf,d)
{
begin_f_loop(f,tf)
{
t0 = THREAD_T0(tf);
c0 = F_C0(f,tf);

c1 = F_C1(f,tf);
t1 = F_C1_THREAD(f,tf);

Thread *tl0 = THREAD_SUB_THREAD(t0,1);
Thread *tl1 = THREAD_SUB_THREAD(t1,1);

if (C_VOF(c1,tl1)<0.5 && C_VOF(c0,tl0)>0.5)
{
/*You reached the interface*/
/*You can compute whatever you want to do when you are at the interface*/
}

}
end_f_loop(f,tf)
}
}

Do not hesitate to contact me in private if you have any other questions about this.

Hope this will help
Good luck
Annan

[elvislf]

Originally Posted by annan

Hello elvislf,
I can give you a sketch UDF as I don't have time to test it, but the idea is the same. At first, I'd suggest you go for a DEFINE_ON_DEMAND UDF and use it for the post-treatment of one of your converged cases, it won't stop the simulation if there is any error and won't impact your final result.

DEFINE_ON_DEMAN(track_interface)
{
Domain d = Get_Domain(1);
face_t f;
cell_t c0,c1;
Thread *tf,*t0,*t1;

thread_loop_f(tf,d)
{
begin_f_loop(f,tf)
{
t0 = THREAD_T0(tf);
c0 = F_C0(f,tf);

c1 = F_C1(f,tf);
t1 = F_C1_THREAD(f,tf);

Thread *tl0 = THREAD_SUB_THREAD(t0,1);
Thread *tl1 = THREAD_SUB_THREAD(t1,1);

if (C_VOF(c1,tl1)<0.5 && C_VOF(c0,tl0)>0.5)
{
/*You reached the interface*/
/*You can compute whatever you want to do when you are at the interface*/
}

}
end_f_loop(f,tf)
}
}

Do not hesitate to contact me in private if you have any other questions about this.

Hope this will help
Good luck
Annan

Thank you for your help, I will try it. This is my e-mail: elvislf@outlook.com. Wish you all the best.