[luisvolumentres]

Hi!

I'm trying to simulate a water tank which has multiple inlets and applying to them a UDF to specify a parabolic inlet profile. The problem comes while calculating the distance from any point of the inlets surface to its centre because the inlets are not in point (0,0,0). I'm using a 3D version of the 2D example found in Fluent Manual:

/************************************************** *********************/
/* vprofile.c */
/* UDF for specifying steady-state velocity profile boundary condition */
/************************************************** *********************/
#include "udf.h"

DEFINE_PROFILE(inlet_x_velocity, thread, position)
{
real x[ND_ND]; /* this will hold the position vector */
real y;
face_t f;

begin_f_loop(f, thread)
{
F_CENTROID(x,f,thread);
y = x[1];
F_PROFILE(f, thread, position) = 20. - y*y/(.0745*.0745)*20.;
}
end_f_loop(f, thread)
}

I assume that I could change y = x[1] to y = x[1] - x0, being x0 the centre of every inlet. However, since I almost have 200 inlets it would be something very annoying if I do it manually. Could anyone give me a hint for obtaining the centre of each of the inlets automatically so I can easily make it work?

Thank you everybody in advance!

[cbooks77]

You know you can just attach a velocity profile without using a UDF? You can also use expressions. I would look into just using the profile. Solve the fully developed flow in a single inlet separate model. Then take that profile data at the boundary and apply it to all your inlets. Also you are using 3D so you would need a conical velocity inlet as parabolic is 2D. Thats going to be more challenging to implement in a UDF...

https://www.youtube.com/watch?v=DQYlnsAfs5s

[luisvolumentres]

Thank you for your reply!

Expressions have the same problem as UDF, you need to know the centre of each inlet in order to have the X,Y and Z coordinates relative to the centre of each inlet, right? I guess the same problem would arise while mapping the inlet with the Velocity Profile since position changes for each inlet. Regarding the conic inlet for 3D case, it is just straight forward to implement it based on the 2D UDF shown in the Fluent UDF Guide and it works properly for a simple geometry (e.g. one pipe).

[pakk]

Something like this might work: calculate the average position in the UDF.



real ysum=0, yavg;
int ycount=0;
begin_f_loop(f, thread)
{
F_CENTROID(x,f,thread);
y = x[1];
ysum +=y;
++ycount;
}
end_f_loop(f, thread)
}
yavg=ysum/ycount;
begin_f_loop(f, thread)
{
F_CENTROID(x,f,thread);
y = x[1]-yavg;
F_PROFILE(f, thread, position) = 20. - y*y/(.0745*.0745)*20.;
}
end_f_loop(f, thread)
}

I did not test it and wrote it on my phone, so typos might be there.

This calculates the average position giving each face the same weight. It's better to make the weight proportional to the area. But if your mesh is not too weird, the error should be very small.

[luisvolumentres]

Originally Posted by pakk

Something like this might work: calculate the average position in the UDF.



real ysum=0, yavg;
int ycount=0;
begin_f_loop(f, thread)
{
F_CENTROID(x,f,thread);
y = x[1];
ysum +=y;
++ycount;
}
end_f_loop(f, thread)
}
yavg=ysum/ycount;
begin_f_loop(f, thread)
{
F_CENTROID(x,f,thread);
y = x[1]-yavg;
F_PROFILE(f, thread, position) = 20. - y*y/(.0745*.0745)*20.;
}
end_f_loop(f, thread)
}

I did not test it and wrote it on my phone, so typos might be there.

This calculates the average position giving each face the same weight. It's better to make the weight proportional to the area. But if your mesh is not too weird, the error should be very small.

Hi, thank you for your reply! I have just tried it in a simple geometry with its inlet centre not in [0,0,0] and it worked perfectly! Thank you so much for your help .

Best regards