|
[Sponsors] |
February 21, 2001, 05:26 |
UDF modification
|
#1 |
Guest
Posts: n/a
|
Hi folks, I'm relatively new to CFD & Fluent and am trying to simulate (in 3D) the flow of water through a pipe. I've spotted a boundary condition UDF on the Fluent documentation site ('parabolic velocity inlet profile in a turbine vane')which I think will help my simulation. The problem is that this UDF is for a 2D problem. My question is (given my very basic knowledge of C) how can I modify the UDF to work in 3D? I've tried to edit here and there, but the results are incorrect. Is it just a case of adding a z component to the UDF? and if so, how? Please help!
Here's the UDF: /************************************************** ***********************/ /* vprofile.c */ /* UDF for specifying a 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) } |
|
February 21, 2001, 07:24 |
Re: UDF modification
|
#2 |
Guest
Posts: n/a
|
Try like shown below: It should work.
#include "udf.h" DEFINE_PROFILE(inlet_x_velocity, thread, position) { real x[ND_ND]; /* this will hold the position vector */ face_t f; cell_t c; real x[ND_ND]; real y,xref,yref,zref; xref=0.0; yref=0.0; zref=0.0; begin_f_loop (f,thread) { F_CENTROID(x,f,thread); y = sqrt((x[0]-xref)*(x[0]-xref)+(x[2]-zref)*(x[2]-zref)); F_PROFILE(f,thread,nv)= 20. - y*y/ (.0745*.0745) *20.; } end_f_loop(f, thread) } |
|
February 27, 2001, 14:21 |
Re: UDF modification
|
#3 |
Guest
Posts: n/a
|
it can't work
|
|
February 28, 2001, 08:10 |
Re: UDF modification
|
#4 |
Guest
Posts: n/a
|
Sorry, but it didn't work. I switched around the x,y,z terms as my flow is in the -ve z direction and it gives a wild overestimation of the inlet velocity at initialisation. It wouldn't compile at first, and when I sorted that out, it didn't seem to work anyway. When I tried using the unmodified 2D UDF supplied by Fluent, the inflow was like a slot within the pipe, so it it not just a case of adding the z direction (x[0] x[1] x[2]) to this UDF?
|
|
March 1, 2001, 07:42 |
Re: UDF modification
|
#5 |
Guest
Posts: n/a
|
This works. You should have y-axis along the length of cylinder.The center of cylinder should pass though the center point. Create a plane slicing at x=0, and see the x-y plot of y-velocity on this sliced plane. You will see the parabola. Your cylinder radius should be 0.0745 to get the parabolic profile. You will see 20m/s at the center and zero at the walls. Try for more cells in radial direction. Enjoy!!!
#include "udf.h" DEFINE_PROFILE(inlet_x_velocity, thread, position) { real x[ND_ND]; /* this will hold the position vector */ face_t f; cell_t c; real y,xref,yref,zref; xref=0.0; yref=0.0; zref=0.0; begin_f_loop (f,thread) { F_CENTROID(x,f,thread); y = sqrt((x[0]-xref)*(x[0]-xref)+(x[2]-zref)*(x[2]-zref)); F_PROFILE(f,thread,position)= 20. - y*y/ (.0745*.0745) *20.; } end_f_loop(f, thread) } |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Help needed on UDF for modification of default syamlal o' brien drag law | caai9 | Fluent UDF and Scheme Programming | 9 | August 20, 2014 09:52 |
UDF parallel error: chip-exec: function not found????? | shankara.2 | Fluent UDF and Scheme Programming | 1 | January 16, 2012 23:14 |
Flowfield temperature modification through UDF | Hypersonicflow | Fluent UDF and Scheme Programming | 2 | April 18, 2011 14:27 |
modification of UDF | ammi | FLUENT | 2 | January 18, 2007 22:35 |
Modification of turb. viscosity using UDF? | moon | FLUENT | 4 | October 2, 2003 12:19 |