CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)
-   -   Transient profile boundary conditions (https://www.cfd-online.com/Forums/cfx/75320-transient-profile-boundary-conditions.html)

Transient profile boundary conditions

Hi everyone,

I am trying to apply transient 3D velocity values from experimental data to the inlet of a model in CFX 12. There is quite a bit of variation to the velocities over the inflow profile and the subsequent secondary flows they produce are of particular interest for my application. I know this is fairly easy to do with a steady state problem using Profile Boundary Conditions but was wondering if anyone had advice for doing this for a transient case. It seems like a logical step to me so I am hoping there is something I have missed in my searches of the manual and CFD-Online. Ideally, I would be able to load a .csv file, with something along the lines of:

x[m] y[m] z[m] u[m/s] v[m/s] w[m/s] t[s]
x1 y1 z1 u1(t1) v1(t1) w1(t1) t1
x2 y2 z2 u2(t1) v2(t1) w2(t1) t1
...
x1 y1 z1 u1(t2) v1(t2) w1(t2) t2
x2 y2 z2 u2(t2) v2(t2) w2(t2) t2
...

or something similar (separated into groups or files by timesteps perhaps), and have CFX interpolate values across the mesh at each timestep. It would also be nice if I could interpolate at intermediate timesteps but I can do the interpolation spatially and temporally myself with scripts if it is an issue. The problem is loading the data into the boundary at each timestep.

To me it seems like my options could possibly be:

1) Separate each velocity component into separate timestep data files. Create a Junction Box Routine that executes at the end of each timestep then increment the name of the data file it reads for the velocity info (u_profile_t1.dat --> u_profile_t2.dat) for each component. Read in the new data, interpolate values across the mesh, solve, repeat.

2) Write a CEL routine somehow capable of passing the current time back and tricking CFX into doing Profile Boundary Conditions at each timestep.

3) Appeal to the helpful team at CFD-Online and see if anyone can give me a push in the right direction.

Cheers,
Dave

 ghorrocks April 22, 2010 07:25

I think you will need a Junction Box routine to do this, as you suggest.

 cosine June 16, 2010 07:48

Hello,

I have a similar problem. I would like to know if you have successfully managed to make your simulation. If so, could you explain how to do this?

Thank you very much
Regards

 SLC July 11, 2011 08:20

Quote:
 Originally Posted by dladd (Post 255789) Hi everyone, I am trying to apply transient 3D velocity values from experimental data to the inlet of a model in CFX 12. There is quite a bit of variation to the velocities over the inflow profile and the subsequent secondary flows they produce are of particular interest for my application. I know this is fairly easy to do with a steady state problem using Profile Boundary Conditions but was wondering if anyone had advice for doing this for a transient case. It seems like a logical step to me so I am hoping there is something I have missed in my searches of the manual and CFD-Online. Ideally, I would be able to load a .csv file, with something along the lines of: x[m] y[m] z[m] u[m/s] v[m/s] w[m/s] t[s] x1 y1 z1 u1(t1) v1(t1) w1(t1) t1 x2 y2 z2 u2(t1) v2(t1) w2(t1) t1 ... x1 y1 z1 u1(t2) v1(t2) w1(t2) t2 x2 y2 z2 u2(t2) v2(t2) w2(t2) t2 ... or something similar (separated into groups or files by timesteps perhaps), and have CFX interpolate values across the mesh at each timestep. It would also be nice if I could interpolate at intermediate timesteps but I can do the interpolation spatially and temporally myself with scripts if it is an issue. The problem is loading the data into the boundary at each timestep. To me it seems like my options could possibly be: 1) Separate each velocity component into separate timestep data files. Create a Junction Box Routine that executes at the end of each timestep then increment the name of the data file it reads for the velocity info (u_profile_t1.dat --> u_profile_t2.dat) for each component. Read in the new data, interpolate values across the mesh, solve, repeat. 2) Write a CEL routine somehow capable of passing the current time back and tricking CFX into doing Profile Boundary Conditions at each timestep. 3) Appeal to the helpful team at CFD-Online and see if anyone can give me a push in the right direction. Cheers, Dave
Did you ever have any luck with this? I am looking at solving a very similar issue.

 pfister July 11, 2011 09:55

Hello!

I had a similar problem and I solved it by using a function.

Create a new function, use the option interpolation and specify the arguments unit (e.g. [s] for time) and the results unit (e.g. [m/s] for velocity).

At the interpolation data option tab use three dimensional and then right click in the field below and choose import data.

If you want to use these data as a boundary condition you have to specify an expression. Create an expression and define it as your function with the corresponding argument (e.g. t for time).

Now you can use your data as bc.

That's it, I think!

Cheers.

 Super Sonic January 8, 2012 15:58

Greate the a csv file

Hi,

I am tryinf to create a file, apparently .csv file, to import it to a transient simulation.But me problem is that i don't know how to create the collums. For expample from where I can find the coordinates of

x , z , y

And secondly, how I can create a .csv file.
Can some give an example.
Thanks

 ghorrocks January 8, 2012 20:25

You can create csv files in excel or any text editor. They are simply comma separated value files (hence CSV).

 Super Sonic January 9, 2012 05:03

Transient Simulation

Thank you very much for your respond.

Is it possible for you to upload an example file and explain me exactly the reason of each factor?

Thank you very much.

 ghorrocks January 9, 2012 05:45

Search google for CSV file format.

 All times are GMT -4. The time now is 08:45.