CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CD-adapco (
-   -   velocity divergence as plot (

Sabatini October 12, 2001 03:28

velocity divergence as plot
Hallo people Is there a way to extract the velocity divergence as plot Thanks

Paal October 16, 2001 05:56

Re: velocity divergence as plot

I am not fully sure that I get your question right: Do you want to plot the diverged "results"?

These results are saved to a file - case.div, and can be loaded in the same manner as normal result files just replacing the case.pst with case.div

steve October 16, 2001 12:11

Re: velocity divergence as plot
I think to get what you want you have to load velocities (GETV,ALL) and then use OPERATE,CURL function. I think CURL is the same as divergence of a function?

Lars Ola Liavåg October 17, 2001 01:49

Re: velocity divergence as plot
The curl of a vector field is itself a vector (cross product of the "upside down" delta and the vector), whereas the divergence (the corresponding dot product) is a scalar. Hence, for the velocity vector V, div V = du/dx + dv/dy + dw/dz, whereas curl V = i(dw/dy - du/dz) + j(du/dz - dw/dx) + k(dv/dx - du/dy), all d's being partial differentials.

In PROSTAR, the OPERATE,CURL function takes the curl of the loaded vector field and returns the curl (vector) of it. You can also calculate the gradient vector, but I don't know how to calculate the divergence in PROSTAR. I guess contacting the support team is a good idea.

daniel January 21, 2002 10:30

Re: velocity divergence as plot
CURL is definitely NOT the same as the divergence of a function, but a DIV command should also be available the same way

cfd January 28, 2002 20:07

Re: velocity divergence as plot
Continuity equation:

div(rho u) = d(rho)/dt

For a converged steady state solution:

div(u) -> 0.

For a transient compressible time step solution:

div(u) = 1/rho [d(rho)/dt - u grad(rho)]

For incompressible time step solution: (rho = const.):

div(u) -> 0.

All times are GMT -4. The time now is 04:17.