I am trying to duplicate the postprocessing results of one model as I only simulated quarter of the whole model. I did the same thing according to the suggestions of the adapco online website. However, I found that I sucessfully duplicated the geometry but the duplicated the velocity results were not right: the commands only copy the original results to the newly produced geometry area i.e, the directions of those vectors were the same as the original vectors, however, the duplicated vector directions should be symmetrically reflected and be changed.

What is wrong with my doing? How can I get the results that I want?

The followings are the commands I used to get the results: !Set up the coordinate systems CSYS 1 ! - Find the highest vertex number in the model: ! *GET VOFF MXVE ! ! - Make one additional copy of the symmetrical grid in the ! Y direction of CSYS 1 based on the vertex offset ! CSET,NEWS,FLUID CSET,ADD,SOLID CSET,ADD,BAFFLE CGEN,2,VOFF,CSET,,,VREF,1,2

! ! - Load the cell based velocity vectors ! TRLOAD,*.PSTT, STORE,ITER,500 PSYS,1 GETC,ALL POPT,VECT PLTY,QHID ! ! - Calculate the cell offset. Make sure the cell set is the ! same as the original starting collection. ! *GET,COFF,NCSET ! - Copy the vector data to the duplicate parts DGEN,2,COFF,CSET CSET,ALL CPLOT

Thank you very much for your help!

Lee

I could be wrong but I think this only works when you have compressed your cells before the simulation. Any 'empty' cell numbers in your original model are not transferred to the copied geometry, therefore your post data is copied to the wrong place. Hope that made sense...

First thank you very much for your help, Ralphie!

However, I have to point out that I did compress my cells by using "ccom and vcom" commands before the simulation and before the time I duplicated the results. As my simulated results have some turbulent flows in the solution domain, so the velocity directions are not the same for different points.

I just wonder that problem should not be come out.

Anyway, I guess maybe STAR-Cd only copy the different velocity components to the newly generated geometries and will not do some correspondent changes, for example, when the whole original geometry is reflected about Y axis, the original positive X component of the velocity will be kept as positive value rather than negative value which should be negative in fact after duplication. The description of the command "dgen" just says this point.

I still want to know how to realize the right duplication, hopefully someone can give me further guide.

Thanks a lot for any help!

Lee

Try loading your velocity components (before you copy them) in cylinrdical coordinate system (check PSYS command)

I understand your problem now. Use the CHANGE command for the component you want to modify. Put all new cells in a set (after loading and copying of your vector data), then do

CHANGE CSET,,,-1 0 1 1

if you want to change the X component. This will do

-1 * X + 0 for post registers 1 to 1

There is a designated Post-Processing manual in the Star V4 documentation. Chapter 6 explains mirroring symmetric geometries and results. Good Luck!