Dump of overlap statistics

Submitted by Tamoghna on Wed, 06/20/2012 - 15:43

Hi!

Could you tell me a way to dump the overlap statistics for each timestep?

Regards and thanks in advance,

TM

cstoltz | Thu, 06/21/2012 - 13:16

Only way I know to do it is to dump positions and then calculate overlaps pairwise. I think that the dump options are pretty much all particle oriented, not contact oriented.

Regards,
Chris

Juergen | Thu, 07/24/2014 - 16:39

Hello!
I'm also working on this problem and the command compute pair/gran/local is a very good hint. But I dont get out, to create a dump with the overlap ratio. Here is my stuff about this:
compute pgl all pair/gran/local pos id force
variable pglres atom sqrt((c_pgl[4]-c_pgl[1])^2+(c_pgl[5]-c_pgl[2])^2+(c_pgl[6]-c_pgl[3])^2)
compute maxpgl all reduce max v_pglres
variable maxpgl atom c_maxpgl

But now I can't dump c_maxpgl or v_maxpgl. I tried much variations, like:
dump dmp_maxpgl all custom ${ts_dump} ${modelname}/vtk/${modelname}.maxpgl id c_maxpgl
(-->Error: Dump custom compute does not compute per-atom info (../dump_custom.cpp:1255)
With local in the dump command it's that error: Invalid attribute in dump local command (../dump_local.cpp:421)
What is wrong?

Regards and thanks in advance,
Juergen

Juergen | Fri, 07/25/2014 - 16:11

Ok, variables can't get input local vectors or local arrays. Is there any possibility to make these in per-atom vectors? Maybe with the atomfile-style of variables to read the data from the local dump?

Regards,
Juergen

Juergen | Mon, 07/28/2014 - 18:22

Now I try to modify the sourcecode of gran_pair_local.cpp to get my wished data.

Good luck for everyone, who wants the same or similar data.