This probably relates to this problem, but I've posted a new topic to attach some images...
I have just tried to verify the coordination number for some simple packings (in this case simple cubic). I have computed and dumped the coordination number using computer coord/gran. I was expecting a coordination number of 6 everywhere apart from at the edges...
I have attached some images showing the overall cube, and then a cubic clip of the centre portion, to avoid edge effects (there are no periodic boundaries in this sim), and a slice through a larger setup. As can be seen, at the centre the coordination number is zero, and the maximum anywhere is 3 - the average from the log file is 2.7.
I tried playing with the neighbor command - changing this value results in changing "Total # of neighbors" and "Ave neighs/atom" but does not change the values from the compute - which are always in the range 0-3 here.
Is this a bug, or am I doing something wrong?
Regards, Mark
| Attachment | Size |
|---|---|
 sc_coord.png | 212.15 KB |
| sc_coord_centre.png | 132.73 KB |
| sc_coord_slice.png | 214.36 KB |
| sc_10per_overlap_coord.png | 162.6 KB |
 in.sc_packing.txt | 1.75 KB |
| sc_packing_zoom_force.png | 131.54 KB |
ckloss_ | Wed, 08/24/2011 - 01:12
Hi Mark, how did you generate
Hi Mark,
how did you generate the packing - do the particles _really_ touch, i.e. overlap? The compute coord/gran only counts particles that are actually in touch (as opposed to neighbors).
If the lattice spacing is exactly equal to particle diameter, this is probably a round-off issue. In that case, you try a slightly smaller lattice spacing. If the problem persists, feel free to send me the script
Christoph
msbentley | Wed, 08/24/2011 - 13:45
Still having problems with 10% particle overlaps
Hi Christoph,
I generated the packing with the lattice command - I've attached the script to the original post. I've modified it to allow for a large (10%) overlap, and to make it periodic in all directions. However, I still don't see the expected coordination number - see the new graphic sc_10per_overlap_coord.png.
I've also checked by computing pair/gran/local and zooming to look at one particle (see sc_packing_zoom_force.png) - here the coordination number is 3, but the forces acting with the nearest 6 particles are also seen. Although they are rather small - I'll go back and check the particle parameters (which were copy and paste from an example!).
Any hints as to what I'm doing wrong gratefully accepted!
Thanks, Mark
ckloss | Wed, 08/24/2011 - 15:37
Hi Mark, thanks for posting
Hi Mark,
thanks for posting the script! seems that this is a bug, I will have a look at it and fix it next week when I am back in office!
Cheers, Christoph