label for different particle groups

Submitted by Oliver.pasqual on Sun, 05/13/2012 - 12:03

dear all,
I want to do a particle mixing test.
Is there any method that I can use to mark different particle groups, all the particle groups have the same material parameters, such as young modulus..possion rate...thermalconductivity and so on, the only difference is the label. (group 1, group 2.....)
For example, after all the particles accumulate in the bottom of the container, the upper half region was settled with label 1, and the lower group with label 2.
is it possible to deal this problem as similar to the temperature setting. the set command was used following a region command.
region up block INF INF INF INF 0.022 INF units box
set region up property/peratom Temp 400

it will be highly appreciated if you can give me some suggestions.
Best regards
oliver

Forums:

Oliver.pasqual | Mon, 05/14/2012 - 02:46

dear all,
the problem has been fixed with the same way as for the temperature.
firstly define a variable using fix property/atom command,
secondly define a region using region command,
then set this region with a modified value to the variable defined in fix command.
thanks
oliver

pfalkingham | Tue, 05/15/2012 - 21:08

At the risk of asking a silly question, could someone explain how to get such a variable attribution into the dump files?

So, in a similar case to the original problem, I'm looking to give a 'layer' value to each particle.

I create the variable with:

fix m7 all property/atom layer scalar yes yes yes 0

Then I set this once all the particles are in and settled with:

region up block INF INF INF INF 0.05 INF units box
set region up property/peratom layer 2

region down block INF INF INF INF INF 0.05 units box
set region down property/peratom layer 1

However, I can't suss out from the documentation how to add this to the custom dump. Presumably I'm doing something wrong somewhere? Liggghts doesn't throw out any errors, so I guess all the particles have this property... I just can't see it.

ckloss's picture

ckloss | Wed, 05/16/2012 - 10:33

You can do so with f_layer[0] (see eg. heatGran example). Btw. I would use "yes no no", there's no need for communicating this property every time-step

Cheers, Christoph

pfalkingham | Wed, 05/16/2012 - 18:14

Thanks Cristoph, I appreciate the help. Unfortunately I'm still getting:
ERROR: Could not find dump custom fix ID

I couldn't reconcile the dump command from heatGran, as Temp/heatFlux (outputted as f_Temp[0] and f_heatFlux[0]) don't seem to be in the input file aside from at that line.

If anyone needs to look at the input I'm using, it's available here:
http://dl.dropbox.com/u/608161/in.scratchFoot

(and the simple cad indenter is here: http://dl.dropbox.com/u/608161/cyl.stl in case anyone needs to run it).

ckloss's picture

ckloss | Wed, 05/16/2012 - 18:54

sorry - I was a bit too fast. if you define the fix like

fix layer all property/atom layer scalar yes yes yes 0

you can access with f_layer[0]

Cheers, Christoph

pfalkingham | Wed, 05/16/2012 - 19:28

Doh! So simple; sorry!

Thanks very much it's working as expected now.

P