Output for compute wall/gran/local contactPoint

Submitted by jsams on Fri, 02/09/2018 - 01:53

Hi everyone,

I'm using the following compute for accessing particle-wall contact information:

compute fc all wall/gran/local pos vel id force contactPoint

and the following dump command for writing that information to an output file:

dump walldmp4 all local 1000 post1000/fc*.dump c_fc[1] c_fc[2] c_fc[3] c_fc[4] c_fc[5] c_fc[6] c_fc[7] c_fc[8] c_fc[9] c_fc[10] c_fc[11] c_fc[12] c_fc[13] c_fc[14] c_fc[15] c_fc[16] c_fc[17] c_fc[18] c_fc[19] c_fc[20] c_fc[21]

Everything seems to work well except the output for contactPoint (last three entries: c_fc[19] c_fc[20] c_fc[21] ), which is 0 for all particle wall contacts. See the following exemplary line in my dump file:

ITEM: ENTRIES c_fc[1] c_fc[2] c_fc[3] c_fc[4] c_fc[5] c_fc[6] c_fc[7] c_fc[8] c_fc[9] c_fc[10] c_fc[11] c_fc[12] c_fc[13] c_fc[14] c_fc[15] c_fc[16] c_fc[17] c_fc[18] c_fc[19] c_fc[20] c_fc[21]
0 0.00413315 6.26724 0.0150834 0.00413315 6.26724 0 0 0 0.00551094 -0.00807788 -0.00915771 4 4 54651 0.696635 0.198186 0.194157 0 0 0

Has anyone experienced something similar, or can give me a suggestion as to what I might have done wrong? Any help is kindly appreciated!

medvedeg | Sat, 02/10/2018 - 14:09

Hallo,

in case of particle-wall contact, first 6 elements of c_fc contain coordinates of particle center and of the contact point.
In case of particle-particle contact these elements contain coordinates of the centers of contacting paticles. In order to be able to dump contact points the keyword "contactPoint" was implemented. For particle-wall contact this keyword is useless and 0 0 0 will be dumped. I agree, this must be clarified in the documentation and in the code.

Alexander Podlozhnyuk

jsams | Sun, 02/11/2018 - 22:31

Thanks for the answer!

I kind of figured that out after a little while, but as you said, I think it wouldn't hurt to clarify it in the doc.