yade-users team mailing list archive

[Janek Kozicki <janek_listy@xxxxx>]

Kien Dang, Mr said:     (by the date of Sat, 9 Feb 2008 15:01:41 -0500)


> I am currently implementing the NBS and MR contact detection
> algorithms, since the cell should be big enough for the largest

what is NBS and MR contact detection? Is it some sort of grid collider?


> element. I intent to divide all the bodies into 2 group: spheres and
> boxes. Then, the contact detection is performed over the spheres. The
> problem arise when I need to detect contact between box and sphere.
> Is there any other way than to loop over all the spheres and boxes?

This will be extremely useful for everyone here if you send us the
code to add it to repository. We are seriously considering to start
giving more people svn acces, so that everyone can contribute their
own code with much less hassle. Just some permission access details
must be looked into.

Also - if code written by you: the yade users, is added to svn
repository - it is automatically converted (by us, the developers) to
work with latest yade version. You *will* like this feature, because
yade is going to evolve.
If you have your own code, which you don't share by adding it to SVN
you will not want to make an upgrade to new yade version, since your
code will stop working with new version....


> Btw, I dont know exactly what algorithm is used in
> PersistentSAPCollider and is it faster than the NBS when I look at
> the code, it seems that the sorting algorithm was employed (not sure
> yet)

I have explained this algorithm in the past, have a look in the
archives (and also this explanation should be put into wiki...)
Look at other posts in thread also, I just did a 'grep collider' on
my inbox archive.

  http://lists.berlios.de/pipermail/yade-users/2006q4/000363.html
  http://lists.berlios.de/pipermail/yade-users/2006q4/000427.html

Only a short paragraph, starting with "In DEM contact detection is
Sweep and Prune collider ....":

  http://lists.berlios.de/pipermail/yade-users/2006q4/000353.html


you could try searching the archive to find more relevant posts. I
just made only a quick search.

best regards
-- 
Janek Kozicki                                                         |
_______________________________________________
Yade-users mailing list
Yade-users@xxxxxxxxxxxxxxxx
https://lists.berlios.de/mailman/listinfo/yade-users