← Back to team overview

yade-users team mailing list archive

Re: Sign_convention_on_YADE


Many thanks Vaclav !

-----Message d'origine-----
De : yade-users-bounces+nejib.hadda=cemagref.fr@xxxxxxxxxxxxxxxxxxx [mailto:yade-users-bounces+nejib.hadda=cemagref.fr@xxxxxxxxxxxxxxxxxxx] De la part de Vaclav Smilauer
Envoyé : mercredi 13 octobre 2010 15:52
À : yade-users@xxxxxxxxxxxxxxxxxxx
Objet : Re: [Yade-users] Sign_convention_on_YADE

> we can display an interaction between shperes in Yade by specifying its
> number with O.interactions, which gives (id1,id2).
> Now if we look for example for the normal vector or the normal force,
> does it correspond (via the convention used by yade in displaying
> contact ids) to the force of id1 applied on id2 or inversely ?
> (same question for the normal vector)

Hi Nejib,

a few things to consider:

1. If you say i=O.interactions[1,2], i is an interaction object, but its
id order can be different: (i.id1==1 and i.id2==2) or (i.id1==2 and

2. For the normal vector, that is always pointing from id1 to id2. If
you have civil engineering background, this corresponds to the axis of
the free surface on a half-beam, if you want.

3. The normal (and shear) force is oriented depending on the consitutive
law, which is considered a bug
(https://bugs.launchpad.net/yade/+bug/493102). I recommend that you
orient normal and shear forces as applied on id1, but as other people
will quickly point out, it is "just" a convention. Yes, it is "just" a
convention, but it should be consistent. In most code I write (like
computing stress tensor etc, where this comes to play), I assume that

It will take a while to unify it, but we will get there.

Hope this helps, Vaclav

Mailing list: https://launchpad.net/~yade-users
Post to     : yade-users@xxxxxxxxxxxxxxxxxxx
Unsubscribe : https://launchpad.net/~yade-users
More help   : https://help.launchpad.net/ListHelp