yade-users team mailing list archive

[Jan Stránský <question690066@xxxxxxxxxxxxxxxxxxxxx>]

Question #690066 on Yade changed:
https://answers.launchpad.net/yade/+question/690066

    Status: Open => Answered

Jan Stránský proposed the following answer:
> but i do not get you concerning the determination of t being tricky.

Maybe "tricky" is not the right word..
the point is that for the contact, you have geom.normal. So the "n" is uniquely defined. Then you have a plane perpendicular to n, which is uniquely defined. Then in that plane, you can choose arbitrarily two perpendicular vectors to be "t1" and "t2" (out of infinitely many possibilities).
This is the "tricky" part.
IMO if you do not really really need it (which should be the case) or do not know exactly what you are doing (you would not ask), then do not use shear components in local coordinate system.

> As far as I know, I thought it was precisely defined in DEM in order to compute contact forces:
> https://yade-dem.org/doc/formulation.html#shear-deformation

contact forces are simply computed in global coordinate system. This is exactly an example of my remark:
>> (often working with global coordinate system but assuming vectors being only in the tangential plane is suitable).

the picture is a 2D illustrative representation (in 2D the situation is
much easier), the computations below it are for the general 3D case.

> How can I get it for my purpose ??

if your purpose is adding forces with O.forces, then use the "global
coordinate system approach" (as O.forces are in global coordinates
anyway).

Or
>> You can choose some two perpendicular vectors arbitrarily if you really need it

cheers
Jan

-- 
You received this question notification because your team yade-users is
an answer contact for Yade.