Clump orientations and their angular velocities

Hello,

> I am a relative novice in using Yade

welcome :-)

> Since the momentums of inertia of clumps are defined in the space of
> principal axes I would expect also the orientations of clumps to be aligned
> with the principal axes, but I can't seem to find any documentation on this
> matter.

it is "moment of inertia" (not momentum). You are right that inertia is
defined in the space of principal axes, let's call it local coordinate
system. So in LCS the inertia tensor T would look like T=diag(inertia).

The orientation is internally stored as a quaternion, but it can be also
transformed to rotation matrix R. It is simply a mapping between local to
global coordinate system. So in GCS the inertia tensor is Rt*T*R. Rt means
transpose of R.

I am afraid that the only documentation is source code itself [1]..

> Regarding the angular velocities - are they defined in the local
> (principal axes) coordinate system or in the global coordinate system?
>
>
angular velocities are in global coordinate system. Be careful if you
wanted to prescribe angular velocity, it would have no effect and one have
to prescribe angular momentum.

cheers
Jan

[1]
http://bazaar.launchpad.net/~yade-pkg/yade/git-trunk/view/head:/core/Clump.cpp

For future reference:
Answer #1 refer to the "aspherical" integration scheme [1]. For spheres, or as soon as the "spherical" approximation is used for non-spherical bodies, angular velocity can be prescribed and angular momentum is not used.
Bruno

[1] https://yade-dem.org/doc/formulation.html#orientation-spherical