yade-users team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

> I dw'/dt + w' cross Iw' = T'
>
> when (.)' is value (.) in local r.f. This equation don't integrated by
> simple leap-frog, because we will have w' on right.
>
> So, problem is not only in a rotation of vectors. We need also other
> integration algorithm for clumps.
>
> Is this correct?
>
>   


Yes, I think this is correct. And this non linear equation cannot be 
inverted explicitely.
It needs an iterative procedure for each clump at each time-step (1), 
and of course a matrix representation of I.

For low-spin rotations however, one can neglect the term in w'^2. If 
aditionally, you neglect out-of-diagonal terms in I (or if I is given in 
principal axis), then you can use a simpler algorithm and leap-frog 
still applies (2)  (and you use a vector representation of I).
This approximation is proposed as an option for clumps in PFC3D.

We could perhaps make sure that situation (2) is handled correctly first 
(with current definition of inertia*, it doesn't need big changes, if 
any), before going to situation (1). Is somebody planning to simulate 
high-spin clumps by the way?

Bruno

(*) Sorry Vaclav, I should have check the definition of inertia before 
claiming it was that of a sphere. You develop too fast for me! ;)





> ==
> Best regards,
> Sergei D.
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> _______________________________________________
> 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
>
>   


--

_______________
Chareyre Bruno
Maître de Conférences

Grenoble INP
Laboratoire 3SR - bureau E145
BP 53 - 38041, Grenoble cedex 9 - France
Tél : 33 4 56 52 86 21
Fax : 33 4 76 82 70 43
________________