yade-users team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

> For this algorithm we need to keep l(t-dt/2) for next time step.
>
> Any comments?
>
>   
Yes, one comment : you don't need to keep any additional variable (as 
stated before). :-P

l(t-dt/2) is I(t-dt/2).w(t-dt/2).
w(t-dt/2),  is already kept in body-> velocity (or the rotational 
equivalent).
For I(t-dt/2), you can use the same approach and consider that you 
define values of I at times (t+k.dt/2).

No additional variable as long as you don't need a value x(t-3.dt/2).

Bruno

--

_______________
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
________________