yade-users team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

> one hope is that instead of saying O.run(1000), you say O.iter, then 
> calculate the new location of the zero point of the minor axis, then 
> iterate again.  the calculations of the new zero point location would 
> take place outside the definitions/declarations of the engines, but 
> the engines would never-the-less have access to, and make reference 
> to, the new calculated zero point of the minor axis.
What about adding a user setable variable in rotation engine that would 
be the velocity of the reference points (or also rotational velocity 
relative to one reference -- heh sounds like nested rotations)?
That way you don't have to update position, run(1,True), update 
position, run(1,True)... The rotation engine would have a function 
updating itself.
With RefPoint velocity = 0 (default), it would be fixed rotation, like 
now. But if user set velocity=(1,0,0), the rotation axis will be 
translated of (1,0,0)*dt at each step (for instance). It doesn't prevent 
large displacement in only one step since you can always write 
velocity=dispYouWant/dt.

Also, if you really need this run(1)+update in the end, there is a 
better trick : you can use PeriodicPythonRunner to execute a command 
from inside the calculation loop without stopping it. I'm sure it is in 
the documentation, you can also see example usage in scripts (e.g. 
scripts/test/shear.py).

Cheers.

Bruno




> I will test this theory..
>
> --- On *Thu, 7/15/10, Michael Jensen /<jenmichael2000@xxxxxxxx>/* wrote:
>
>
>     From: Michael Jensen <jenmichael2000@xxxxxxxx>
>     Subject: Re: [Yade-users] hello? -help with a spinning bucket!
>     To: yade-users@xxxxxxxxxxxxxxxxxxx
>     Received: Thursday, July 15, 2010, 1:05 PM
>
>     salutations once more :-D
>
>     I have another question for you, very interested to know what the
>     answer will be..
>
>     one of the partial engines is the RotationEngine.  This engine
>     takes several variables, notably a bolean, a point, and an axis. 
>     so far so good.  with a point and axis defined, everybody is happy.
>
>     what if I want my point itself to move?  for example, I want my
>     thing to rotate around an axis whose direction is fixed in time,
>     but I want the location of that axis, defined by the point, to
>     itself rotate around another point and axis, a second point/axis
>     pair.  this is the situation that occurs when you have one object
>     rotating, attached to another one that is also rotating.
>
>     that is the most direct way to approach what I want to do.  the
>     other way is to introduce the manually the forces that result when
>     my bucket spinns, so that I don't have to make it spinn at all,
>     and I can go ahead and add my second spinning device.
>
>     I will have to implement one of these two solutions, or something
>     equivalent, if my simulations are ever to completely represent my
>     device.
>
>     it seems to me like the first option should be relatively easy to
>     code, if it does not yet exist, and that it could be a powerful
>     way to extend yade, in that a user would not be limited to two
>     such spinning things, but could in theory simulate much more
>     complex motion.
>
>     -mike
>
>
>
>
> _______________________________________________
> 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
>    


--
_______________
Bruno Chareyre
Associate Professor
ENSE³ - Grenoble INP
Lab. 3SR
BP 53 - 38041, Grenoble cedex 9 - France
Tél : +33 4 56 52 86 21
Fax : +33 4 76 82 70 43
________________