yade-users team mailing list archive

[Jan Stránský <question313648@xxxxxxxxxxxxxxxxxxxxx>]

Question #313648 on Yade changed:
https://answers.launchpad.net/yade/+question/313648

    Status: Open => Answered

Jan Stránský proposed the following answer:
Hi Jean-François,

In fact I initially tried the difference between 2 cylinders you suggest and
> I then moved to the revolution surface because I found that the
> inside cylinder was fully filled as soon as its diameter  was close to the
> sphere diameter.


I am not sure if I got it, could you please send a cript reproducing this
problem? There might be some bug in the code..

I thought that the revolution surface approach I posted will give me
an explicit
> control of the angular discretization. I found with a few experiment that
> it was indeed playing a role on the performance of the algorithm but I
> could never come close to the wire I need to model... so my post.


The problem with inGtsSurface is that it evaluates the sphere in/ousideness
only approximately in 6 axis aligned directions [1]. It should not be
difficult to make it more exact.

I assume when you mention other aproach that you are suggesting gravity filling
> with sphere factory ?  something else?


gravity deposition is one option, some kind of compression within desired
volume is another similar option.

Random dense pack first create a packing an then it is "cropped" to the
geometry, so near boundary the packing is not optimal. You can also use
randomDensePack(...,cropLayers=3), from certain point of view it improves
the results [2,3]. Using this parameter and difference of two cylinders, I
got (from my point of view :-) OK results.

cheers
Jan

[1]
https://yade-dem.org/doc/yade.pack.html#yade._packPredicates.inGtsSurface.__init__
[2] https://answers.launchpad.net/yade/+question/291326
[3] https://yade-dem.org/doc/yade.pack.html#yade.pack.randomDensePack


2016-08-01 23:42 GMT+02:00 Jeff <question313648@xxxxxxxxxxxxxxxxxxxxx>:

> Question #313648 on Yade changed:
> https://answers.launchpad.net/yade/+question/313648
>
>     Status: Answered => Open
>
> Jeff is still having a problem:
> Hi Jan,
> Thank you for your answer.
>
> in fact I initially tried the difference between 2 cylinders you suggest
> and I then moved to the revolution surface because I found that the inside
> cylinder was fully filled as soon as its diameter  was close to the sphere
> diameter.
>  I thought that the revolution surface approach I posted will give me an
> explicit control of the angular discretization. I found with a few
> experiment that it was indeed playing a role on the performance of the
> algorithm but I could never come close to the wire I need to model... so my
> post.
>
> I assume when you mention other aproach that you are suggesting gravity
> filling with sphere factory ?  something else?
>
> Thanks anyway
> JF
>
> Jean-François Leon
> Galtenco
>
> Tel [Main-Global]: +1 (415) 322-0376
> Cell [France]: +33 (651) 168-014
>
> On Mon, Aug 1, 2016 at 9:12 AM, Jan Stránský <
> question313648@xxxxxxxxxxxxxxxxxxxxx> wrote:
>
> > Question #313648 on Yade changed:
> > https://answers.launchpad.net/yade/+question/313648
> >
> >     Status: Open => Answered
> >
> > Jan Stránský proposed the following answer:
> > Hi Jeff,
> >
> > I am new to Yade
> >
> >
> > welcome :-)
> >
> > You can define the predicate as a difference of two cylinders [1]. If the
> > gap between inner wire and inner particles is too big, then you would
> have
> > to use different initial packing method than random dense pack.
> >
> > ####################
> > pred = pack.inCylinder((0,0,0),(0,0,length),rOut) -
> > pack.inCylinder((0,0,0),(0,0,length),rIn)
> > sp=pack.randomDensePack(pred,radius=rS,returnSpherePack=True)
> > sp.toSimulation()
> > ####################
> >
> > cheers
> > Jan
> >
> > [1] https://yade-dem.org/doc/user.html#boolean-operations-on-predicates
> >
> >
> > 2016-08-01 14:13 GMT+02:00 Jeff <question313648@xxxxxxxxxxxxxxxxxxxxx>:
> >
> > > New question #313648 on Yade:
> > > https://answers.launchpad.net/yade/+question/313648
> > >
> > > Hi All
> > >
> > > I am new to Yade and this is my first post.
> > >
> > > I am trying to generate a random dense packing in  a cylinder with  a
> > fine
> > > wire on its axis:
> > > my problem : the wire size being much smaller than the sphere diameter,
> > > the spheres overlap it completely and make it "useless as such"
> > > Can someone could provide some advice for improving this packing...? (
> I
> > > would like the sphere against the wire not inside..)
> > > Thanks a lot
> > >
> > > JF
> > >
> > > the instructions I use are pretty simple ( copied and slightly adapted
> > > from some example):
> > >
> > > ###geometry definition
> > > rIn=0.75 # wire radius
> > > rOut=13 #cylinder radius
> > > length=320 # cylinder length
> > > rS=2.6 #sphere radius
> > >
> > > poly=((rIn,length),(rOut,length),(rOut,0),(rIn,0),(rIn,length))
> > > thetas=arange(0,2*pi,pi/24)
> > > pts=pack.revolutionSurfaceMeridians([[(pt[0],pt[1]) for pt in poly] for
> > > theta in thetas],thetas)
> > >
> > >
> >
> surf=pack.sweptPolylines2gtsSurface(pts,capStart=True,capEnd=True,threshold=1e-2)
> > > ### add surface to the simulation
> > > O.bodies.append(pack.gtsSurface2Facets(surf,color=(1,0,0)))
> > > #####generate the packing and add it to the simulation
> > > sp=SpherePack()
> > >
> > >
> >
> sp=pack.randomDensePack(pack.inGtsSurface(surf),radius=rS,returnSpherePack=True)
> > > sp.toSimulation()
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > --
> > > You received this question notification because your team yade-users is
> > > an answer contact for Yade.
> > >
> > > _______________________________________________
> > > 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
> > >
> >
> > --
> > You received this question notification because your team yade-users is
> > an answer contact for Yade.
> >
> > _______________________________________________
> > 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
> >
>
> --
> You received this question notification because your team yade-users is
> an answer contact for Yade.
>
> _______________________________________________
> 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
>

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