2D makeCloud

Hi,
The check on input parameters was excessively restrictive. I removed it. Now the command you show works correctly (update yade-daily package) even if periodic=False.

>what is the way to fill a box with the maximum number of spheres of the same radius?

Translating you question into input parameters gives:
sp.makeCloud(minCorner=(0,0,0),maxCorner=(1,1,0),rRelFuzz=0,rMean=something)

... which works like a charm. :)

Thanks,
the 2D makeCloud command works well now.

Actually the second question was not really about having the same radius for all the spheres but more about the input "porosity" in makeCloud(). which is the range of this input?
moreover the command "num" seems to be uneffective: if I run this command

sp.makeCloud(minCorner=(-.1,0,-.1),maxCorner=(.1,0,.1),rMean=.01,periodic=False,num=100)

I get
WARN /build/buildd/yade-daily-3+3367+42~lucid1/pkg/dem/SpherePack.cpp:213 makeCloud: Exceeded 1000 tries to insert non-overlapping sphere to packing. Only 58 spheres was added, although you requested 100.

but the makeCloud box seems to have space for more than 58 spheres

I would like to understand if input "porosity" and "num" are somehow linked and how they affect the generation of the sphere cloud

Hope to have explained clear the issue

Thanks

Riccardo

>but the makeCloud box seems to have space for more than 58 spheres

The number of additional spheres that could fit in the voids looks very limited, and the fact that you see a few free spots does not mean that the algorithm is capable of finding them.

num, meanRad, and porosity are linked of course. Please state your problem more clearly, as for now the only thing I could do is to rephrase the documentation on how they interact.

ok,
using makeCloud() ,the maximum number of particles of the same radius I can put into a box defined by minCorner and maxCorner is achieved by calling the function with the default inputs:
(num=-1,porosity=.5)
am I right?

Does the porosity have something to do with the amount of void space in the makeCloud box? Because I can not see difference between porosity=0 and porosity=1 (or porosity=10)

let's now call

radius=.005
sp=pack.SpherePack()
sp.makeCloud(minCorner=(-.1,0,-.1),maxCorner=(.1,0,.1),rMean=radius)
sp.toSimulation(material=MatSp)
box=O.bodies.append(utils.geom.facetBox((0,0,0),(.1,radius,.1),material=MatBox,wallMask=31)) # the same box of makeCloud
O.engines=[
 ForceResetter(),
    InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb(),Bo1_Wall_Aabb()]),
    InteractionLoop(
        [Ig2_Sphere_Sphere_ScGeom(),Ig2_Facet_Sphere_ScGeom(),Ig2_Wall_Sphere_ScGeom()],
        [Ip2_FrictMat_FrictMat_FrictPhys()],
        [Law2_ScGeom_FrictPhys_CundallStrack()]
    ),
 NewtonIntegrator(gravity=(0,0,-9.81),damping=.1)
]
O.dt=utils.PWaveTimeStep()

at the end of the run, which is a trivial gravity deposition, I see almost half of the box empty:
is the makeCloud command able to produce a particles distribution like the random one achieved in this way filling the entire box?
For example I see that regularHexa() packs spheres in such a way, but of course following a regular distribution; what I'd like to have is a random one.

Riccardo

Thanks Chareyre, that solved my question.

got it!
thanks