yade-dev team mailing list archive

[Bruno Chareyre <bruno.chareyre@xxxxxxxxxxx>]

I don't see anything wrong in Law2_ScGeom_CpmPhys, but I think there is 
a problem with contact point in ScGeom for distant interactions. In 
Scgeom detectionFactor>1 just changes the rule for return value of go() 
and Ig instanciation (for e.g. letting capillaryLaw get something even 
if there is no contact). detectionFactor doesn't enter any equation 
relative to contact geometry itself or contact law.

As a consequence, this is perhaps not what you want :

Ig2_Sphere_Sphere_ScGeom.cpp :
Real penetrationDepth=s1->radius+s2->radius-distance;
scm->contactPoint=se31.position+(s1->radius-0.5*penetrationDepth)*normal;

Later used in CpmLaw like this :
applyForceAtContactPoint(BC->normalForce+BC->shearForce,contGeom->contactPoint,I->getId1(),pos1,I->getId2(),pos2);

I'm also wondering if avoidGranularRatcheting=true is not a problem in 
your case, since it means rad1/rad2 will be used whatever the distance. 
I added avoidGranularRatcheting attribute in Ig2_Sphere_Sphere_ScGeom to 
let you choose.
Or maybe you are changing refRadius in the interaction, leading to 
refRad1+refRad2=initialDistance? In that case it should work even with 
avoidGranularRatcheting=true.

I see scripts/test/cpm-dem3dof-scgeom.py is giving very different 
results if nSteps=1000. Do you know why?
Also a bit strange : unbalanced force in uniax.py looks relatively 
similar in compression, but very different in traction (it is increasing 
with ScGeom, decreasing with Dem3Dof!)

I have more questions than answers for now, sorry.

Bruno



On 18/07/10 23:24, Václav Šmilauer wrote:
> Hi Bruno,
>
> I tried to use ScGeom for the Cpm model, in previous three commits
> (2374-2376). Macroscopic results look very differently (esp. resistance
> in compression, and also the failure pattern - in
> examples/concrete/dem3dof-scgeom.table) although contact-level behavior
> is more or less the same, modulo incremental effects (as it seems to me)
> -- figures can be obtained from scripts/test/cpm-dem3dof-scgeom.py.
>
> Could you please check that Law2_ScGeom_CpmPhys_Cpm uses ScGeom
> correctly? I am only rotating shear displacement using the rotate
> function, then adding the current increment. Is that sufficient? I am
> not sure how prevNormal is used and so on.
>
> Cheers, Vaclav
>
> PS I might be out of mail in the following few days, sorry if I don't
> reply promptly.
>
>
>
> _______________________________________________
> Mailing list:https://launchpad.net/~yade-dev
> Post to     :yade-dev@xxxxxxxxxxxxxxxxxxx
> Unsubscribe :https://launchpad.net/~yade-dev
> 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
________________