← Back to team overview

yade-users team mailing list archive

Re: twist Moment and Bending Moment in cohesionlessMomentRotation


Thank you all for you reply about my question. Of course I agree with all of you about the fact that there is no contradiction about considering a twist moment and we can even find a physical meaning. However the disadvantage is that we need to use two more parameters, representing a total of 7 parameters. Hence it is always a question of compromise.

About your suggestion Bruno to include all that clearly in Law2_ScGeom_CohFrictPhys_CohesionMoment, it raises for me the usual dilemma: should I use the last uptodate trunk version or a fixed version. Personnally I made the choice to use a fixed version, as long as a given work is not finished. The main reason is that otherwise, if I want to compare results, I don't know if differences are due to changes in the code or not. Then, when the work is finished and I start something new, I upgrade to the last version. Hence currently I use Yade 0.50, and I was interested in CohesionlessMomentRotation, because among others, in this version the code of this contact law is quite clean and short.


Bruno Chareyre a écrit :

I am still investigating the cohesionlessMomentRotation.
It is written in the sources files that this code has been "verified with the paper of Plassiard in GM".

However, Plassiard considers in his paper "and in his thesis" only the rolling part of the relative rotation of particles (and thus only bending moment), whereas in the code rolling and twist part of the relative rotation are considered (and thus bending and twist moment).
There would be no contradiction here, if only it was possible to set twist stiffness = 0 (so the law with bending and twist would "contain" the law with bending alone).
For now, Ktwist=Kbending=Kr, so it is indeed not possible...

My question: is there a particular reason for that? What is the motivation of the person who wrote this code? I don't say that is bad or good, but I would like to have an idea about advantages and disavantages, and physical meaning for considering a twist moment.

As soon as there is a finite area of contact between the solids (i.e. always), it makes sense to include twist resistance I think. The testing has been done by Boon IIRC.

By the way, I think rolling resistance, Ktwist!=Kbending, and other features would be ideally implemented in Law2_ScGeom_CohFrictPhys_CohesionMoment, which is now the cleanest, and shortest code for moments. Funky factors η, α, β, would be contained in a collection of Ip functors, since they are not needed by the law itself, which just needs 4 stiffnesses, and the definition of maximum values for each force/moment.

Any help on this task of unifying duplicates would be welcome.



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

Luc Sibille

Université de Nantes - Laboratoire GeM UMR CNRS

IUT de Saint Nazaire
58, rue Michel-Ange - BP 420
44606 Saint-Nazaire Cedex, France

Tel: +33 (0)2 40 17 81 78