yade-users team mailing list archive

[Luc Sibille <luc.sibille@xxxxxxxxxxxxxx>]

First, please reply to the yade-users mailing list and not to me. Your 
questions may be usefull for others, and others can help you much better 
than me!

About your questions: replies below

nejib hadda a écrit :
> Hi Luc, 
>
> I have noticed it thanks to your last mail, but still there are some
> questions that I'm not sure about the answer and it would be great if
> you clarify things to me :
>
> 1-  Is the default triaxialtest response with lateral move of the walls
> corresponds to the triaxial test with axial load. Elsewhere, what shall
> I do?

I am not sure to very well understand your question... do you mean: how 
to impose stress paths different of the axisymmetric triaxial compression?

> 2-  in theory q = sigma11-sigma33, but on Yade it is in fact
> sigma22-sigma33. (triaxialtest, triaxialtestwater)

Yes by default with TriaxialCompressionEngine the axial direction is the 
direction 22, but if you sample is isotropic the compression direction 
should influence the result (and if you want to add gravity, you add 
gravity in the direction you want to, direction 22 for instance).
Actually sigma11 may be a little bit different of sigma33 so you can 
compute q = sigma22 - (sigma11+sigma33)/2.
> 3-  Why does Yade go on with calculations until epsilon22 is equal to
> 50% which does not corresponds to small deformations hypothesis.

You can change this value of 50% in the xml, so you stop the simulation 
when you want, it is just a default value. Strains computed in the 
TriaxialStressController are logarithmic strains, thus they should be 
valid for large displacements. Have a look in the archive of the mailing 
list, I think there is questions / answers about that.

Luc

> Nejib
>
> Le jeudi 10 décembre 2009 à 14:17 +0100, Luc Sibille a écrit :
> > Hi Nejib,
> >
> > In the TriaxialCompressionEngine, I think there is a booleen that you 
> > can swith from 1 to 0 to stop the computation when the compaction is 
> > finished. I don't remember the name of the booleen, but it should be 
> > something like "AutoCompressionActivation".
> >
> > About the stress deviator q: generally speaking, during the compaction 
> > phase Yade tries to reach the same stress value for the three space 
> > direction. During the compaction, sigma_1, sigma_2 et sigma_3 may be a 
> > little bit different, thus you can compute a non vanishing q value 
> > (negative or positive), but q value should be very small (and it is not 
> > really a problem).
> > Does it solve your problem?
> >
> >    Luc
> >
> >
> > nejib hadda a écrit :
> > > I don't see how to seperate these two phases on yade.
> > > In fact what I want to do is to compact my specimen before launching the
> > > calculus so that I get the density I wish to have.
> > > But here, yade do both phases simultaniously, and the phase at which q
> > > takes negatives values is the phase of compaction which corresponds to
> > > the first ietrations.
> > >
> > >
> > > Le mardi 08 décembre 2009 à 17:55 +0100, Luc Sibille a écrit :
> > > > For which phase have you got negative q values? The compaction phase, 
> > > > the deviatoric phase, ... ?
> > > >
> > > > Hadda Nejib a écrit :
> > > > > I don’t know if it is common in YADE to have negative values for the 
> > > > > deviatoric stress q   in both triaxial test and triaxial test water for 
> > > > > the first iterations (numerical problem) until the specimen is 
> > > > > stabilized. I have also always had this sudden fall in the value of 
> > > > > sigma22 after some iterations at the beginning of my simulations.
> > > > >
> > > > > Shall I take the initial value of sigmaisocompactation near to value of 
> > > > > the confining pressure to avoid such problem ?
> > > > >
> > > > >  
> > > > >
> > > > >  
> > > > >
> > > > >
> > > > > ------------------------------------------------------------------------
> > > > >
> > > > > _______________________________________________
> > > > > 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