yade-users team mailing list archive

[Bruno Chareyre <question675377@xxxxxxxxxxxxxxxxxxxxx>]

Question #675377 on Yade changed:
https://answers.launchpad.net/yade/+question/675377

Bruno Chareyre proposed the following answer:
Hi Mehdi,
If I understand correctly the main problem is that you expect from the transformation tensor properties which only apply in the small strain + small rotation approximation. Typically: exy =0.5 (O.cell.trsf[0,1] + O.cell.trsf[1,0]). The problem is this is not a frame independent quantity in large deformations.
When you don't reset trsf to identity it has components of the order of 0.7, which is far away from a "small" thing. This is reflected by the fact that your definition of shear is not frame independent. It is to be expected.

What needs to be constant is the change of trsf, with or without
reseting it to identity. And it is the case, see code below replacing
your last section. Note that the composition of transformations implies
products, not additions, therefore calculating an "increment" between
two transformations needs multiplication by an inverse, while you where
doing substractions.

Cheers

Bruno


######## to be activated for comparison
O.cell.trsf=Matrix3.Identity
O.run(1,True)
initTrsf= O.cell.trsf

##############
dummy_deps = -1
triax.goal=[-100,-100 ,triax.stressTensor[2,2]] #-100 is meaningless. Just to make sure straining is occurring in the right diretion
nT = 100
O.run(nT,True)

#print "incremental trsf:",
dF=O.cell.trsf*(initTrsf.inverse())
print 'dFxx=', dF[0,0], ' dFyy=',dF[1,1], ' dFxy=',dF[0,1]



Bruno

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