> 1) //! stretch ratio ?(n) > (http://en.wikipedia.org/wiki/Finite_strain_theory#Stretch_ratio) > > This comment is wrong, the strectch ratio is |F.x|/|x|, which is a > function of non-diagonal terms. The values returned by getStretchRatio > are not stretchRatios, they have no clear meaning and are not invariants > of F. > As mentionned before, out of diagonal terms have an effect on stretch > ratios (this is why I don't know how to work with _size without Hsize). I will fix that. Actually the collider needs lengths of transformed boundaries (i.e. lengths of Hsize column vectors), and it will also fix the 1/cos(φ) term -- it will be the same algebraically, but the angle will refer to true shew of the cell, regardless of (arbitrary) rotation.
I will do this today. > 2) The stress update is not done at each step, this is a problem since > the size of the period will be controlled based on outdated values. Where? In PeriTriaxController? Yes, but in the in-between steps, it is estimated from stiffness value (which is perhaps computed wrongly, but doesn't invalidate such approach by itself). Plus, you can control the update interval manually and set it to 1 or 2 if unsure. (Remember the strain-stress diagram I sent, with the saw-like shape as stiffness was overestimated beween true stiffness updates?) > > I think I'll commit to a branch, I have too many changes in view and I > don't want to mess with the trunk. Plese commit to trunk, I don't mind if that code is broken temporarily. At least you get testers (me) and no other people probably use that code much. Vaclav _______________________________________________ Mailing list: https://launchpad.net/~yade-dev Post to : [email protected] Unsubscribe : https://launchpad.net/~yade-dev More help : https://help.launchpad.net/ListHelp
