bucking - composites

[oxygen7]

Hello,

The problem which bothering me for a some time is related to buckling analisys of a composite material. In particular I am considering following
problem:
two heterogenous, linear-elastic halfplanes (2D plain strain elements) bonded with pefect connection manner and introduced analitical crack

(tangent: frictionless, normal: hard) on a interface between halfplanes. The compressive, buckling, load (non-zero displacement) is applied parallel

to the interface and a crack (in this case COD and SIF are zero).
From analitical solution one can find that for isotropic (for example: E=60GPa, n=0.30 and 170GPa, 0.25) halfplanes with introduced analitical crack

critical buckling strain are equal to 50% in addition in the case of compression along interface microcracks the critical strain does not depend on

the number of microcracks and the distance between them - all this data was confirmed by ABAQUS analisys and it means that my FE model is

appropriate.

The question starts with orthotropic materials - halfplanes (engineering constans for Graphite-Epoxy), cross-plies in a sense E1 for one half-plane

is orthogonal to the E1 for another half-plane; and E1 for one half-plane is parallel to the E1 for another half-plane. For analitical solution the

critical strain is equal to 2.8% and for FE model critical strain is equal to 5.7%. Introducing any other orthotropic materials the critical strain

for FE model is always two times larger then analitical solution.
Another interesting observation is that introducing 0.5*G12 and 0.5*G13 in FE model the critical strain is convergent with analitical solution.
Why ABAQUS gives two times larger critical strains?

Thanks in advance

oxygen7