The application of classical fracture mechanics to laminated com posites is discussed. A convenient method is presented for predicting the static strength of a flawed specimen. Theoretical predictions are compared with experimental data for specimens containing two types of flaws.

Strength characterization of the fiber-reinforced laminated composite materials is quite often accomplished through the utilization of the finite-width tension coupon specimen. Recent analytical studies have shown that the state of stress in the vicinity of the free edge of such a laminate is fully three dimensional in nature and may not be predicted by laminated plate theory. Furthermore, experimental results have revealed that while the strength of angle-ply tensile coupons of certain fiber orientations are predicted adequately by laminate strength theories, coupons of other fiber orientations are not. As an explanation of this result, it is argued that large, nonlinear shear strains occur at the interfaces for fiber orientations leading to near maximum shear coupling compliances. Further, these strains allow the initiation of matrix cracks at the free edge, which are ultimately responsible for the premature rupture of the coupon. Experimental evidence which supports these arguments is presented.

This paper demonstrates the statistical relationships which exist between test specimens of different designs and how these relationships can be utilized in strength predictions. The specimens selected were the straight-sided tension coupon, uniaxial sandwich beam, and the acceptance-type flexure coupons. Experimental data were obtained for the [0]c, [90]c, [02/±45]c, [±45/902]c, and [0/±45/90]c laminates fabricated from Narmco 5505 boron-epoxy at various temperatures ranging from -65 to 420 F.

A discussion of several failure criteria from their initial application to homogeneous, orthotropic materials to their extension to “quasi-homogeneous,” anisotropic materials is presented. Also discussed are the assumptions, general limitations, and the physical and analytical significance of the respective “material constants.” The basic criteria include the Lame'-Navier maximum stress theory, Henky-von Mises distortional energy theory, Tresca maximum shear stress theory, and the St. Venant maximum strain stream theory.