PLAYING WITH THE ANISOTROPY OF THE MATRIX AND FIBER FOR THE IMPROVEMENT OF THE STRENGTH OF COMPOSITES

The paper deals with the theory of the coupling of strongly anisotropi c fibers along their axis with anisotropic matrices along either the f iber direction or the transverse plane to the direction of the fibers. It is shown that the first arrangement deteriorates, whereas the seco nd improves significantly the mechanical behavior of the composites. I ndeed, it is shown that anisotropy of the matrix, increasing its mecha nical properties on the transverse isotropic plane of the composite, i ncreases also the transverse-transverse Poisson's ratio, nu(23), where as decreases the longitudinal shear modulus, G(12). This results in va lues of the eigenangle omega(c) approaching the corresponding value om ega(ic) for the equivalent isotropic material and caused a deteriorati on of the strength and toughness of the composite, since the material now has the tendency to develop higher stress concentrations for equiv alent loadings [1,2]. On the contrary, a strong anisotropic matrix alo ng the direction of the fibers yields the inverse results for the resp ective moduli of the anisotropic composite. Finally, the most importan t result is a perceptible decrease of the ratio E(11)/2G(12) which yie lds values of the eigenangle omega(c) tending to approach the critical value omega(i) for the isotropic material. The decrease of omega(c) i ndicates the improvement of the quality of the composite, since it dev elops relatively lower stress concentration factors, which approach th eir respective isotropic values [2]. Thus, the anisotropic composite m aterial is approaching an equivalent state of quasi-isotropy and there fore improves its strength by reducing considerably the eventual aniso tropic stress concentration factors of the respective structural eleme nts [1]. Examples with T300/N5208 graphite-epoxy composites and Borsic -1100 aluminum metal-metal composites indicate clearly the beneficial effect of the anisotropy of their matrices.