Effect of Stress Concentration near Grips on the Tensile Strength of Composites

Abstract

The strength of a unidirectional composite along the fibers is the most important characteristic for the design calculation of composite products, being, however, most difficult for correct determination in an experiment. The main problem is attributed to the shape and methods of fixing the specimen providing the minimum impact of the stress concentration near the grips on the strength. The impossibility of direct use of standard self-tightening grips because of lateral crushing of the specimen led to the necessity of using either specimens with glued pads or special grips with a constant transverse force. Specimens with fillets should not be used for unidirectionally reinforced plastics. When using standardized specimens in the form of rectangular strips with abruptly varying shear stresses applied to their lateral surface, which results in a significant concentration of tensile stresses, an incorrect result can be obtained. We present the calculated values of the stress concentration factor (SCF) near the grips using a simplified model of shear analysis and the finite element method (FEM). Since the strength reduction factor is always less than the theoretical stress concentration factor, a phenomenological parameter of the material with the dimension of length as a characteristic size of the stress averaging zone should be introduced to assess the strength value. Three methods are proposed to find this characteristic size according to the results of testing specimens of different kind: (1) smooth with various thicknesses; (2) smooth with various thicknesses and with holes in which the failure occurs at the hole, away from the grips; (3) with a series of holes with a decreasing radius, to find that small radius the influence of which on the decrease in strength is equivalent to the influence of grips. Proceeding from the experimentally determined characteristic size, the dependences of the strength on the specimen thickness, on the ratio of Young’s and shear modulus, on the length of the grips and the working zone, and on the degree of stress smoothing near the grip edges are calculated. The proposed calculation and experimental technique provided more accurate assess of the “true” tensile strength of the composites, which can be approximately 10% higher than that determined on the specimens of large thickness. Taking this correction for the effect of grips into account, it appeared possible to reduce the safety factor and, hence, to reduce the mass of composite structures.