Experimental Identification of Statistical Correlation between Mechanical Properties of FRP Composite

Abstract

Recent prediction on the heavy statistical correlation between the mechanical properties of fiber reinforced composite (FRP) raises new concerns on the accurate reliability evaluation of composite structures, but such statistical correlation still lacks experimental verification. In this work, an experimental methodology is proposed to determine the statistical correlation between mechanical properties of unidirectional FRP composite. A rectangular shaped carbon fiber reinforced plastic (CFRP) specimen with a circular hole is loaded by tension, and 3D digital image correlation (DIC) is employed to characterize the heterogeneous strain field around the hole. Virtual field method (VFM) is used to derive E11, E22, ν12, and G12 of specimens. Specimen configuration considering fiber angle and hole diameter is optimized to achieve accurate determination of correlation coefficients. Experimental results on the linear correlation coefficients between E11, E22, ν12, and G12 agree well with previous theoretical predictions.