Evaluating Measurement Bias in Flexural Modulus of Orthotropic UHMWPE Fiber/Nano-Epoxy Composites: A Finite Element Study

Reactive graphitic nanofibers (r-GNFs) dispersed in epoxy increase the resin’s flexural modulus by approximately 20% at a 0.3 wt% optimum loading, serving as matrices for ultra-high-molecular-weight polyethylene (UHMWPE) fiber composites. This study reanalyzes published datasets, constructing a traceable orthotropic material library evaluated in three-point bending via finite element analysis. The isotropic matrix cards successfully reproduce source experimental values, showing a maximum error of 0.50% at a 16:1 span and 0.020% at 32:1. However, for the unidirectional composite, the standard 16:1 span is predicted to yield an apparent flexural modulus of only 77.6–80.7% of the true longitudinal modulus. Crucially, this loss is formulation-dependent: r-GNFs increase the longitudinal modulus considerably more than the axial shear modulus, increasing the anisotropy ratio. Consequently, a 16:1 span underreports the true modulus gain at the optimum loading by 13%—an error that drops significantly when using a 32:1 span. Furthermore, we document issues in the source dataset: a lack of common fiber volume fraction across groups invalidates direct raw modulus comparisons, and the empirical model’s optimum coordinates appear transposed. Finally, a rule-of-mixtures decomposition confirms the 0.3 wt% optimum, revealing a statistically significant residual excess stiffness exclusively at this loading.