Perpustakaan Digital ITB

Transverse cracking is one of the earliest forms of damage in fiber-reinforced composite laminates, and its progressive multiplication often precedes more severe failures such as delamination and fiber fracture. Conventional fracture models, however, require crack paths to be defined in advance and dissipate fracture energy in a mesh-dependent manner, which limits their ability to capture how cracks spontaneously form and multiply. This study develops a phase-field fracture model to analyze the initiation, propagation, and multiplication of transverse cracks in cross-ply laminates, and evaluates it against cohesive-zone model and experimental results. The model was implemented in Abaqus through a user-material subroutine, extended from an isotropic baseline to include orthotropic elasticity, ply-orientation transformation, and an undamageable 0° ply. The AT2 formulation was solved monolithically, with a spatially random fracture-toughness field assigned to the transverse plies. A glass-fiber/polypropylene [0/90]? laminate at 45% fiber volume fraction was loaded in uniaxial tension, and the subroutine was verified on simple benchmarks and a mesh-convergence study. The model reproduced the experimental stress–strain response well, and captured transverse crack initiation, propagation, and multiplication without any predefined crack paths. The predicted initial modulus fell between the rule-of-mixtures and classical-laminate-theory estimates, deviating from both by less than 1.3%, and its continuous stiffness decay followed the experimental trend more closely than the cohesive-zone model at higher strains. Overall, the approach offers a reliable, mesh-objective framework for predicting transverse damage evolution in fiber-reinforced composites.