This study investigates the influence of pre-crack conditions (introduced under Mode I and Mode II loading prior to fracture testing) and specimen compliance on the Mode II fracture characterization (G_IIC) of adhesively bonded composite joints. Calibrated End-Loaded Split (CELS) and 3-Point-Bending ENF tests were performed using structural AF163-2 K adhesive. Various data reduction schemes were employed to account for pre-crack morphology and compliance in the development of the R-curve. The data reduction schemes showed significant scatter, ranging from 8.07 ± 0.17 to 17.3 ± 1.19 N/mm, depending on the pre-cracking conditions and compliance effect. Mode I pre-cracked specimens consistently exhibit higher G_IIC values compared to Mode II pre-cracked specimens, a difference governed by the morphology and extent of the fracture process zone (FPZ). Mode I pre-cracking forms a localized FPZ that subsequently transitions into a shear-dominated FPZ for G_IIC evaluation during the subsequent Mode II fracture test. In contrast, Mode II pre-cracked specimens contain an already-developed shear FPZ that is broader and more diffuse, resulting in lower strain-energy release rates and lower G_IIC values. High compliance effects cause significant bending, additionally introducing high derogatory energy deformation from the test fixtures, obscuring the actual crack tip. The apparent crack length methods demonstrated reliable estimates of fracture energy and R-curve behavior by accounting for the effects of large FPZ, thereby capturing both crack-tip and distributed dissipation mechanisms. The experimental findings correlate with computational results, displaying stable cohesive disbond growth in the adhesive layer. This study indicates that pre-cracking and compliance effects significantly influence Mode II fracture characterization and, therefore, need to be properly addressed.

The fracture process zone (FPZ) significantly influences the damage tolerance of adhesively bonded composite joints, governing crack-growth mechanisms and migration. Existing fracture characterization approaches generally evaluate pure-mode behavior independently and extend these results to mixed-mode conditions using a power-criterion, such as the Benzeggagh-Kenane (B–K) criterion. This process assumes that FPZ-dependent mode-mix behavior from a standard mixed-mode test is transferable to another complex loading condition. This assumption remains unchecked for toughened adhesive joints, where FPZ morphology varies with loading conditions. This study addresses this gap through experimental and numerical investigation using digital image correlation (DIC) and cohesive zone modeling (CZM). The pure mode I test displayed localized FPZ ahead of the crack tip, influenced by carrier bridging. Two different pure Mode II tests demonstrated that the apparent crack length method accurately accounts for the large FPZ ahead of the crack tip. The mixed-mode bending (MMB) test linked pure modes through the B–K criterion. The Crack-Lap Shear (CLS) specimens exhibited evolving FPZ and mode II-dominated fracture. The fracture toughness predicted by the B–K criterion deviated from the CLS tests as the loading became more mode II dominant. It was observed that the FPZ morphology during the CLS test differed significantly from that observed during the MMB test, through DIC and CZM. These results highlight that differences in FPZ affect the mixed-mode fracture toughness and demonstrate the limitations of applying a single empirical power-criterion. It underscores for FPZ-sensitive approaches to accurately predict the fracture resistance of toughened adhesive joints under evolving mixed-mode conditions.