An experimental campaign on glass-fiber/aluminum laminated specimens was conducted to assess the interlaminar fracture toughness of the metal/composite interface. Asymmetric end-notched flexure tests were conducted on specimens with different fiber orientation angles. The tests were also modeled by using two different analytical solutions: a rigid interface model and an elastic interface model. Experimental results and theoretical predictions for the specimen compliance and energy release rate are compared and discussed.Metals 2020, 10, 56 2 of 17 the elastic interface model including bending-extension and shear deformability [15]. Valvo analyzed the delamination of shear-deformable laminated beams with bending-extension coupling based on a rigid interface model [16]. Tsokanas and Loutas extended the above-mentioned analysis to include the effects of crack-tip rotations and hygrothermal stresses [17]. Bennati et al. furnished a complete analytical solution for a crack-tip element made of two multidirectional laminated beams connected by an elastic-brittle interface [18].Moreover, many experimental investigations have been carried out to characterize the delamination behavior of FMLs [19][20][21][22]. Cortés and Cantwell conducted single cantilever beam (SCB) tests on magnesium alloy based FMLs [19]. Abdullah et al. carried out similar tests on glass fiber-reinforced polypropylene (GFPP) based FMLs [20]. Bieniaś et al. conducted end-notched flexure (ENF) tests on CARALL and GLARE laminated specimens and used Valvo's rigid interface model to interpret their results [21]. Bieniaś and Dadej extended the previous research to investigate fatigue delamination growth [22].However, it is long known that a rigid interface model, which neglects the relative rotation and deflection between sublaminates in the neighborhood of the crack-tip, may underestimate the ENF specimen compliance with respect to experimental tests and therefore lead to a wrong evaluation of the energy release rate [23]. To overcome this drawback, Sundararaman and Davidson introduced rotational springs at the crack-tip in a theoretical model of bi-material ENF test specimens [24]. Yang and Sun tested and modeled multidirectional laminated asymmetric end-notched flexure (AENF) test specimens accounting for bending-extension coupling, but did not partition the energy release rate into its mode I and mode II contributions [25]. To the best of our knowledge, in the available literature, there are no specific analytical solutions for multidirectional laminated asymmetric ENF tests that fully account for the bending-extension coupling and mixed-mode fracture conditions.In this paper, we present the results of AENF tests on multidirectional glass fiber-reinforced polymer/aluminum (GFRP/Al) specimens and show how these can be interpreted based on both the rigid [16] and elastic [18] interface models. Section 2 describes the geometry and material properties of the tested specimens. Section 3 presents the experimental methods and the theoretical models adopted for...