The bendability of extruded aluminium alloys is a critical property in automotive design. Numerical simulations can be used to provide key insights into their bending performance. In this study, we investigate the bending behaviour of three strongly textured aluminium alloys---extruded AA6060-T6, AA6082.25-T6, and AA6082.50-T6---using numerical finite element simulations of the bending test VDA238-100. This test is widely used for material characterization in automotive applications. To model anisotropic plasticity and ductile failure, we employ a porous plasticity framework that combines the Yld2004-18p anisotropic yield criterion with the Gurson-Tvergaard-Needleman (GTN) model [1]. In the simulations, the critical porosity in the GTN model is determined from the material softening condition, as proposed by Morin et al. [2] and used by Espeseth et al. [3]. Thus, the softening-based GTN model dynamically adjusts the critical porosity based on the stress state, enabling controlled softening and ductile failure predictions while reducing the number of model parameters. Additionally, we calibrate the anisotropy parameters for generic Cube and Goss textures and assign them in the finite element model of the VDA test with randomly varying material orientations. This serves as a simplified and cost-efficient alternative to full crystal plasticity models. Our results illustrate the significant influence of texture and porosity on the bendability of these alloys, contributing to a better understanding of their mechanical performance in automotive applications.