Recovery of Interlaminar Tensile Stresses (ILTS) in Curved Laminates Subjected to Bending: Development and Validation of a Computationally Efficient Numerical Method

Curved composite laminates are widely employed in aerospace and automotive structures due to their high specific stiffness and strength-to-weight ratio. However, the presence of geometric curvature significantly amplifies out-of-plane stress components, particularly interlaminar tensile stresses (ILTS), which are a primary driver of delamination. Despite their importance, ILTS are not directly available in standard shell finite element formulations, which remain the preferred modeling strategy in industrial practice due to their computational efficiency. This thesis proposes and validates a computationally efficient numerical post-processing methodology for the recovery of interlaminar tensile stresses in curved laminates subjected to bending. The approach is grounded in laminate theory and shell equilibrium equations, and combines symbolic derivation with numerical implementation. A complete workflow is developed: from theoretical formulation and symbolic manipulation in Maxima, to numerical implementation in Octave, including through-thickness integration, energetic assembly, kernel identification, static condensation, and field reconstruction. The proposed method reconstructs the interlaminar normal stress component $\sigma_{zz}$ from shell-based strain and curvature fields without requiring full three-dimensional solid discretization. Validation is performed against high-fidelity 3D finite element models under butterfly bending loading for multiple material configurations, including isotropic aluminum, quasi-isotropic CFRP, sandwich, and "anti-sandwich" laminates. Additional comparison with reference results from literature further confirms the reliability of the approach. Results demonstrate that the recovered ILTS distributions closely match solid FEM predictions within a moderate thickness-to-radius range, while significantly reducing computational cost. The methodology enables accurate interlaminar stress estimation within shell-based industrial workflows, thus bridging the gap between computational efficiency and delamination-sensitive design. The proposed framework represents a robust and extensible strategy for interlaminar stress recovery in curved composite structures and provides a solid foundation for future developments involving complex loading conditions and damage modeling.