PSI - Issue 77

Tianyu Wang et al. / Procedia Structural Integrity 77 (2026) 512–520 Wang et al./ Structural Integrity Procedia 00 (2026) 000 – 000 the performance changes significantly. The failure coefficient for the [+75°/−75°] 2 layup rises to 0.55, an 57% increase, indicating the inherent disadvantage of high-angle fibres in resisting shear stress generated by torque. Meanwhile, the [+45°/−45°] 2 layup exhibits remarkably stable performance, with failure coefficients decreasing from 0.97 to almost 0.94 over the same torque range. This stability derives from the fortuitous alignment of ±45° fibres with the principal stress directions under combined pressure-torsion loading. The medium-angle configuration essentially transforms the biaxial stress state into a fibre-dominated response, maximising the utilisation of the composite ’ s superior fibre properties. The [+15°/−15°] 2 layup consistently shows the highest failure coefficients (11.6 to 12.1), primarily due to its inability to efficiently resist the substantial hoop stresses from internal pressure whilst also lacking adequate shear stiffness for torsional loads. A crucial observation emerged regarding interlaminar stress evolution: the variation in torsional load disproportionately increases interlaminar shear stress, with peak values occurring at ply interfaces with maximum angle mismatch. This finding suggests that torsional loads may trigger delamination failures before in-plane failure occurs, particularly in designs with large differences in adjacent ply angles, a failure mode often overlooked in pressure-vessel design codes developed primarily for static loading conditions [2, 14]. The research findings demonstrate that no universally ‘ optimal ’ winding angle exists. This fundamental insight challenges the traditional approach of selecting ‘ standard ’ angles (typically ±55° ) [7] regardless of loading conditions. Instead, optimal design requires a thorough understanding of the operational loading and careful matching of fibre orientation to principal stress trajectories. 517 6

Fig. 4. Effects of stacking sequence on failure coefficients

The second phase of the parametric studies studied the effects of stacking sequence under high torque (30 kN·m), representing severe operational scenarios. That is particularly important because the stacking sequence can be modified without changing material costs or manufacturing complexity, offering a ‘ free ’ optimisation parameter. For the [15°/45°] angle combination, two sequences were compared: [+15°/−15°/+45°/−45°] versus [+45°/−45°/+15°/−15°] , see Fig 4. The analysis revealed that positioning the 45° plies on the interior, closer to the pressure-loaded surface, reduced peak failure coefficients by 22%. This improvement occurs because the internal placement allows the 45° plies to carry a larger portion of the hoop stress whilst maintaining their contribution to torsional resistance. Thus, this arrangement creates a more gradual stress transition through the thickness, reducing stress concentrations at ply interfaces. Similarly, for the [75°/45°] combination, comparing [+75°/−75°/+45°/