Objectives Composite materials have been widely applied in ship superstructures due to their advantages, including low weight, high specific strength, high specific stiffness, and excellent corrosion resistance. These properties significantly reduce the top weight of warships and enhance stealth capabilities, as exemplified by the composite superstructure of the British Navy's Type 45 destroyer. However, critical shipborne equipment (e.g., radar and communication antennas) typically have metal bases that are connected to composite frames of superstructures through mechanical means such as bolts. Under harsh marine conditions, the alternating loads from wind-wave impacts and ship motions can easily induce stress concentration around the bolt holes in composite structures. This leads to various damage modes, including fiber breakage, matrix cracking, and interface debonding. As these damages accumulate, they weaken the structure's load-bearing capacity and may even result in severe failure. Currently, research on the ultimate strength and damage assessment of large-scale steel-composite connections in ship superstructures is limited. Therefore, to optimize the performance of composite connection structures, improve structural safety in practical marine engineering, and support the future optimal design of similar composite connection structures, this study focuses on investigating the ultimate strength and progressive failure process of large composite connectors, specifically the bolted connection between ship equipment bases and composite stiffened plates.

Method This study investigates the bolted connection structure between the steel base of shipborne equipment and a composite stiffened plate, focusing on ultimate strength and material damage from two complementary perspectives: structural testing and numerical simulation. For the structural test, first, composite stiffened plate specimens were fabricated using a vacuum-assisted molding process. The plates consisted of a sandwich panel with a 3-mm-thick face sheet composed of 45°/−45°₂ glass fiber and 0°/90°/0°/90°/45°/−45°/0°/90°/45°/−45°/0°/90°/0°/90° carbon fiber, and a P100 PVC foam core. The plates also featured hat-shaped stiffeners, with an additional 6 layers of 0° unidirectional carbon fiber cloth at the top to achieve a 4 mm thickness, and a Q355 steel flange base fixed by 12.9-grade high-strength M16 bolts. Then, a test fixture was designed to apply horizontal tensile loads. The specimen was fixed by 20 M27 bolts, with a 0.3-ton column (representing the equipment weight) connected to the flange. A 25-ton actuator was used to apply the loads. Seventeen strain measurement points, including both bidirectional S-type and unidirectional D-type points, were arranged to monitor strain changes. For the numerical simulation, an ABAQUS finite element model was established to represent the skin, core material, adhesive layer, steel flange, and bolts. Composite fiber layers were simulated with SC8R continuous shell elements, the adhesive layer with COH3D8 cohesive elements, and PVC foam, flange, and bolts with C3D8R 3D stress elements. The Hashin failure criterion was employed to evaluate failure modes such as fiber tension/compression and matrix tension/compression, while the Tserpes stiffness degradation model was used to quantify damage accumulation through stiffness reduction. The progressive failure of composite fibers was simulated using the VUSDFLD user subroutine. Bolt preload (corresponding to a torque of 16.0 N∙m, equivalent to a 5 166.6 N preload) was applied by exerting pressure on the bolt head section. Additionally, a mass point (representing the 0.3-ton equipment weight) was added to apply gravity.

Results The structural test results showed that the structure reached its ultimate load-bearing capacity at a bending moment of 50.92 kN∙m. This was determined based on the complete debonding of the hat-shaped stiffeners from the panel at the bolt holes, despite the residual load-bearing capacity of the panel. During loading, the load-displacement curve remained linear (indicating elastic deformation) below 24.58 kN∙m. Beyond this value, the curve began to flatten (indicating plastic deformation), and adhesive layer failure between the stiffener and panel began. At 44.78 kN∙m, debonding occurred on the tension side of the stiffened plate. The primary failure modes included severe debonding of the adhesive layer between the hat-shaped stiffener and the sandwich panel face sheet (most pronounced on the tension side and mildest on the compression side), fiber breakage in the sandwich panel around the bolt holes, and "X"-shaped matrix damage around three bolt holes on the tension side of the flange (with the middle hole exhibiting the most severe damage). In contrast, the numerical simulation predicted an ultimate bending moment of 52.86 kN∙m, with a 3.81% error compared to the experimental result, demonstrating good consistency. The simulation also accurately replicated the structural damage, including adhesive layer debonding at the stiffener-panel interface, fiber/matrix damage around the bolt holes and flange edges, and more severe PVC core damage on the compression side. Failure initiated at 2.05 MPa for the sandwich panel core and 2.4 MPa for the stiffener core.

Conclusion This study confirms that the progressive damage analysis method—integrating the Hashin failure criterion, Tserpes stiffness degradation model, and ABAQUS finite element simulation—effectively and accurately simulates the failure process of complex bolted connection structures between ship equipment bases and composite stiffened plates. The ultimate strength results (with a 3.81% error) and the accurate reproduction of damage modes (e.g., adhesive debonding, fiber/matrix cracking) validate the reliability of the model. This research not only addresses the gap in studies of large-scale steel-composite connection structures for ship superstructures but also provides a scientifically validated analytical tool for assessing structural safety and optimizing design in marine composite applications. Future work can build on this foundation to tackle remaining engineering challenges, such as improving structural stiffness (critical for dynamic equipment) and enhancing the integrity of bolted connections.