Evaluation on torsional moment of interior wall-slab joints in reinforced concrete box-shaped wall structures with thick wall and slab
Jiehui Wang, Zelin Wang, Koichi Kusunoki, Zhibin Li, Enrique del Rey Castillo, Tamon Ueda
Accurate evaluation of torsional resistance at wall–slab joints is essential for assessing the punching shear safety of reinforced concrete box-shaped wall structures with thick walls and slabs (TWS structures), particularly under seismic loading. However, existing design approaches for conventional slab–column connections cannot be directly extended to TWS joints because of their substantially different wall geometry and reinforcement configuration. This study develops a practical framework for evaluating the torsional moment at the damage limit and the torsional strength of interior TWS wall–slab joints based on Nadai’s Sand Hill Analogy. Three-dimensional nonlinear finite-element models, validated against available experimental results, were used to isolate the local torsional response and investigate the effects of wall length and slab transverse reinforcement. The results demonstrate that Nadai’s Sand Hill Analogy provides a reasonable basis for evaluating the torsional resistance of thick wall–slab joints. For the representative models with a 500 mm wall length, the torsional moments at the damage limit were 7.72 and 7.74 kN·m without and with slab transverse reinforcement, respectively, indicating that transverse reinforcement has little influence before significant cracking. In contrast, the corresponding torsional strengths increased from 16.62 to 30.10 kN·m, representing an increase of approximately 81% when transverse reinforcement was included. After cracking, the reinforcement therefore substantially enhances torsional strength and deformation capacity through combined reinforcement restraint, crack bridging, shear-transfer redistribution, and dowel-related mechanisms. The simplified localized dowel-force component accounts for approximately 5.65%–12.93% of the strength enhancement within the investigated models, indicating that the reinforcement contribution cannot be represented by dowel action alone. Parametric analyses covering a wall-length-to-thickness ratio