Parametric Sensitivity of Axial–Flexural Interaction in Reinforced Concrete Shear Walls for Optimized Design and Structural Efficiency

Purpose: This study develops a code-agnostic, mechanics-first framework to quantify the parametric sensitivity of axial–flexural (N–M) interaction in reinforced concrete (RC) shear walls and to produce transferable rankings of key design knobs controlling N–M response metrics. Design/methodology/approach: A strain-compatibility, fiber-based sectional solver is formulated for rectangular, T, I/H, and U-shaped wall sections. The mechanics engine is decoupled from a modular code-profile layer (ACI-/EC2-/BS-consistent mappings) to enable cross-code comparisons. Interaction curves are normalized to isolate mechanics-driven shape effects, and scalar metrics are extracted at multiple axial levels. Sensitivity is quantified using local elasticities, Morris screening, and Sobol variance-based indices; numerical reproducibility is verified through convergence and mesh-independence controls. Findings: Normalization collapses most cross-code variability in curve shape, while design-level curves retain separable safety-format differences. Sensitivity rankings vary with axial level and section family, revealing nonlinear interactions and regime shifts especially for irregular sections where flange participation can dominate near peak or balanced states. Practical implications: The framework supports cross-code traceability and efficiency-based design guidance across axial regimes. Originality/value: The study delivers a reproducible, code-agnostic N–M engine integrated with rigorous global sensitivity analysis to bridge the gap between sectional mechanics and design-oriented decision-making.