The increasing functional complexity of composite structures requires material selection approaches that extend beyond traditional property-centric methods. This work presents a Structure–Property–Function (SPF) framework for the function-oriented design of composite blades, demonstrated through carbon nanotube (CNT)-doped adhesive systems. The framework integrates representative volume element (RVE) modelling, microstructure-informed finite element simulations, and a Functional Area Clustering strategy to link material architecture directly to regional functional requirements. Mechanical and electrical properties were computed across 72 CNT–matrix combinations, revealing systematic trends in stiffness enhancement, percolation-driven conductivity, and sensitivity to CNT aspect ratio and volume fraction. Mapping these combinations onto multi-functional design spaces enabled the identification of viable material candidates for five critical blade regions, highlighting design trade-offs between stiffness, insulation, EMI shielding, and recyclability. By shifting the emphasis from material benchmarking to function-driven design, the proposed SPF framework offers a generalizable and scalable methodology for developing multifunctional composite components. The approach provides a foundation for future integration of optimisation and experimental validation, supporting the development of next-generation composite structures in renewable energy and beyond.