This paper proposes a structured framework for Network Digital Twins (NDTs) in 6G, addressing the lack of formal definitions and standardised architectural guidelines in the field. By refining the conceptual foundations of NDTs, it introduces a functional architecture, clarifies key components such as AI-driven workflows, a simulation framework, data management, and orchestration, and provides concrete examples illustrating their role in network automation, optimisation, and predictive analytics. The goal is to offer a cohesive reference model that guides the community in shaping NDTs development, ensuring interoperability, scalability, adaptability, and seamless integration into AI-native 6G networks for improved intelligence and efficiency.