Unlike the standardized classifications applied to traditional fixed-wing and rotary-wing aircraft, low-altitude aircraft are evolving towards greater flexibility and intelligence in system configuration. A prime example is the Electric Vertical Take-Off and Landing (eVTOL) aircraft, which utilizes pure electric propulsion and achieves runway‑independent operations through adaptive control of aerodynamic surfaces. This capability significantly expands potential application scenarios but imposes demanding control requirements on the avionics system across distinct flight phases. Consequently, avionics system architectures are shifting from single‑mode to multimodal configurations to accommodate these dynamic needs. Effectively modeling and assessing the safety of such multimodal avionics systems within a Model‑Based Systems Engineering (MBSE) framework has thus become a critical focus for both academia and industry. To address the safety assessment challenges arising from state transitions in multimodal avionics systems, this paper adopts an MBSE methodology. Integrating the forward‑design philosophy of civil aviation and employing the Systems Modeling Language (SysML), a digital design chain encompassing requirements, functions, and architecture is established. Following a “black‑box opening and hierarchical decomposition” approach, the study develops static architectural models of multimodal avionics systems and dynamic behavioral models under typical flight missions. Furthermore, it extends failure mode analysis across different transition states, creating a coherent modeling pathway from requirements to architecture. This overcomes the limitation of existing methods that often focus on isolated failure modes. The proposed approach is validated through a case study on the avionics state transitions of a tilt‑rotor eVTOL aircraft during flight. The research provides a systematic engineering solution for the safety assessment of next‑generation low‑altitude aircraft.