Towards Certification-Ready Designs

Abstract

As aerospace and automotive industries increasingly adopt advanced materials and manufacturing methods, traditional certification processes, which rely on safety factors, historical data, and physical testing, struggle to address the complex failure modes of new materials like custom composites and additively manufactured structures. Digital twins—virtual replicas of physical systems— can enhance certification workflows by enabling continuous monitoring and providing real-time insights into system performance through the integration of sensor data with physics-based models. This approach has the potential to improve failure predictions, optimize performance, and reduce over- engineering, supporting more efficient and lightweight designs. This study presents the development and validation of a digital twin for a sensorized Unmanned Aerial Vehicle (UAV), focusing initially on a propeller boom arm mounted on a test rig for real-world testing in a controlled environment. The digital twin leverages both physics-based analysis provided by surrogate models of Autodesk Nastran and real-world sensor data. We perform a three- way comparison between Nastran, the digital twin, and sensor data to validate both the hardware and software setups, with ongoing efforts to reduce discrepancies and improve sensor placements. Our initial results highlight the potential of digital twins to significantly accelerate certification processes, reduce costs, and enable the faster adoption of new materials, ultimately driving innovation and transforming engineering practices across industries.