Structural Assessment Based on Computational Models and Digital Twin Technologies

<p>Modern structural design and assessment of aerospace and automotive structures relies on numerical modelling for the analysis of complex components under diverse loading conditions and material properties, reducing the reliance on costly experimental testing. However, these models often struggle to fully replicate the intricate real-world behavior of structures. Discrepancies arise due to factors like variable loading, environmental effects, material properties scatter, manufacturing imperfections, and other effects that can only fully observable in operation. This paper explores the potential of digital twins to bridge this gap, enhancing the accuracy and reliability of structural assessments. As virtual replicas integrating numerical models with sensor data, operational history, and machine learning, digital twins offer a dynamic and comprehensive understanding of structural behavior throughout a structure's lifecycle. Approaches to enable model updating, real-time structural health monitoring, and the development of new methods for model calibration and validation that combine simulations with experimental data are proposed in this paper. This path can revolutionize design, evaluation, and maintenance processes, optimizing structures for performance and lightweight while upholding rigorous safety factors.</p>