Multi-input Multi-output Loewner Framework for Vibration-based Damage Detection on a Trainer Jet

원문 초록

Structural health monitoring of aerostructures often faces challenges identifying damage, especially in complex systems. Multi-input multi-output modal parameter identification methods are known to offer enhanced insight compared to single-input multi-output testing, as they allow for the identification of additional out-of-plane modes. The improved Loewner Framework presents a computationally efficient approach to extracting these modal parameters, focusing on natural frequencies and mode shapes as indicators of structural health. To address the challenges of damage detection, a numerical case study involving a cantilever beam with variable cross-sections is used to simulate various damage scenarios. Additionally, a full-scale experimental dataset from the BAE Hawk T1A trainer jet aircraft is employed for SHM for the first time. The modified total modal assurance criterion (MTMAC) is proposed as a standalone metric for assessing damage severity, while the coordinate modal assurance criterion (COMAC) is applied for localising damage. Benchmarking against methods such as least-squares complex exponential (LSCE) and stochastic subspace identification with the canonical variate analysis (SSI-CVA) demonstrates the effectiveness of the improved Loewner Framework in accurately identifying even small changes in modal parameters. The MTMAC and COMAC are shown to be valuable tools for, respectively, damage quantification and localisation.