Advanced shape sensing of a full-scale composite UAV stabilizer under non-ideal boundary conditions using the inverse finite element method

The inverse Finite Element Method (iFEM) is a robust and efficient computational technique for reconstructing full-field displacements from strain measurements, requiring only structural geometry, boundary conditions, and sensor data—without the need for material properties or external load information. While its independence from loading conditions makes it particularly suitable for aerospace Structural Health Monitoring (SHM) and Digital Twin (DT) frameworks, three major challenges limit its application to aerospace composite structures: (i) most existing studies focus on simplified structural configurations rather than realistic, full-scale assemblies, (ii) sparse sensor layouts, which reduce reconstruction accuracy, and (iii) uncertainty or degradation of boundary conditions, which can significantly compromise shape sensing performance. This study presents a novel combined numerical–experimental methodology for advanced shape sensing of a full-scale composite Unmanned Aerial Vehicle (UAV) stabilizer under non-ideal boundary conditions. The approach integrates Smoothing Element Analysis (SEA) into the iFEM framework to pre-extrapolate strain fields from limited sensor measurements. Given the structural complexity of the stabilizer, SEA is applied on different planes of the structure, starting from instrumented areas and progressively extending to non-instrumented regions. Furthermore, the methodology explicitly accounts for non-ideal boundary constraints arising from joint loosening and support compliance. The methodology is experimentally validated on a full-scale composite stabilizer equipped with a LUNA distributed fiber optic sensor network, with reconstruction accuracy assessed under various ideal and non-ideal support configurations. Several experimental tests are performed in which the stabilizer is subjected to two external forces, inducing bending and torsion in a cantilever beam configuration. Results show that, when only ideal boundary conditions are assumed, displacement reconstruction errors can exceed 14 % at key measurement points, which is not acceptable for SHM purposes. By explicitly modeling joint loosening and support compliance, reconstruction errors are reduced to approximately 2 %, achieving the accuracy required for real-world applications. These findings highlight the potential of the proposed SEA–iFEM approach for real-time monitoring of complex composite aircraft structures in realistic operational scenarios, where non-ideal boundary conditions cannot be neglected.