Damage Detection in FGM Plates Using Modal Strain Energy and Gaussian Process Regression

The extensive use of composite materials, particularly Functionally Graded Materials (FGMs), across a wide range of industrial applications has generated a critical demand for robust and reliable structural health monitoring (SHM) and damage detection systems to ensure operational safety and long-term structural integrity. In this paper, a numerically validated comprehensive SHM framework is proposed for FGM plates through finite element investigations, based on modal strain energy analysis and employing isogeometric analysis (IGA) for accurate dynamic response evaluation. The proposed damage detection methodology exploits variations in modal strain energy characteristics induced by structural degradation. In the first stage, a damage detection index derived from modal strain energy distributions is used to efficiently and accurately identify the location of damage within the structure. In the second stage, the identified damage indices are employed as input features within a Gaussian-Process-Regression (GPR)-based framework, while the corresponding damage-severity levels are treated as outputs, enabling reliable damage quantification for FGM components. A series of numerical damage scenarios are systematically investigated to assess the performance of the proposed approach. The numerical results demonstrate the high predictive accuracy and effectiveness of the proposed methodology under the investigated damage scenarios and simulation conditions, highlighting its potential for vibration-based damage assessment of FGM plates.