IMU-Based Strategies for Drift Estimation and Damage Diagnosis in Building Structures

Structural Health Monitoring has gained significant traction with the advancement of vibration-based damage detection techniques. Conventional approaches primarily utilize accelerometers to capture structural responses and derive modal characteristics, signal properties, or model-based estimations for assessing the presence, severity, and location of damage. However, such methods often face limitations in directly inferring displacement-related indicators (e.g., drift ratios) or in effectively identifying multiple damage locations using acceleration data alone. Therefore, this study introduces an inertial measurement units-based structural health monitoring framework for drift ratio estimation and damage detection, emphasizing accurate localization and quantification of structural damage. The proposed approach utilizes inertial measurement units (IMUs) to enhance conventional acceleration sensing, thereby enabling the improved capture of both translational and rotational responses. A novel modal fitting method is developed to reconstruct continuous mode shape slopes and curvatures from discrete IMU data, forming a refined structural representation that strengthens damage localization capability. Building upon this model, an integrated damage index is formulated using modal strain energy, incorporating shear- and flexure-dominant weighting factors to ensure adaptability across structural systems. A sequential filtering strategy further refines the indices for quantitative evaluation of damage severity. Validation through numerical simulations and experimental testing exhibits that the proposed IMU-based approach effectively detects and quantifies multiple damage scenarios under realistic monitoring conditions.