Structural Damage Detection Using the Modal Assurance Criterion – Differential Evolution (MAC-DE) Method

Abstract

In the field of civil engineering, structural damage detection has become an important concern due to increasing demands for safety and cost-efficient maintenance. In general, the process of structural damage identification is divided into four levels: (1) detecting the presence of damage, (2) determining the affected components or elements, (3) estimating the location and the reduction in stiffness or strength of the damaged elements, and (4) predicting the remaining service life of those elements.This study aims to develop and evaluate a structural damage detection method at the third level by combining the Modal Assurance Criterion and the Differential Evolution algorithm (MAC-DE), enabling accurate identification of both the location and the magnitude of stiffness reduction in damaged structural elements. Prior to the implementation of MAC-DE, the Damage Location Vector approach (VBPLK) is first employed to identify areas potentially experiencing degradation. Two two-dimensional portal frame models are tested under various damage scenarios to assess the effectiveness of the proposed method. Based on numerical simulation results, the MAC-DE approach is proven to accurately identify both the position and the degree of strength and stiffness reduction in damaged structural elements.