Damage detection and quantification in deck type arch bridges using vibration based methods and artificial neural networks. (January 2020)

• Record Type: Journal Article
• Authors: Jayasundara, N.; Thambiratnam, D.P.; Chan, T.H.T.; Nguyen, A.
• Is Part Of: Engineering failure analysis. Volume 109(2020)
• Publication Date: 2020-01
• Subjects: Bridge failures -- Arch bridges -- Vibration based damage detection (VBDD) -- Artificial neural networks (ANN) -- Non-destructive testing; System failures (Engineering) -- Periodicals; Fracture mechanics -- Periodicals; Reliability (Engineering) -- Periodicals; Pannes -- Périodiques; Rupture, Mécanique de la -- Périodiques; Fiabilité -- Périodiques; Fracture mechanics; Reliability (Engineering); System failures (Engineering); Periodicals; Electronic journals; 620.112
• DOI: 10.1016/j.engfailanal.2019.104265
• Languages: English
• ISSNs: 1350-6307

Abstract

Highlights: Health monitoring is important for arch bridges for their reliable operation. The proposed method treats damage in arch ribs and vertical columns of arch bridges. Modified modal flexibility and strain energy methods are used to detect and locate damage. Artificial neural networks and fusion method provide effective damage quantification. Abstract: Vibration based methods can be used to detect damage in a structure as its vibration characteristics change with physical changes in the structure. Extensive research has been carried out on the use of such methods to detect damage in a number of simple and some complex structures. Arch bridge is a popular type of bridge with rather complex vibration characteristics which pose a challenge for using existing vibration based methods to detect damage in them. Further, its complex form of damage detection, even with modified vibration based methods makes the quantification process harder and challenging. This paper develops and applies a vibration based method especially suited for arch bridges to detect, locate and quantify damages in the structural components. In the proposed method, modified forms of the modal flexibility (MMF) and modal strain energy (MMSE) based damage indices coupled with the Artificial Neural Network (ANN) technology is used to provide an overall damage assessment. The procedure to detect and locate damage was experimentally validated and applied to a full scale long span arch bridge under a range of damage scenarios. Damage indices obtained from noise polluted vibration data are then used as input data for training and validation of the neural networks. Two neural networks were trained separately using MMF and MMSE damage indices and a network fusion approach is used to obtain unambiguous and accurate results for detecting, locating and quantifying damages. The trained neural network system was then successfully applied to identify unknown damages using only vibration data of damaged structural elements of arch bridges. The findings of this paper will contribute towards the safe and efficient operation of arch type bridges.