Time-Domain Damage Detection of Structures with Complex Modes Under Variable Environmental Conditions Using Bayesian Virtual Sensors

Abstract

A time-domain algorithm for damage detection is introduced. It is based on hardware redundancy in which the number of sensors is greater than the number of excited modes plus the number of environmental variables. Therefore, for a structure with complex modes, the minimum number of sensors is expected to be higher than that of the same structure with real modes. A two-step detection algorithm is proposed. First, the accuracy of each sensor is increased by Bayesian virtual sensing. Second, the signal of each sensor is estimated using the remaining sensors utilizing a correlation model of the training data under different environmental conditions. The residual is used to detect damage. The algorithm was studied in a numerical experiment of a frame structure having a discrete damper element, which resulted in complex mode shapes. A comparison was made with the same structure having real modes due to proportional damping. The performance of damage detection was higher with real modes and virtual sensors outperformed the raw measurements. Damage localization was also relatively successful revealing the region close to the actual damage.