Infrasound-Based Noncontact Sensing for Bridge Structural Health Monitoring

Abstract

This study demonstrates the use of infrasound measurements from microphones as a means of noncontact sensing to capture the dynamic properties of structures for structural health monitoring (SHM). A pilot study using an in-service highway bridge in Connecticut is conducted to compare infrasound and accelerometer-based SHM using a frequency domain peak-picking method. A three-dimensional finite-element (FE) model is developed to validate the results. Potential benefits and limitations of infrasound-based SHM are discussed. An attention model from machine learning is further proposed to increase the signal-to-noise ratio of the microphone measurements and provide an unbiased rapid means of identifying the modal frequencies of the bridge.