Structural health monitoring using sparse distributed networks of guided wave sensors

Abstract

The motivation for using guided acoustic waves as the sensing mechanism for large area structural health monitoring (SHM) is explained and the logic for using baseline signal subtraction as the fundamental signal processing tool is presented. In the first part of this paper, a simple experimental example is presented to illustrate how a guided wave SHM using baseline subtraction could be used to detect and locate simulated damage in a 1 m by 1.5 m by 3 mm thick aluminum plate. The experiment shows for an SHM system to be useful it must have a coherent noise floor around 40 dB lower in amplitude that the amplitude of a signal reflected from the edge of the structure. The experiment demonstrates that the sensitivity is severely limited by the stability of the baseline subtraction procedure which deteriorates rapidly over time. In the second part of the paper, the factors affecting the stability of the reference signal subtraction approach are investigated. Experimental and modeling studies on a simple test structure are presented that show that a change in temperature of a few degrees leads to coherent artifacts after baseline subtraction that are of a similar magnitude to the signals arising from defects. A possible strategy for overcoming this barrier to reliable baseline signal subtraction is then considered and shown to provide an improvement in sensitivity of around 10 dB.