Noncontact laser ultrasonic wavefield imaging and laser lock-in thermography techniques for structural damage detection

Abstract

Laser ultrasonic and infrared (IR) thermography techniques have great potential for noncontact, nondestructive, noninvasive and instantaneous detection of a structural damage. In particular, their wavefield imaging with high temporal and spatial resolutions enables to identify and localize even an incipient damage. This dissertation aims to develop new laser ultrasonic wavefield imaging (LUI) and laser lock-in thermography (LLT) techniques for instantaneously automated detection of structural damages and validate their applicability for various target structures and damage types. The main objectives of the proposed techniques are listed as (1) noncontact damage detection, (2) reference-free damage diagnosis for automated and instantaneous damage alarms, (3) damage visualization for damage identification and localization and (4) adaptable and versatile damage detection techniques for real structures. These objects are able to significantly contribute to the evolution of nondestructive testing (NDT) to structural health monitoring (SHM). To achieve them, the LUI and LLT techniques are developed and validated through the following four steps: (1) development of noncontact LUI and LLT systems, (2) development of reference-free damage diagnosis algorithms, (3) numerical and experimental validations using lab scale specimens and (4) application to real structures. For the LUI technique, a new complete noncontact LUI system is developed by synchronizing and integrating a Q-switched Nd:Yag pulse laser for ultrasonic wave generation, a laser Doppler vibrometer for ultrasonic wave measurement and a galvanometer with a focal lens for laser scanning. In particular, the characteristics of laser-generated ultrasonic waves are investigated through thermoelastic finite element modeling, and various laser scanning schemes for ultrasonic wavefield imaging are introduced and compared to address practical issues. Then, reference-free damage diagnosis algorithms, which are able to...