Application of Pattern Recognition Algorithms and Nondestructive Evaluation Techniques for the Structural Health Monitoring of Civil Structures
MetadataShow full item record
Safety and durability of civil structures play a very important role in ensuring the economic and industrial prosperity of society. Currently, assessing the performance and safety of civil structures relies on visual inspection, and unfortunately, even with the recent advances in automated ground-based nondestructive evaluation (NDE) methods, there is a potential that indications of structural degradation could be missed. This dissertation aims at designing integrated Structural Health Monitoring (SHM) systems for the nondestructive evaluation of civil structures. Overall, it is proposed to use NDE techniques, such as acoustic emission (AE), guided ultrasonic waves (GUW), and automated vision-based inspection, coupled with advanced statistical signal processing and pattern recognition algorithms to locate and quantify the extent of damage in structural elements, such as reinforced concrete shear walls, pipelines, and post-tensioned concrete beams. Experimental tests are carried out to validate the proposed approaches. In particular, an automated damage diagnosis method is developed for reinforced concrete structures using acoustic emission. In addition, an advanced vision-based NDE technique for crack pattern quantification in reinforced concrete structures is proposed and validated experimentally. A reference-free corrosion damage identification using guided ultrasonic waves is introduced for post-tensioned concrete structures. And finally, a real time monitoring system is designed for rehabilitated pipelines using acoustic emission.