Back to results list
Please use this identifier to cite or link to this item:
|Title:||Structural health monitoring-oriented damage characterization using nonlinear ultrasonic waves and active sensor networks : from fundamental investigations to engineering applications||Authors:||Hong, Ming||Advisors:||Su, Zhongqing (ME)
Cheng, Li (ME)
Zhou, Li-min (ME)
|Keywords:||Structural health monitoring.
Structural failures -- Prevention.
|Issue Date:||2016||Publisher:||The Hong Kong Polytechnic University||Abstract:||In order to maintain structural integrity to the best extent and avoid structural failure, it is highly desirable to identify damage in engineering structures as accurately as possible when the damage is still in its embryo stage. Traditional nondestructive evaluation (NDE) and structural health monitoring (SHM) techniques based on guided waves have been well developed over the years, and are now on the verge of maturity for real-world applications. Guided waves, and Lamb waves in particular, feature desired merits for practical implementation in a cost-effective manner, including fast propagation, strong penetration, and omni-directional dissimilation in the structure under inspection. Nonetheless, the effectiveness and accuracy of the majority of available techniques, which use linear signal features such as the time of flight (TOF) of a particular waveform, has been demonstrated only for gross damage ("gross" here means the characteristic dimension of the damage is comparable to the wavelength of the guided waves used for detection). Motivated by this, nonlinear attributes of guided elastic waves, as typified by second harmonics, have been increasingly studied and employed in NDE and SHM for engineering structures, owing to their higher sensitivity to small-scale damage (e.g., fatigue cracks) compared to their linear counterparts. Nevertheless, in the context of SHM, which is aimed at evaluating structural health in a condition-based, automatic, and real-time fashion, a couple of challenges remain in utilizing the nonlinear attributed of Lamb waves. First and foremost, few nonlinear methods have been targeted at quantitative damage characterization, such as determining the damage's location and/or size, in conjunction with the use of sensor networks. Without the use of an SHM-friendly sensor network (e.g., attachable to structures with multiple sensing paths), the information collected on structural integrity is highly limited. Second, the generation and acquisition of nonlinear Lamb wave features using miniaturized lead zirconium titanate (PZT) wafers have not been thoroughly scrutinized. Last but not least, the feasibility of using nonlinear Lamb wave techniques for real-world SHM practices has not been fully validated. The uncertainty of popular nonlinear features, such as the relative acoustic nonlinearity parameter (RANP), is yet to be quantified and interpreted for practice. To address the above issues, a quantitative damage detection technique using nonlinear Lamb wave features and active sensor networks is proposed in this study, oriented at real-world SHM of plate-like engineering structures. First, rigorous fundamental investigation has gone into the modeling of nonlinearities involved in metallic materials. A dedicated simulation approach is proposed, in which various sources of nonlinearities (including the material itself and damage) are accounted for. The nonlinearities in acquired Lamb waves, manifested as second harmonics, are studied and quantified using the defined nonlinear factor - RANP. This parameter is then employed to calibrate the distance of a sensing path from the damage site, so that localization of the damage may be achieved using an active sensor network.
To better comprehend the nonlinearities of Lamb wave signals, advanced signal processing techniques are applied in the time, spectral, and time-frequency domains, respectively. Damage indices (DIs) based on TOF, signal correlation, RANP, and the TOF of second harmonic Lamb waves are developed for diagnostic imaging, through a probabilistic data fusion scheme. Furthermore, the concept of DI synthesis is proposed by combining the previously defined DIs, so that the resulting index can potentially adapt to different levels of measurement noise, while maintaining its sensitivity to damage of various sizes. To apply the proposed methodology, an integrated SHM system is developed, along with the conception of "decentralized standard sensing" for configuring active sensor networks. Proof-of-concept studies are performed, which well demonstrate the accuracy and efficiency of the approaches for small-scale damage localization in a quantitative manner, in comparison to results obtained from traditional, linear techniques. Finally, the developed methodology and the configured system are implemented for quantitative damage characterization in two engineering structures, from a simple plate structure tested in the laboratory to a complex engineering system in the real world. In the first one, the practical concern of uncertainties involved in RANP is addressed, by developing a probabilistic model and using it in repetitive experiments for detection of barely visible impact damage (BVID) in carbon fiber laminates. In the second study, the integrated SHM housing the developed damage detection algorithms are deployed for in-situ bogie structure monitoring on a running high-speed train, operated on the Beijing-Shanghai High-Speed Railway. The system has delivered consistent performance, accommodating the rugged working conditions throughout the test. Using a synthesized DI that combines linear and nonlinear signal features, a mock-on damage attached to the bogie panel is accurately located and presented in diagnostic images in real time. Conclusively, through theoretical modeling, simulation, and experimental investigations, a stream of novel damage characterization techniques are developed in this study based extensively on nonlinear Lamb wave attributes. The methodology, drawing on advanced damage identification mechanisms, signal processing techniques, and diagnostic imaging algorithms, has demonstrated a great potential for real-life SHM applications in conjunction with the use of active sensor networks.
|Description:||PolyU Library Call No.: [THS] LG51 .H577P ME 2016 Hong
xxxi, 232 pages :color illustrations
|URI:||http://hdl.handle.net/10397/40931||Rights:||All rights reserved.|
|Appears in Collections:||Thesis|
Show full item record
Files in This Item:
|b28563918_link.htm||For PolyU Users||208 B||HTML||View/Open|
|b28563918_ira.pdf||For All Users (Non-printable)||7.47 MB||Adobe PDF||View/Open|
Citations as of Mar 19, 2018
Citations as of Mar 19, 2018
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.