Condition assessment of pressurized pipelines using transient waves and time-reversal technique

Abstract

The time-reversal (TR) of water hammer (WH) waves in a viscous fluid bounded by a visco-elastic pipe is studied analytically and numerically. The quasi-steady part of friction is represented by the Darcy–Weisbach relation while the unsteady part is represented by the Vardy–Brown relation. The visco-elastic damping is represented by the Generalized Kelvin–Voigt model. It is shown that damping does not hinder the refocusing property of the time-reversed WH waves. In addition, it is analytically shown that (i) refocusing of WH waves reaches the well-known diffraction limit in physics, and (ii) TR is optimal in the sense of maximum signal-to-noise ratio (SNR) for white noise. Moreover, the TR technique is used to develop noise-tolerant (i) model-based method (method 1), and (ii) model-free method (method 2) for condition assessment of pressurized pipes. Method 1 is a one-dimensional technique that decouples the search for the location and strength of a defect (e.g., leak or blockage). The applicability of method 1 to diagnose multiple defects in both elastic and polymeric (visco-elastic) pipe systems is demonstrated numerically and experimentally. The results confirm that the proposed TR technique successfully detects defects with a resolution equal to half-wavelength (i.e., achieves the theoretical diffraction limit of waves). Method 1 is also successfully applied to localize real leaks in (i) a fire sprinkler system, located in HKUST, and (ii) a transmission main, located in Yuen Long, Hong Kong. Method 2 uses TR to identify differences (if any) between a baseline transient signal and a current (post-baseline) transient signal to localize the defects responsible for the difference between the two signals. That is, method 2 detects the defects that have emerged between the time the baseline measurements was conducted and the current time. The applicability of method 2 is demonstrated using numerical examples and experimental test cases in two laboratory-scale systems, a field-scale transmission-like pipeline system, and using real-field data collected from a district metered area (DMA), located in Ngau Tau Kok, Hong Kong. This technique successfully detected a change in the DMA that was unbeknown to the author and the water authority. This test case demonstrates the real-field applicability of method 2. Both methods are non-iterative; hence, computationally efficient. Additionally, their computational time increases linearly with the number of defects for both methods; thus, they are scalable to large systems and are easily amenable to automation — a fact that is demonstrated in the Water Research Resources Laboratory of HKUST. The use of metamaterials to manipulate and control WH wave propagation and ultimately achieve higher defect detection resolution is also explored. Metamaterial structures are modeled by imposing periodic variation of the impedance along the length of the pipe, where the modulation period is much smaller than the typical wavelength of the WH waves. It is shown that the transient wave propagation in a pipe made of metamaterial (named meta-pipe) is dispersive (i.e., nonlinear wavenumber-frequency relation) and produces sub-wavelength resonance as well as spectral band-gaps. For source localization using the TR technique, it is shown that the meta-pipe produces a far superior resolution than the normal pipe.

Date of Award	2022
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology
Supervisor	Moez LOUATI (Supervisor) & Mohamed Salah Ben Habib Ghidaoui (Supervisor)