Transient wave characterization and utilization for viscoelastic pipe analysis and diagnosis

Abstract

Viscoelastic pipelines are increasingly used in water supply systems of different scales. Although plastic pipes have many advantages in engineering applications, many studies have shown that their behavior during transient is different from elastic pipes. Moreover, most of the developed leak detection methods based on elastic pipelines cannot be directly applied to viscoelastic systems. Although accurate characterization can solve these problems, the current viscoelastic identification methods are either time-consuming or ill-posed. Therefore, the developed transient-based viscoelastic parameter identification methods and leak detection methods should be re-examined. The main contributions and finds of this work can be summarized as follows: (1). Through analytical analysis in the frequency domain, this study proposes a novel and efficient frequency domain multistage method for viscoelastic parameter identification in reservoir pipe valve systems. Moreover, the different roles of retardation times on the frequency shift are revealed and understood by this research, and the main idea of the method is to use the frequency shifts caused by different retardation time scales to identify as many viscoelastic parameters as possible for a viscoelastic pipeline. Both experimental and numerical tests have shown the effectiveness and efficiency of the proposed method; (2). Coupled with (1), a frequency response function-based two-step strategy for viscoelastic characterization and then leak detection is developed, and the method only uses resonant frequencies of the leaky transient trace to fulfil the goal of leak detection in simple viscoelastic pipeline systems (e.g., reservoir pipe valve systems). (3). For more complex systems, a new efficient globalized frequency-domain inverse transient analysis method is innovatively proposed for simultaneous viscoelastic parameter identification and leak detection using a part of the frequency response function. Its effectiveness has been validated in both single and branched systems at different flow and leak conditions. Moreover, the different influential factors are revealed in the subsequential numerical investigation. (4). A novel forward and backward transient analysis method for efficient leak detection in the time domain is proposed based on the reconstructive method of characteristics. The method uses the intact pipe assumption, splits the leak localization and quantification, and therefore, does not need to perform any optimization procedure for leak detection. This significantly contributes to the transient-based leak detection methods. (5). The transient energy analysis is then conducted to analyze different energy forms and leak-wave interactions in both elastic and viscoelastic pipeline systems. The results deepen the understanding of the roles and mechanisms of the frictional and the viscoelastic effects in a transient system. Moreover, the influence of leak location and size on energy dissipation is also revealed through the analysis. The main conclusions, insufficiencies, and recommendations for the whole study are summarized at the end of this thesis.