Acoustic emission (AE) is employed in this study to monitor piping failure process in the dike subjected to varied head action. Piping is the main reason for the seepage failure of dikes and dams, and the AE characterization of piping process is investigated. The laboratory experiments and data analyses were carried out using specially designed physical model and data acquisition systems. The monitored data were analyzed using four methods, namely, AE ringing count, fractal dimension, frequency domain analyses and b value. The analysis results showed that the piping failure process can be divided into 3 phases: seepage erosion, piping generation, and piping development. In different stage, the activity of the AE signals is related to seepage flow and seepage pressure. Based on the fractal dimension and waveform energy spectrum distribution analysis in the entire process of piping failure, the measured acoustic signals in the test can reflect both a seepage-induced and a soil-seepage induced AE. The AE ringing count fractal dimension and waveform energy spectrum can be used as new index monitoring the occurrence and development of piping. The abrupt change of AE fractal dimension and AE energy spectrums curve with time can be considered as the precursor to piping occurrence and piping development. What’s more the b value of the AE signal can be used to describe the piping state.

在这项研究中，采用了声发射（AE）来监测受不同水头作用的堤防中的管道故障过程。管道是堤坝大坝渗流破坏的主要原因，并研究了管道过程的声发射特性。使用专门设计的物理模型和数据采集系统进行了实验室实验和数据分析。使用AE振铃计数，分形维数，频域分析和b值这四种方法对监测数据进行分析。分析结果表明，管道破坏过程可分为三个阶段：渗漏侵蚀，管道生成和管道发展。在不同阶段，声发射信号的活动与渗流和渗压有关。基于管道破裂整个过程中的分形维数和波形能谱分布分析，在测试中测得的声信号既可以反映渗透引起的AE，也可以反映土壤渗透引起的AE。AE振铃计数的分形维数和波形能谱可用作监测管道发生和发展的新指标。AE分形维数和AE能谱曲线随时间的突然变化可以认为是发生管道和发展管道的先兆。此外，AE信号的b值可用于描述管道状态。AE振铃计数的分形维数和波形能谱可用作监测管道发生和发展的新指标。AE分形维数和AE能谱曲线随时间的突然变化可以认为是发生管道和发展管道的先兆。此外，AE信号的b值可用于描述管道状态。AE振铃计数的分形维数和波形能谱可用作监测管道发生和发展的新指标。AE分形维数和AE能谱曲线随时间的突然变化可以认为是发生管道和发展管道的先兆。此外，AE信号的b值可用于描述管道状态。