Acoustic emission testing (AE) is an established method for the detection of leaks in gas and liquid containments. It is currently being considered for the continuous monitoring of piping systems in a range of government facilities that require safeguards against accidental release. The ability of AE to locate leaks quickly using remote sensors is of particular interest. Most AE leak testing work is only qualitative. There is increasing demand for it to be made more quantitative. It is often desired to be able to estimate the leakage rate from the measured AE signal. This has been done very successfully, but only in certain well-defined, pre-calibrated situations (e.g. valves in nuclear plant and oil refineries). A harder task is to design AE monitoring installations that will detect or locate leaks of a pre-specified magnitude. To develop this kind of design capability for a wide range of structures, a plan is in place based on quantitative understanding of the four elements of the AE monitoring process: the release of acoustic energy at the source, the propagation through the structure, the sensing, and the signal conditioning and processing. In this paper, these four elements are separately treated in the quantitative analysis of AE from gas leaks in a model piping system. Leak sources with sizes ranging from 0.1 to 1 .0mm were installed in a double-walled stainless steel piping section. With pressures ranging from 13 to lOOpsi, leakage rates ranged from 5 to 1400 std.ml/s. AE measurements were made at various frequencies in the range 30 to 500kHz. In the first approximation at the lower frequencies, the AE signal was shown to be proportional to the pressure and to the orifice cross-section, and thus to the leakage rate. Variations from this linear relationship were associated with the generation of edge tones. The nature of the source process was clarified through an innovative energy analysis. Based on this analysis, it is concluded that the AE is generated outside the pipe, downstream of the leak orifice, in accordance with the established theory of aerodynamic sound generation. When this airborne sound impinges on the pipe wall, a very small fraction of the energy passes into the metal. It is this very small amount of acoustical energy that is detected by the remote AE sensor.
展开▼
