Acoustic Emission Testing – Data Analysis

{0 Comments}

The data obtained from an acoustic emission test of a pressure vessel are analyzed to determine the existence of flaws and their locations in the shell of the vessel. Accurate analysis of the data is highly dependent upon the expertise of the vendor performing the test. Only those vendors that can demonstrate that they have acquired appropriate knowledge and experience should be used. It is vitally important for the vendor to be able to distinguish between acoustic emissions attributable to flaws, and other “noises” detected by the acoustic sensors during the test.

The acoustic emission signals detected by each sensor during a test are generally characterized for analysis by number of counts above a threshold amplitude that determines the sensitivity of the test, amplitude, duration, energy (area under signal envelope), and rise time as illustrated in Figure 700-25. The data are then displayed on CRT monitors as illustrated in Figure 700-26.

Characteristics of Acoustic Emission Signal

Plots of cumulative count, and count rate vs. time are shown in Figure 700-26a and b respectively. A plot of cumulative count rate vs. test pressure is shown in Figure 700-26c. An abrupt increase in counts vs. either time or test pressure is an indication of the existence of a significant flaw. Innocuous flaws are generally characterized
by a more gradual increase in counts.

A plot of the number of counts vs. amplitude is shown in Figure 700-26d, which can be useful for distinguishing between various plastic deformation and flaw propagation mechanisms, or for separating emissions associated with flaws from background noise. Figure 700-26e is a plot of the cumulative number of counts of equal or greater amplitude, which is more useful for evaluating the severity of a flaw.

Figure 700-26f is an acoustic emission source location display, which is basically a map of the vessel with the computer location of each emission source. This map can be used to focus subsequent NDE, such as MT and UT, that are usually desirable to confirm the existence of flaws in the vessel and to determine their sizes for fitnessfor-service evaluations (see Section 750).

The location of flaws is determined by analyzing the arrival time of the same acoustic emission from a flaw at different acoustic sensors attached to the shell of the vessel. The arrival time of an acoustic emission at a sensor is dependent on the distance of the flaw from the sensor, and the speed of sound in the material. Therefore, the differences in arrival time of the same emission at different sensors can be used to locate the flaw by triangulation. A computer is used to perform this function using the raw time-of-arrival data recorded during the test, and to drive an X-Y recorder to plot the location of each flaw on a “map” of the vessel. The computer is programmed to calculate the distances between the flaw and the sensors following the curvature of the vessel shell, and not as straight lines through the air between the flaw and the sensors.

The above displays of acoustic emission data can be obtained during the test to monitor the results, and can subsequently be recreated from the recorded data for more detailed analyses.

Typical Displays of Acoustic Emission Data

Leave a Comment

Your email address will not be published.