Page 208 - CW E-Magazine (26-9-2023)

Special Report

with a range from 10 KHz to 200 expression:AF = AE counts/Dura-
KHz (under software control). tion, (kHz)

4 Low Pass filters for each channel Rise time – The time from the first Zone
with a range from 100 KHz to 2.1 threshold crossing to the maximum
MHz (under software control). amplitude. Sensor
32 bit Digital Signal Processor. Count rate – Number of countsper
1 Mbyte DSP and Waveform buffer. time unit.
Location of
AE source
There are two types of AE signal The above parameters are shown Sensor with the
namely: ‘Burst’ and ‘Continuous’. pictorially in Figure-12. highest output
Burst AE is a qualitative description of Rise Fig. 13: Fault Zone Location
the discrete signal’s related to individu- time Peak sequence of signals can be detected
al emission events occurring within the amplitude giving a more accurate result using
material as shown in the left side of- Counts time differences and attenuation
Figure-11. Continuous AE is a qualita- Threshold characteristics of the wave.
tive description of the sustained signal
produced by time-overlapping signals Voltage (v) Pressure vessel leakages will gradu-
as shown in the right side of Figure-11. ally cause deterioration, which can re-
sult in catastrophic damage. AE signals
Different AE parameters used to o Duration generated by leakage have the potential
analyse AE signal for leakage/crack Time (s) of being used for online monitoring.
detection of a pressure vessel are as fol- Unfortunately, AE signals have the
lows(10, 11): Fig. 12: AE Parameters characteristics of being non-stationary,
Peak amplitude – The maximum Leakage/crack zone location can wide-band and with strong noise in-
of AE signal, measured in the units be found by adopting the following terference, which causes the monitor-
of dB (dB=20log (Vmax/1µvolt)- methodology (as per the details given ing results to have low reliability. To
10
preamplifier gain. in Figure-13): address the poor robustness of tradi-

Energy – Integral of the rectified Zone location determination is tional time-domain and time-frequency
voltage signal over the duration of based on the principle that the sen- domain-based monitoring methods, an
the AE hit. sor with the highest amplitude or online monitoring method based on
Duration – The time from the first energy output will be closest to the adaptive singular value decomposi-
threshold crossing to the end of the source. tion (ASVD) is proposed. Firstly, sin-
last threshold crossing. Zonal location aims to trace the gular value decomposition (SVD) is
Counts – The number of AE signal waves to a specific zone or region used to divide the signal space into a
exceeds threshold. around a sensor. signal subspace and a noise subspace.
Average Frequency (AF) – Deter- Zones can be lengths, areas or vol- Experiments indicate that SVD can
mines the average frequency in umes, depending on the dimensions distinguish leakages under conditions
kHz over the entire AE hit. AF can of the array. of different pressures and variable
be determined by the following With additional sensors added, a temperature, which means that SVD is

sensitive to changes in signal. Subse-
quently, update iteration-based ASVD
algorithms are proposed for long-term
online health monitoring and ASVD is
shown to be successful in distinguish-
ing the different statuses of intact, leak-
age and repaired condition. To improve
the robustness of ASVD, a novel energy
indicator is proposed, which can iden-
tify the status change more effectively.
Burst AE Signal Continuous AE Signal With the proposed methodology, an on-
Fig. 11: Burst and continuous AE Signals line monitoring application for pressure

208 Chemical Weekly September 26, 2023

Contents Index to Advertisers Index to Products Advertised