136
Z. Wang et al.
and spectrum envelope with acoustic emission signal. Yoshiok and Fujiwara [6]’s
study points out that AE event rate can be used to predict the faults earlier than RMS
of the vibration signal. Elforjani and Mba [7–9] carried out an accelerated fatigue
test with SKF-51210 ball bearing. RMS and absolute energy information of event
are considered to be the good performance degradation indexes. Then, Warren, etc.
[10] analyzed the rolling contact fatigue process of hyperfine polished surface. It is
pointed out that amplitude and RMS are more sensitive. The study of rolling contact
[11] shows that there is a close relationship between the AE event count and the
initial damage size of rolling surface. Filip, etc. [12] did trend analysis with event
count, duration, rise time and amplitude. It is concluded that the ball bearing has two
stages before the appearance of slight pitting. Abdullah, etc. [13] investigated RMS,
amplitude and kurtosis. The relationship between acoustic emission duration and
damage size was studied. There still some other scholars keep on doing the research
with AE indexes to detect damages inside materials [14–16]. From the papers above,
it can be fully demonstrated that acoustic emission technology is an effective means
to detect the internal damage of materials from the studies above. The damage degree
of ball bearing can be expressed by acoustic emission index.
Conventional AE indexes are specified from each AE event. The threshold voltage
of acoustic emission events is usually determined manually. That is easy to introduce
subjective errors. If AE data is large, and the fatigue life of ball bearing is long. It
will discount computing efficiency greatly. It is not good for a real-time monitoring
test. Moreover, the conventional acoustic emission index is difficult to correspond
with statistical index in time. It is hard to make a comprehensive judgment of fatigue
damage of ball bearings. To solve the problems above, a method of ball bearing
fatigue monitoring with improved acoustic emission new index is proposed. Then,
AE data from a self-made accelerated fatigue test rig for ball bearings are studied.
Compared with the monitoring with conventional acoustic emission indexes and
statistical indexes. A more comprehensive analysis and damage sensitive indexes of
ball bearing are obtained. The effectiveness of the proposed algorithm is verified.
13.2 Fatigue Monitoring with Acoustic Emission
Technology
13.2.1 Acoustic Emission Detection and Acoustic Emission
Source of Ball Bearing
The phenomenon of transient elastic waves produced by the rapid energy release
from local sources in materials is called acoustic emission [17]. Materials produce
acoustic emission signals due to plastic deformation or crack propagation. The simplified waveform parameter method is usually used in acoustic emission analysis. The
waveform parameters of conventional acoustic emission signal include: event count,
ring count, rise time, duration, etc. The method extracts the structure parameters from
Z. Wang et al.
and spectrum envelope with acoustic emission signal. Yoshiok and Fujiwara [6]’s
study points out that AE event rate can be used to predict the faults earlier than RMS
of the vibration signal. Elforjani and Mba [7–9] carried out an accelerated fatigue
test with SKF-51210 ball bearing. RMS and absolute energy information of event
are considered to be the good performance degradation indexes. Then, Warren, etc.
[10] analyzed the rolling contact fatigue process of hyperfine polished surface. It is
pointed out that amplitude and RMS are more sensitive. The study of rolling contact
[11] shows that there is a close relationship between the AE event count and the
initial damage size of rolling surface. Filip, etc. [12] did trend analysis with event
count, duration, rise time and amplitude. It is concluded that the ball bearing has two
stages before the appearance of slight pitting. Abdullah, etc. [13] investigated RMS,
amplitude and kurtosis. The relationship between acoustic emission duration and
damage size was studied. There still some other scholars keep on doing the research
with AE indexes to detect damages inside materials [14–16]. From the papers above,
it can be fully demonstrated that acoustic emission technology is an effective means
to detect the internal damage of materials from the studies above. The damage degree
of ball bearing can be expressed by acoustic emission index.
Conventional AE indexes are specified from each AE event. The threshold voltage
of acoustic emission events is usually determined manually. That is easy to introduce
subjective errors. If AE data is large, and the fatigue life of ball bearing is long. It
will discount computing efficiency greatly. It is not good for a real-time monitoring
test. Moreover, the conventional acoustic emission index is difficult to correspond
with statistical index in time. It is hard to make a comprehensive judgment of fatigue
damage of ball bearings. To solve the problems above, a method of ball bearing
fatigue monitoring with improved acoustic emission new index is proposed. Then,
AE data from a self-made accelerated fatigue test rig for ball bearings are studied.
Compared with the monitoring with conventional acoustic emission indexes and
statistical indexes. A more comprehensive analysis and damage sensitive indexes of
ball bearing are obtained. The effectiveness of the proposed algorithm is verified.
13.2 Fatigue Monitoring with Acoustic Emission
Technology
13.2.1 Acoustic Emission Detection and Acoustic Emission
Source of Ball Bearing
The phenomenon of transient elastic waves produced by the rapid energy release
from local sources in materials is called acoustic emission [17]. Materials produce
acoustic emission signals due to plastic deformation or crack propagation. The simplified waveform parameter method is usually used in acoustic emission analysis. The
waveform parameters of conventional acoustic emission signal include: event count,
ring count, rise time, duration, etc. The method extracts the structure parameters from