=Paper=
{{Paper
|id=Vol-2135/SEIM_2018_paper_39
|storemode=property
|title=Development of software package for data analysis of acoustic emission control
|pdfUrl=https://ceur-ws.org/Vol-2135/SEIM_2018_paper_39.pdf
|volume=Vol-2135
|authors=Victoria Belousova,Anastasia Grigorieva
}}
==Development of software package for data analysis of acoustic emission control==
https://ceur-ws.org/Vol-2135/SEIM_2018_paper_39.pdf
Development of a software package for acoustic
emission control data analysis
Victoria Belousova Anastasia Grigorieva
Saint Petersburg State University Saint Petersburg State University
Email: vsbelousova64@yandex.ru Email: a.v.grigorieva@spbu.ru
Abstract—The following article addresses a software package entire structure of the tested object. The sensitivity of
developed for working with data obtained during monitoring this method is sufficient to register even microscopic crack
the detection of material defects via the acoustic emission (AE) growth (by 0.001 mm), which allows to detect cracks in
method. Timely detection of cracks allows to prevent contin- time. The AE method can be employed for testing of a
gencies and accidents at early stages. This paper describes the wide variety of technological processes, as well as processes
architecture of this software, as well as the used calculation
methods, provides visualized results of their work, and com-
of changes in properties and condition of materials. This
pares them with other analysis methods. The innovation of broad spectrum of tasks and the variety of control objects
this work is the use of the moving window method for AE data requires constant improvement of data processing tools.
analysis. Obtained results indicate the practical importance and
relevance of our research in this area. III. Overview of existing solutions
There are several acoustic emission systems made by
I. Introduction
different manufacturers. A review of the characteristics
Evaluation of the current condition of varying indus- of their software, technical parameters, capacity of their
trial and infrastructural objects is one of actual problems AE equipment, and certain abilities for the analysis of
of modern material science. These objects include, but are registered data is presented in study [2]. Unfortunately, the
not limited to the oil and gas and chemical industry equip- post-experimental analysis capabilities of reviewed systems
ment, thermal and nuclear power equipment, aerospace are limited — mostly to creation of standard plots. For
equipment, pipeline and railway transportation, bridge example, in AMSY-5, it is easy to create an number
constructions, and concrete and reinforced concrete struc- of impulses-amplitude histogram or an amplitude-time
tures. The risks of equipment failure increase substantially correlation plot, but it is not always possible to implement
after it has been in use for a long time under mechanical a custom user formula [3]. Moreover, the system itself is
and thermal loads. The development of methods that quite expensive.
allow to study the physical nature of material degradation
Furthermore, a considerable number of studies dedi-
processes is an important task within the field of technical
cated to experimental and practical results of using AE
diagnostics. Acoustic emission testing is prominent among exists, for example: [4], [5], [6]. The moving window
these methods. It allows to identify the coordinates and method is widely applied in different scientific fields, such
estimate the danger level of defect-associated acoustic as economics [7], geophysics [1], social networks analysis
emission sources that appear in a loaded object. [8], audio encryption, and so on.
In this work we present a software package intended
for analyzing a large volume of specific data that is Our software applies the moving window method to
being studied by a considerable number of researchers all three statistical quantities and one composite parameter
around the world. Our software is designed for analyzing specifically during analyzing data of acoustic emission
different impulse responses of AE signals (amplitude, testing of loaded structures. This allows to filter peak
energy, length, etc) with the use of the moving window values of these quantities and observe the trends in changes
method. By employing this method, the software identifies of process phases in general.
the change dynamics of the arithmetic mean, median,
standard deviation, and b-value of these responses. IV. Software architecture
II. Acoustic emission method Currently, our software allows to build the trend of
a time series. We employ moving averaging for trend
Acoustic emission testing is an efficient method of determination. The calculations are performed on impulse
nondestructive testing. It is based on detecting elastic responses of acoustic signals — mainly amplitude values.
waves during deformation of stressed material. These This parameter is one of the most informative, because it
waves travel from the source to sensors that transform indicates the detectability of a signal, which is why it is
them into electrical signals. From the standpoint of the frequently used during AE testing. For example, in [9], the
AE method, a defect can produce its own signal [1]. The average amplitude of events is used for forming the P(R)
AE testing devices measure these signals, and then display criterion used for determining the necessity of additional
data used for evaluating the condition and behavior of the testing of detected areas of AE activity.
61
This software is written in C#. It accepts an input of If the size is too large, then the trend will be identified
a file containing data from a certain time interval. Every slower, however, there will be fewer false vibrations as well.
line of this file contains data collected from a single sensor,
in particular, registration time (up to a microsecond),
Fig. 1-6 display the plots of the moving average for the
sensor number, and the value of the parameter chosen
first experiment (destructive testing of steel-reinforced
for processing, e.g., amplitude. The user can set up the
concrete beams), where the optimal window size is
parameters required for their research. The software allows
100. This size has been identified via the possibility of
to select the sensor whose data will be analyzed to
real-time adjustment of program parameters.
localize the process, indicate whether the analysis will
be performed with respect to the number of signals or
time, designate the calculation method for one of the four Hereinafter in this section, all plots are built with the
statistical indicators (arithmetic mean, median, standard window size set to 100 signals, for time windows this size
deviation or b-value). The selection of the window size is set to 100 milliseconds, used data slice — the entire
adjusts the accuracy grade of data evaluation and resolves duration of the experiment.
the problems associated with the possible non-uniformity
of data distribution in the time domain.
V. Experiments
In this study we use the data collected during two
experiments designed to model the use of real-life con-
structions with different load types. In the first experiment
(strength load) we have used steel-reinforced concrete
beam samples which were being bent according to a 3-
point scheme. The total volume of AE testing data we
have obtained is quite significant. Thus, in some figures Figure 1. Simple moving average. X-axis indicates the number of
we have only plotted the data of sensor 2 (Fig. 1, 3, 5, 7, signals, Y-axis indicates amplitude.
10), and in others we have used the data of all four sensors
(Fig. 9, 8).
In the second experiment (strength-thermal load), we
have monitored a large-sized object contained in a cylin-
drical concrete construction. During the experiment, the
control object was uniformly heated to 400◦ . The data
were being registered by ten sensors. Fig. 2, 4, 6, 11.
VI. Processing methods
The developed analysis system provides the ability to
average the processed parameter by calculating three indi-
Figure 2. Simple moving average. X-axis indicates time (in millisec-
cators: arithmetic mean, median, and standard deviation. onds), Y-axis indicates amplitude.
All of them are calculated with the use of the moving
window method. This method can be explained as follows:
the calculations are performed on same length sets of B. Statistical median
consecutively registered data, which are shifted by one Statistical median is the middle element of an ordered
value relatively to each other during consecutive scanning sample. We use the following algorithm to determine the
of the entire measurement interval. The data set size median values: enumerate all values from 0 to N in an
(moving window size) is determined by the user. ascending order, then the median values are the elements
indexed 0.5N and (0.5N+1) for an even N, and 0.5(N+1)
A. Simple moving average for an odd N. Fig. 3 and 4 display the plots of the medians.
Simple moving average (or arithmetic mean) is calcu-
lated as follows: C. Standard deviation
Pn
xi (t) Standard deviation is calculated as follows:
v
σt = i=1 (1) uP n
n u (xi (t) − x(t))2
t
where t is the time interval; n is the smoothing interval; σ = i=1 (2)
xi (t) is the time series. n
where x(t) is the time series, x(t) is the arithmetic mean,
Fig. 1 and 2 present the plots of the simple moving
and n is the size of the moving window.
average. The smaller the size of the window, the faster
the moving window method identifies the new trend, but A larger standard deviation value indicates a larger
with that, the final plot contains more false vibrations. scatter in the presented sample. A smaller value points
62
VII. Plot analysis
In the beam destruction experiment, the plots of the
simple moving average and the statistical median are
virtually the same. This is correct for both the time scale
of the entire experiment (Fig. 2 and 3) and smaller time
scales (Fig. 7). Our analysis has revealed that standard
deviation is the least significant parameter out of all three.
Although it is of theoretical interest, statistical median and
Figure 3. Statistical median. X-axis indicates the number of signals, simple average have turned out to be more informative in
Y-axis indicates amplitude. practice. These parameters help to trace the dynamics of
crack formation. According to the study [6], during static
loading of metal with a crack there is no increasing trend
in the time domain of AE signal amplitude, but there
are individual AE signals whose amplitude exceeds the
average by 45 dB. We observe a similar situation in the
beam loading experiment. This change of trend can be
observed on the resulting plots: in Fig. 7, which represents
the phase of active formation of main cracks, the averaged
amplitude values do not increase uniformly. Furthermore,
the increases of amplitude of certain signals are filtered
via the moving window method which allows to see the
whole picture of trend change.
Figure 4. Statistical median. X-axis indicates time (in milliseconds),
Y-axis indicates amplitude.
to set values being aggregated around the mean (Fig. 5 и
Fig. 6). Standard deviation can be calculated in a different
way if variance, which is equal to the radicand in formula
2, has been found previously.
Figure 7. Statistical median and simple average for a shorter time
slice. Windows size is set to 20.
The selection of an optimal size of the moving window is
important. Our software allows the user to make this choice
empirically for the whole duration of the experiment, and
Figure 5. Standard deviation. X-axis indicates the number of signals, then change this value proportionally to the total number
Y-axis indicates amplitude. of signals in a particular smaller data slice. For example,
in the beam experiment, one sensor has registered 1700
events. The clearest picture was obtained with the size
of the window L = 100 events. Thus, for the macrocrack
formation period that contains 305 events, L was set to
305*100/1700 ≈ 18 (Fig. 7). Sometimes, it is reasonable to
reduce the window size, for example, to avoid missing the
registration of relatively rare events such as macrocrack
formation.
Consider Fig. 8. In this figure, arrows denote the
moments of crack formation that were registered directly
(visually) during the experiment. It is interesting that this
plot also contains similar decline peaks at different time
points. It is highly likely that those points correspond to
Figure 6. Standard deviation. X-axis indicates time (in milliseconds), internal cracks in the structure that could not be identified
Y-axis indicates the calculated indicator.
visually.
63
change of the predominant destruction type that were
presented in the following studies: [13], [15], [16], [17].
We have implemented the ability to calculate the b-
value in our software package. In some cases, it turns
out to be a more informative evaluation parameter for
crack formation dynamics than other statistical indicators.
For example, in experiments with thermal or composite
(strength-thermal) loading of a large-size reinforced con-
Figure 8. Simple moving average for different stages of beam crete structure, the change of b-value trends in regards
destruction. X-axis indicates the number of signals, Y-axis indicates
amplitude.
to the defect formation stage are more pronounced in
comparison to the experiments with strength loading of
small samples. Fig. 10 and Fig. 11 show the b-value plotted
VIII. b-value analysis by the system during analyzing the results of the strength
and thermal load experiments respectively.
It is possible to use the Gutenberg-Richter law (widely
applied in seismology [1]) to study the scaling of the
amplitude distribution of AE signals that appear during
crack formation. In AE terms, this formula can be written
as follows:
log10 N = a–b ∗ Amax dB (3)
where Amax dB is the maximum amplitude in the window
(in decibels), a is an empirical constant value set to
4.8, b is a value obtained from this equation and then
multiplied by 20 to be comparable to the value used in
seismology[14]. The b-value is used for identifying the Figure 10. Dependency of b-value on the number of signals. Strength
predominate destruction type and determining trends in load experiment.
construction damage development.
Figure 11. Dependency of b-value on the number of signals. Thermal
load experiment.
The resulting plots demonstrate the b-value fluctuating
in a significantly wider range for the second structure type.
However, its use is sufficiently informative for the objects
Figure 9. Plots of amplitude distributions for different stages of of the first type as well. It can be seen that destruction
beam destruction and their b-values processes of different intensity are being considered, which
is confirmed by the presented dependencies.
Fig. 9 provides an example of express analysis for b- Analyzing these kinds of dependencies allows AE test-
value estimation for the experiment conducted by non- ing specialists and material scientists to obtain useful and,
destructive testing specialists during test destruction of sometimes, unexpected information on the behavior of
a reinforced concrete construction. In this analysis, the materials under the influence of different kinds of loads
angle of inclination of a line (which is build via the [18].
least squares method) determines the b-value. Fig. 9, (a)
shows the amplitude distribution for the stage that directly
precedes the destruction of reinforcement metal; at this IX. Results and conclusions
stage, macrocracks have already formed, the main material The following results have been obtained:
has unloaded, and microcracks were forming intensively on
the last loaded area in the vicinity of the reinforcement 1) We have developed a software package that
metal. Fig. 9, (b) displays the data for the stage of main enhances an AE control analysis system with
macrocrack formation. The obtained results correspond several data analysis methods designed to
with the notions of b-value evolution associated with the increase the informativeness of testing. This
64
software was tested on real experimental data. Using these estimates, we have performed the calculations
on data of impulse responses of acoustic signals. The final
2) The correctness of the employed algorithms is results indicate the efficiency of the suggested moving av-
confirmed by the obtained results matching the eraging methods in the task of analyzing acoustic emission
previously known facts on the development of testing data and the practicability of using the considered
defect formation. software.
3) Employing this software package allowed the AE In our further research we plan to perform the analysis
testing specialists to perform a more accurate of data obtained in experiments with composite loads and
and detailed analysis of data, which substantially then generalize the results.
increased the informativeness of testing.
The development of methods for detecting signals
associated with defect growth with the use of information
In this paper, we have presented elements of analysis theory methods could be a fruitful area for further work
of data collected during both a laboratory experiment as well.
of loading and destruction of reinforced concrete beams
and real-world testing of a large-size reinforced concrete
construction. The conditions of conducting the first type of References
experiments have made it possible for us to observe certain
[1] Lyubushin A.A. Analiz dannykh sistem geofizicheskogo i eko-
key phases of the sample destruction process and identify logicheskogo monitoringa. Moscow, Izd: Nauka (2007). (In Rus-
them with the corresponding AE testing data. Software sian)
algorithms have performed well in these experiments: the [2] Oglezneva L. A. Sravnitel’nye kharakteristiki akustiko-
type of obtained dependencies corresponds to the real- emissionnykh sistem. Vestnik nauki Sibiri - Siberian Journal of
world processes that occurred in control objects according Science (2011) (In Russian)
to the known facts about the mechanics of destruction of [3] Metodika akustiko-emissionnogo kontrolya s ispol’zovaniem AE
this type of materials. Thus, the proposed algorithms were sistemy AMSY-5 firmy Vallen-Systeme GmbH. Russia, Vol-
tested successfully. gograd (2010). (In Russian)
[4] Zotov K., Rastegaev V., Gomera V., Sokolov V., Fedorov A.,
During our research, we have corrected the moving Smirnov A. The Detection of Different Stages of the Delami-
window size with respect to estimating the maximum nating in the Pressure Vessels by the Ultrasonic and Acoustic
informativeness of this parameter during the destruction Emission Technique. 19th World Conference on Non-Destructive
Testing (WCNDT), 13-17 June 2016, Munich, Germany, Book
stage. The resulting estimates are employed for the moving of Abstracts, p.209.
window method in the developed software for both the [5] Nefedyev, E. J., Gomera, V. P., Smirnov, A. D. (2016). Use of
described experiment and other situations in which crack the Capabilities of Acoustic-Emission Technique for Diagnostics
formation (deformation or other internal destruction) pro- of Separate Heat Exchanger Elements. In Advances in Mechan-
cesses are obscured and occur inside of the object. In this ical Engineering (pp. 183-194). Springer, Cham.
kind of experiments, the value of such analysis increases [6] Rastegaev I. A., Chugunov A. V., Vinogradov A. Y., Merson
due to the inability to visually observe material structure D. L., Danyuk A. V. The specific features of acoustic-emission
degradation processes and having to resort to evaluating testing of vessel equipment with a wall delamination of a
technological origin. Russian Journal of Nondestructive Testing,
them by indirect indicators. 51(5), 2015, pp. 280-291.
The second kind of structure that was considered in this [7] Anantchenko I. V., Musaev A. A. Programma dlya torgovli
study is an object of this type. The results of additional na rynke Foreks na osnove skol’zyashchikh srednikh. Vektory
razvitiya sovremennoy nauki (2014), pp. 14-18. (In Russian)
examination of material structure via destructive methods
[8] Bagretsov G.I., Shindarev N.A., Abramov M.V., Tulupyeva
confirmed the correctness of conclusions that were made T.V. Approaches to development of models for text analysis of
during the AE testing with the use of our software. information in social network profiles in order to evaluate user’s
vulnerabilities profile // Soft Computing and Measurements
The software of existing AE systems is generally limited (SCM), 2017 XX IEEE International Conference on. – IEEE,
to a set of standard plots used for a formal representation 2017. – P. 93–95.
of testing results. It usually lacks advanced tools of data [9] Nefed’yev E. Yu., Smirnov A. D., Gomera V. P. Razrabotka
analysis. This is typical even for the most modern AE metodicheskikh priemov dlya povysheniya effektivnosti AE kon-
testing systems, such as AMSY-5 (developed in Germany) trolya teploobmennikov. Modern Mechanical Engineering: Sci-
[3], which was employed during the discussed experiments. ence and Education, 4 (2014), pp. 408-418. (In Russian)
[10] Tikhonova O.A. Sravnitel’nyy analiz razlichnykh algoritmov
skol’zyashchego usredneniya s ispol’zovaniem kriteriya mini-
X. Conclusion mal’nogo srednego kvadrata oshibok. DonNTU (2008). (In Rus-
sian)
In this work we have developed a software package
intended for analyzing acoustic emission testing data. [11] NDT Recource Center: https://www.nde-ed.org (accessed
17.04.2018)
We have shown that it can be successfully employed for
[12] Vrije Universiteit Brussel: Damage testing, prevention and
both laboratory experiments and large-size construction detection in aeronautics, chapter 10, 2007.
testing. We should note that this kind of analysis requires
[13] Golaski L., Gebski P., Ono K. – Diagnostics of Reinforced
determining the optimal size of the moving window. All Concrete Structures by Acoustic Emission - 25th European Con-
of the required estimates of this value for the arithmetic ference on Acoustic Emission Testing, Prague, Czech Republic,
mean, median and standard deviation were obtained. 2002, pp. I/207-I/215.
65
[14] Colombo S., Main I.G., Forde M.C. Acoustic emission for
bridges: experiments on reinforced concrete beams. 25th Euro-
pean Conference on Acoustic Emission Testing, Prague, Czech
Republic, 2002 pp. I/127-I/134
[15] Calabrese, L., G. Campanella, and E. Proverbio. "Use of
cluster analysis of acoustic emission signals in evaluating damage
severity in concrete structures." Journal of Acoustic Emission 28
(2010).
[16] Carpinteri A., Lacidogna G., Pugno N. Time scale effects on
acoustic emission due to elastic waves propagation in monitored
cracking structures. Physical Mesomechanics, 5, 2005. pp. 85-89.
(In Russian)
[17] Carpinteri A., Lacidogna G., Puzzi S. Prediction of cracking
evolution in full scale structures by the b-value analysis and Yule
statistics. Politechnico di Torino, 2008.
[18] Nefedyev E. Yu. Ispol’zovanie metoda akusticheskoy emissii s
primeneniem spektral’nogo analiza signalov dlya opredeleniya
parametrov techi v truboprovodakh ITER. Modern Mechanical
Engineering: Science and Education, 3 (2013), pp. 347-355. (In
Russian)
66