RU2569078C1 - Method of identifying sources of acoustic emission signals arising from degradation of material, cracking and structural failure - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: method includes measuring maximum pulse amplitude, the number of emissions and the duration of signal pulses, after which acoustic emission signal sources are identified based on said measurements.

EFFECT: high reliability of identifying acoustic emission signal sources.

7 dwg

Description

The invention relates to methods for non-destructive testing of materials and products according to strength conditions and is intended to identify sources of acoustic emission signals arising from the degradation of the material, the formation of cracks and the destruction of the structure.

A known method is when the registration of cracks in brittle stress coatings is carried out using acoustic emission control [1]. In this case, the acoustic emission system registers not only the signals arising from the formation of cracks in the brittle coating layer, but also interference signals, as well as signals caused by degradation of the substrate material during its deformation and destruction.

The closest technical solution adopted for the prototype is Patent No. 2403564 of the Russian Federation: IPC G01N 29/14, published on 10.11.10 [2].

In the studied areas of the structure, fragile strain gauges are installed, for remote monitoring of the state of which: registration and location of the resulting cracks is used acoustic emission system. To separate “useful” signals (cracking in a brittle layer of strain gauges) from all other signals, they are filtered.

The disadvantage of this technical solution is that during the filtering process, useful information is lost, which is provided by the method of acoustic emission (AE) on the degree of degradation of the substrate material during its deformation.

The task to which this technical solution is directed is to develop a method that allows, in the process of acoustic emission diagnostics of products, to identify the sources of recorded signals, to recognize signals that occur during material degradation, cracking and structural destruction during its deformation.

When implementing the claimed technical solution, the task is carried out by introducing a diagram of parameters characterizing the pulse attenuation rate and the average emission frequency.

The specified technical result in the implementation of the invention is achieved by the fact that when registering the signals with an acoustic emission system, an algorithm is additionally used for recognizing signals by the shape of a damped wave.

As the essential features, the most informative for recognizing signals by the shape of a damped wave, the combined parameters are used:

u _m / N _AND , N _AND / t _AND ,

where u _m is the maximum amplitude of the pulse,

N _And - the number of emissions

t _And - pulse duration.

To classify detected AE signals in the field of these parameters specially developed software that allows the diagram descriptors u _m / N _{D and} N _AND / t _and trace the dynamics of the formation of clusters of signals in material degradation, the formation of cracks and destruction of the structure during its loading.

The boundaries of the clusters formed by AE signals that occur during material degradation, cracking and structural failure for the parameter u _m / N _And depend on the detection threshold u _th and are calculated by the formula:

, $${\left(\frac{u_m}{N_u}\right)}_{B,H}=\mathrm{twenty}\cdot K_{B,H}\cdot \left(\lg \left(u_{th}\right)-\lg {\left(u_{th}\right)}_{\min }\right)\cdot {\left(\frac{u_m}{N_u}\right)}_{B,H_{{\left(u_{th}\right)}_{\min }}}$$

,

where (u _th ) _min is the minimum possible value of the registration threshold, corresponding to 26 dB, at which the electrical noise of the AE equipment used is not recorded,

K - the coefficient for determining the boundaries of the cluster is set depending on the nature of the signal source and the threshold level u _th .

The technical and economic efficiency of the invention follows from the technical result obtained by the implementation of the invention, i.e. monitoring the state of the material of the structure during its deformation, and therefore, diagnosing the degree of its degradation and preventing the destruction of materials and products according to the strength conditions.

As examples of demonstration of the proposed method, the application of the parameter diagram u _m / N _И and N _И / t _{И is considered} in the study of the acoustic emission properties of a multilayer oxide tensile coating. The examples are accompanied by graphical illustrations, which are presented in FIG. 1-7:

FIG. 1 is a diagram of crack formation in an oxide strain gauge;

FIG. 2 is a graph of strain changes along the length of the beam during bending;

FIG. 3 - comparative location picture of the distribution of signals in a brittle layer;

FIG. 4 - waveform of signals during the formation of cracks in strain gauges;

FIG. 5 - a cluster of AC signals during cracking in a brittle layer of a strain gauge at a threshold level u _th equal to 26 dB;

FIG. 6 - cluster of acoustic emission signals during the formation of cracks in the brittle layer of the strain gauge at a threshold level of u _{th = 60 dB} ;

FIG. 7 - a cluster of acoustic emission signals during the formation of cracks in the brittle layer of the strain gauge at a threshold level of u _{th = 32 dB} .

The brittle oxide strain gauge of FIG. 1 is a thin aluminum foil subjected to electrochemical anodization to obtain a transparent oxide film (15-35 μm thick) and glued to the structural element under study [3].

When deformations ε ₁ occurring in the aluminum foil substrate exceed the threshold strain ε ₀ in the oxide film of strain gauges, crack patterns are formed that reflect the force field of the highest principal stresses (deformations) on the surface of the structure. Using the test characteristics of the strain sensitivity (σ ₀ , ε ₀ ) and the graph of the change in the number of cracks in a brittle oxide film (Ψ) versus the level of deformations in the substrate, it is possible to estimate the values of principal stresses (deformations) on the test surface with an error of not more than 15% structures in the area of crack propagation [3].

The brittle oxide strain gauge is a multilayer coating that includes a brittle outer layer with a thickness of 10-35 μm, a non-porous barrier layer with a thickness of about 1 μm, a layer of aluminum foil with a thickness of 60-90 μm and a brittle layer of epoxy adhesive with a thickness of 60-100 μm. All layers of the strain gauge are potential sources of AE signals. As the level of deformation in the substrate increases, the destruction of the strain gauge occurs in the following sequence. First, the outer brittle layer of alumina is destroyed, then the barrier layer, then plastic deformations occur in the aluminum foil and, finally, the adhesive layer connecting the strain gauge to the surface of the product is destroyed. Each layer generates its own signals, which differ in both energy and time parameters.

To study the acoustic emission properties of oxide strain gauges, A7 aluminum foil 100 μm thick was used, which was glued to the surface of test samples using epoxy glue. The electrochemical anodization of aluminum foil was carried out in a 15% aqueous solution of sulfuric acid using pre-worked oxidation modes. The duration of the electrochemical anodizing process varied from 20 to 60 minutes, the current density from 4 to 6 A / dm², the temperature of the electrolyte is from 0 to 20 ° C. This method of obtaining strain gauges made it possible to regulate their sensitivity from extremely low ε₀equal to 400 ÷ 500 μm / m, up to ε₀, equal to 2000 ÷ 2500 μm / m.

Experimental studies of the AE properties of oxide strain gauges were carried out on test samples in the form of calibration beams with dimensions of 200 × 20 × 5 mm. When testing the strain gauges, calibration beams made of ST-1 plexiglass and ED-20 epoxy were used. These materials have a relatively low modulus of elasticity E equal to 2900 MPa, tensile strength σ _B equal to or more than 100 MPa, and elongation at break of at least 3% [4]. In addition, they are not sources of AE signals, which is especially important to reduce the amount of interference arising from test tests of strain gauges.

Figure 2 shows a diagram of the location of the strain gauges and AE converters on the surface of the calibration beams when tested for cantilever bending. Between the level of the load applied at the end of the beam and the magnitude of its deflection, the following relationship takes place:

$$P=\frac{E\cdot b\cdot {\mathrm{<figure-callout\; id="3"\; label="h"\; filenames="00000007.png,00000008.png"\; state="\{\{state\}\}">h</figure-callout>}}^3}{6\cdot {\mathrm{<figure-callout\; id="3"\; label="L"\; filenames="00000007.png,00000008.png"\; state="\{\{state\}\}">L</figure-callout>}}^3}\cdot f,gde(\mathrm{one})$$

E is the modulus of elasticity of the material of the beam,

L is the length of the console from termination to the place of application of the load,

b and h are the width and thickness of the calibration beam.

With a deflection f equal to 1 mm, deformations in the substrate of the strain gauge changed depending on the distance to embedment from 60 to 200 μm / m.

To record AE signals when testing samples of the materials under study and testing strain gauges, we used the multi-channel multi-parameter acoustic emission system A-Line 32D of OOO Interyunis and the two-channel DiSP-2 system of RAS. As acoustic emission transducers (PAE), GT200 transducers of GlobalTest LLC and resonance transducers DP-151 of the RAS company were used.

The intrinsic noise of the A-Line 32D AE electrical circuit, including the PAE, preamplifier, and data collection and preprocessing unit, was 26 dB, and for the DiSP-2 AE with DP-151 integrated converters, it was 24 dB. Depending on the acoustic noise arising from the tests, the threshold for detecting AE signals was set in the range from 26 to 60 dB.

A comparison of the distribution of AE signals, which are localized during loading of the calibration beam, with the pattern of crack propagation in the brittle coating layer is shown in FIG. 3. In the upper part of the figure, a location pattern of the distribution of AE signals is presented, which reflects the process of propagation of cracks in a brittle layer of an oxide film upon loading of a strain gauge. The figure shows that the number of localized signals and their distribution between PAEs 1 and 2 are in good agreement with the number of cracks and their location along the length of the strain gauge. Each located AE signal reflects the crack propagation process in a brittle coating layer.

The occurrence and propagation of cracks in the brittle layer of the strain gauge generates AE signals, the parameters of which depend on many factors, including the thickness of the brittle layer. The greater the thickness, the greater the amplitude of the pulse generated by the formation of a crack in the tensile coating. The signals recorded during the formation of cracks in the brittle layer of the tensile coating are characterized by a significant amplitude and a high attenuation rate. Figure 4 shows the waveforms of typical pulses of signals generated by the formation of cracks in the strain coatings with a brittle layer thickness of 21 and 33 μm.

Using the developed program, which allows us to trace the dynamics of the formation of signal clusters in the process of deformation and destruction of the AE diagnostic object in the descriptor field u _m / N _И , N _И / t _И , we tested the multilayer structure of the strain gauge during its deformation and destruction. The parameter diagram u _m / N _И , N _И / t _И defines the clusters formed by AE signals arising from the destruction of a brittle layer, aluminum foil (substrate) and adhesive layer.

An analysis of the diagrams of the recorded AE signals caused by the formation of cracks in the brittle layer of the tensile coating showed that the boundaries of the parameter u _m / N _{AND of the} clusters of the recorded signals depend on the level of the detection threshold u _th . With increasing threshold u _th decreases the number of recorded emissions, and therefore, increases the value of the parameter u _m / N _And .

In FIG. 5 and FIG. Figure 6 shows diagrams of the considered parameters for AE signals caused by the formation of cracks in the brittle layer for strain gauges with a threshold strain ε₀equal to 500 μm / m, at the level of the registration threshold u_th, equal to 26 and 60 dB. With increasing threshold level, the interval of u_m/ N_AND the recorded signals increases significantly from 20 to 80 μV / unit. at u_thequal to 26 dB, and also increases to 200-1000 μV / unit. at u_thequal to 60 dB. The interval of the averaged emission frequencies N_AND/ t_AND detected pulses during the formation of cracks in the brittle coating layer is stored in the same boundaries from 90 to 170 kHz. For signals whose amplitude exceeds the registration threshold by no more than two times (less than 6 dB), the parameter N_AND/ t_AND can reach 200 kHz.

As shown by experimental studies, the boundaries of the clusters formed by the signals of crack formation in the strain coating and the signals arising from the destruction of the substrate material for the parameter u _m / N _И depend on the set level of the detection threshold u _th and can be calculated by the formula:

, где $${\left(\frac{u_m}{N_u}\right)}_{B,H}=\mathrm{twenty}\cdot K\cdot \left(\lg \left(u_{th}\right)-\lg {\left(u_{th}\right)}_{\min }\right)\cdot {\left(\frac{u_m}{N_u}\right)}_{B,H_{{\left(u_{th}\right)}_{\min }}}$$

where

(u _th ) _min - the minimum possible value of the set detection threshold, corresponding to 26 dB, at which noises of the AE equipment used were not recorded,

K is the coefficient for determining the boundaries of the cluster, the value of which, depending on the nature of the signal source and the threshold level u _th, varies from 0.1 to 1.1.

As an example, we define the cluster boundaries for AE signals that arise when cracks form in the brittle layer of a strain gauge at a threshold level u _th equal to 32 dB. In this case, the coefficient for determining the boundaries of the AE signal cluster will be 0.33. This coefficient value is applied when u _th exceeds (u _th ) _min by 2 dB or more.

In the above formula, the boundaries of the clusters of the parameter u _m / N _И are calculated as the ratio of the maximum signal amplitude (μV) to the number of emissions (units) before the wave goes below the threshold u _th , the value of which is also measured in μV. To go to decibels in the formula, the following dependence is used: 20 · (log (u _th ) -lg (u _th ) _min ).

Based on the fact that at (u _th ) _min equal to 26 dB, the boundaries of the parameter u _m / N _{AND of the} cluster formed by AE signals arising from the formation of cracks in the brittle layer of the strain gauge with an oxide film thickness of 33 μm are 20 and 80 μV / units (see Fig. 5), we determine the approximate boundaries of the cluster at u _th equal to 32 dB:

(u _m / N _I ) _H = 0.33. (32-26) .20 = 40 μV / unit;

(u _m / N _AND ) _B = 0.33. (32-26) .80 = 160 μV / unit.

The calculated values of the boundaries of the parameter u _m / N _{AND of the} cluster are in rather good agreement with the experimental data obtained at the threshold level u _th equal to 32 dB (see Fig. 7) for AE signals arising from the formation of cracks in the brittle layer of the tensile coating.

AE signals generated by different layers of the strain gauge in the diagram u _m / N _И , N _И / t _И have their own specific clustering areas. Two methods were used to record AE signals generated during plastic deformation and rupture of aluminum foil: the first, when the foil sample was directly subjected to a tensile test, and the second, when the foil was glued onto a plexiglass calibration beam, which was loaded by cantilever bending.

In order to record AE signals observed during the destruction of the adhesive layer and peeling of the foil, the following experiments were carried out. A strip of aluminum foil 200 mm long and 20 mm wide was glued to calibration plexiglass beams with dimensions of 240 × 22 × 6 mm. The sticker was carried out from one end of the foil over a length of 140 mm. The free end of the foil 60 mm long was bent at a right angle and clamped from the edge at a length of 40 mm between the steel plates. Then, the beam was rigidly fixed in fixed supports installed at a distance of 200 mm. DP-151 transducers were placed on the upper plane of the beam at a distance of 160 mm. Tear foil tests were carried out in several stages. At each new stage, the load increased by 10 N and was maintained for 30 seconds if foil was not torn off. After this, the sample was unloaded and aged for 3 minutes before the next test.

In Fig. 7, for the level u _th equal to 32 dB, the diagram of descriptors u _m / N _И , N _И / t _И shows clusters of AE signals generated by various layers of the strain gauge during its destruction with an increase in the level of tensile deformations in the substrate.

As can be seen from the figure, the boundaries of the cluster formed by the signals arising from the destruction of the brittle layer of the strain gauge (1) are partially overlapped by the cluster of signals arising from the destruction of the adhesive layer and peeling of the foil from the substrate (3). The boundaries of clusters (2) and (4) formed by signals arising from plastic deformation and rupture of aluminum foil are noticeably lower.

To separate the signals of clusters (1) and (3), parameters of time, load, or strain can be used. Active destruction of the brittle layer of the strain gauge occurs in the region of elastic deformations, long before the occurrence of irreversible structural changes in the substrate material. The destruction of the adhesive layer is observed only with significant plastic deformations, usually exceeding 10 ⁴ μm / m, when a high density of cracks already occurs in the brittle layer of the strain gauge and the process of formation of new cracks slows down noticeably, requiring an increasing load increment. Therefore, at such strain values, the increase in AE signal activity is mainly due to the processes of peeling of the coating and destruction of the substrate.

The above examples showed the high information content of the proposed method, not only for recording signals of cracking in the strain gauge, but also for assessing the degree of degradation of the substrate material during its deformation.

Literature

1. Makhutov N.A., Shemyakin V.V., Ushakov B.N., Petersen T.E., Vasiliev I.E. The use of acoustic emission to control the process of cracking in brittle oxide strain gauges // Factory Laboratory. - 2011. No. 6. 41-44 p.

2. Patent No. 2403564 of the Russian Federation, IPC G01N 29/14. A device for diagnosing the limit state and early warning about the danger of destruction of materials and products / Vasiliev I.E., Ivanov V.I., Makhutov N.A., Ushakov B.N .; Applicant and patent holder A.A. Blagonravova RAS, No. 2009100183/28, application. 01/11/09, publ. 10.11.10, Bull. No. 31.

3. Stress-strain states of the rocket engine. / Ed. Makhutova N.A., Rachuk B.C. M .: Nauka, 2013, 646 p.

4. Methods and means of determining the fields of deformations: Reference / Under. ed. N.I. Prigorovsky. M.: Mechanical Engineering, 1983. 248 p.

Claims (1)

A method for recognizing sources of acoustic signals arising from the formation of cracks and destruction of the structural material, including acoustic emission diagnostics of the controlled object, characterized in that the maximum pulse amplitude, the number of emissions are measured by the shape of the damped wave of the acoustic emission signals during cracking and structural destruction and the pulse duration of the signals, after which establish the ratio of the parameters u _m / N _AND , N _AND / t _AND
where u _m is the maximum amplitude of the pulse,
N _And - the number of emissions
T _And - the pulse duration of the signals,
the boundaries of the parameter u _m / N _{And the} clusters of the detected signals are set depending on the level of the registration threshold u _th and determined by the formula:

,
where K _{B, N} is the coefficient for determining the upper and lower boundaries of the cluster is 0.9-1.1.

Images