SYSTEM FOR THE ACOUSTIC DETECTION OF PRECURSORS OF MATERIAL FRACTURE IN A SAMPLE UNDER TEST

Abstract

System (10) for determining a trigger amplitude indicating a precursor to material failure in a test specimen (12), the system (10) comprising: a microphone (20) that converts the sound emission emitted by the test specimen (12) into electrical signals (40), wherein a load (38) is applied to the test specimen (12) and the sound emission is emitted when the load (38) causes deformation of the test specimen (12) prior to material failure; a control module (22) in electrical connection with the microphone (20), the control module (22) being configured to execute instructions for: monitoring the electrical signals (40) generated by the microphone (20); filtering the electrical signals (40) generated by the microphone (20) to allow frequencies within a range of interest and to attenuate frequencies outside the range of interest;Converting the electrical signals (40) generated by the microphone (20) into individual frequency components (66) based on a fast Fourier transform (FFT) (62), wherein each of the individual frequency components (66) contains a peak intensity that represents audible sound; and determining the trigger amplitude based on the peak intensity of the individual frequency components (66) of the FFT (62); wherein the individual frequency components (66) of the FFT (62) define an amplitude track (82); wherein the control module (22) is executed to determine the trigger amplitude by: analyzing the amplitude track (82) using a kernel density estimator (KDE) function, wherein the KDE function produces a smoothed estimate of the amplitude track (82), and wherein the smoothed estimate of the amplitude track (82) contains a plurality of peaks (84), each representing a number of times that a peak amplitude (68) occurs in the FFT (62);and identifying a highest peak (84A) of the smoothed estimate of the amplitude trace (82), wherein the highest peak (84A) represents the peak amplitude (68) that occurs most frequently in the FFT (62).

Description

The present disclosure relates to a system and a method for determining a trigger amplitude that indicates precursors of material failure in a sample under test based on audible sound. The present disclosure also relates to a system for capturing images of the sample under test by a camera, wherein the system instructs the camera to capture images at a different frame rate when it is determined that the trigger amplitude has occurred. Image processing systems are becoming increasingly popular. However, image processing requires storing and processing relatively large amounts of data, which in turn can demand significant processing and storage resources. One factor influencing the storage space required for large image files is the frame rate of the camera capturing the images. The frame rate is expressed in frames per second (fps) and indicates the number of images the camera captures per second. A higher frame rate results in a larger image file that requires more storage space, while a lower frame rate results in a smaller image file that requires less storage space. Although image processing systems fulfill their purpose, there is therefore a need for an approach to reduce the memory requirements of an image processing system. DE 694 16 410 T2 describes a system for the continuous monitoring of stressed reinforcements in a structure, such as bridges or modern buildings. The system has a large number of detectors that are randomly and purposefully positioned around the structure. The detectors respond to acoustic or seismic energy emissions and generate a signal. The signals associated with a failure of the reinforcement are analyzed to locate the point of failure. US 2010 / 0058867A1 describes an AE detector comprising an AE sensor and a destruction assessment section. The destruction assessment section evaluates a bearing as a warning of destruction when a predetermined number or more points are present, defined by parameters calculated based on signals from the AE sensor within a predetermined range of a parameter space defined by a variety of parameters that can be generated based on signals from the AE sensor. US 2005 / 0204818A1 describes a method for determining the vibration amplitude limits of a mechanical device. The method includes identifying a mechanical device and a frequency range for a spectrum to be analyzed, retrieving vibration spectra that include an individual spectrum for the mechanical device and the frequency range, calculating the frequency for the individual spectrum, and identifying the individual spectrum with the fewest frequency lines. Furthermore, the method includes calculating noise bandwidths and a maximum noise bandwidth, removing outlier data, calculating the conditional kernel density, and calculating vibration amplitude limits to detect defects in the mechanical device. The object of the invention can be considered to be to provide an image processing system that reduces the storage requirements of large image data.According to the invention, a system for determining a trigger amplitude indicating a precursor to material failure in a test specimen is described. The system includes a microphone that converts the acoustic emissions emitted by the test specimen into electrical signals. A load is applied to the specimen, and the acoustic emissions are emitted when the load causes deformation of the specimen prior to material failure. The system also includes a control module that is electrically connected to the microphone and executes instructions for monitoring the electrical signals generated by the microphone. The control module executes commands to filter the electrical signals generated by the microphone to allow frequencies within a range of interest and to attenuate frequencies outside the range of interest.The control module converts the electrical signals generated by the microphone into individual frequency components based on a fast Fourier transform (FFT), with each frequency component containing a peak intensity that represents audible sound. The control module determines the trigger amplitude based on the peak intensity of each frequency component of the FFT. According to the invention, the individual frequency components of the FFT define an amplitude track. According to the invention, the control module determines the trigger amplitude by analyzing the amplitude track using a kernel density estimator (KDE), wherein the KDE function generates a smoothed estimate of the amplitude track and the smoothed estimate of the amplitude track contains a plurality of peak values, each representing a number of times that a peak amplitude occurs in the FFT. According to the invention, the control module determines the trigger amplitude by identifying the highest peak of the smoothed estimate of the amplitude trace, wherein the highest peak represents the peak amplitude that occurs most frequently in the FFT. In one embodiment, the highest peak represents the audible sound generated by background noise. In one embodiment, the control module determines the trigger amplitude by identifying a second-highest peak of the smoothed estimate of the amplitude trace, wherein the second-highest peak represents the peak amplitude that occurs most frequently after the highest peak in the FFT. In one embodiment, the second highest peak represents the acoustic emission emitted by the sample under test when it is deformed before material fracture. In one embodiment, the control module determines the trigger amplitude by identifying a valley between the highest peak and the second-highest peak and selecting an amplitude value corresponding to the valley as the trigger amplitude. In one embodiment, the frequencies outside the range of interest represent background noise. According to the invention, a method for determining a trigger amplitude indicating a precursor to material failure in a test specimen is disclosed. The method comprises converting acoustic emissions emitted by the test specimen into electrical signals using a microphone. A load is applied to the specimen, and the acoustic emissions are emitted when the load causes deformation of the specimen prior to material failure. The method includes monitoring the electrical signals generated by the microphone using a control module. The method also includes filtering the electrical signals generated by the microphone to allow frequencies within a range of interest and to attenuate frequencies outside the range of interest.The method also includes converting the electrical signals generated by the microphone into individual frequency components based on a fast Fourier transform (FFT), where each frequency component contains a peak intensity representing audible sound. Finally, the method includes determining the trigger amplitude based on the peak intensity of each frequency component of the FFT. According to the invention, the individual frequency components of the FFT define an amplitude track, and the method further comprises analyzing the amplitude track using a KDE function. The KDE function generates a smoothed estimate of the amplitude track, and the smoothed estimate of the amplitude track contains a plurality of peaks, each representing a number of times a peak amplitude occurs in the FFT. According to the invention, the method further comprises the identification of a highest peak of the smoothed estimate of the amplitude track, wherein the highest peak represents the peak amplitude that occurs most frequently in the FFT. In one embodiment, the method further comprises the identification of a second-highest peak of the smoothed estimate of the amplitude track, wherein the second-highest peak represents the peak amplitude that occurs most frequently after the highest peak in the FFT. In another embodiment, the method also includes determining a valley between the highest peak and the second highest peak. In another embodiment, the method further includes the selection of an amplitude value corresponding to the valley as the trigger amplitude. Another embodiment is that the frequencies outside the range of interest represent background noise. In one application, an image acquisition system includes a system for determining a trigger amplitude according to the invention. The image acquisition system comprises a specimen to be tested, wherein a load is applied to the specimen to cause it to deform before material failure. The system also includes a camera that captures images of the specimen, the camera capturing images at a first frame rate and a second frame rate, the first frame rate being lower than the second frame rate. The system also includes a microphone that converts the sound emissions emitted by the specimen to electrical signals, wherein a load is applied to the specimen and the sound emissions are emitted when the load causes the specimen to deform before the material failure.Finally, the system includes a control module that is electrically connected to the microphone and camera. The control module executes instructions to monitor the camera, which captures images at the first frame rate. The control module also executes instructions to monitor the electrical signals generated by the microphone for a trigger amplitude, where the electrical signals represent the sound amplitude. The control module executes commands to determine that the electrical signals generated by the microphone indicate that the trigger amplitude has occurred. In response to the determination that the trigger amplitude has occurred, the control module instructs the camera to capture images of the sample under test at the second frame rate, where the trigger amplitude represents a precursor to material failure in the sample. In this application, the sample to be tested is subjected to one of the following tests: an open-hole tensile test, a tensile strength test, a notched tensile test, a compression test, and a torsion test. Another aspect of this use case is that the test specimen consists of at least one of the following materials: a glass fiber composite, a carbon fiber composite, a basalt fiber composite, a plastic, a filled plastic, and a fiber-reinforced polymer. Another aspect of this use case is that the first frame rate is approximately five frames per second (fps) and the second frame rate is approximately fifty fps. Fig. 1 is a schematic diagram showing a system for determining a precursor to material failure in a sample under test, wherein the system includes a control module in electronic communication with a microphone; Fig. 2 is a block diagram representing the control module shown in Fig. 1; Fig. 3 shows a diagram of an exemplary fast Fourier transform (FFT) of the electrical signal received from the microphone shown in Fig. 1; Fig. 4 shows a diagram illustrating a smoothed estimate of the FFT shown in Fig. 3, determined using a kernel density estimator (KDE); Fig. 5 is a process flow diagram illustrating a method for determining the trigger amplitude using the system shown in Fig. 1; and Fig.Figure 6 is a schematic diagram of a system for capturing images of the sample to be tested using a camera. Figure 1 shows an exemplary system 10 for determining a precursor to material failure in a specimen 12 under test. In the illustrated, non-limiting embodiment, the system 10 comprises the specimen 12 under test, a holder 14 for securing the specimen 12 under test, a microphone 20, and a control module 22. In the example shown, the specimen 12 under test is a tensile specimen with two enlarged ends 30 and a measuring section 32 arranged between the two enlarged ends 30, but the specimen 12 under test can also have other configurations. In one embodiment, the specimen 12 under test consists of at least one glass fiber composite, a carbon fiber composite, a basalt fiber composite, a plastic, a filled plastic, or another fiber-reinforced polymer, such as...carbon fiber reinforced polymers (CFRP), although other types of materials can also be used. Fig. 1 shows the device 14 with two clamps 34 that clamp and secure the enlarged ends 30 of the specimen 12 under test. It should be noted, however, that Fig. 1 is only exemplary and other types of devices can also be used. The holder 14 exerts a force 38 on the specimen 12 under test. In Fig. 1, the load 38 is shown, for example, as a tensile force that pulls the specimen 12 apart. In one embodiment, the specimen 12 under test is subjected to an open-hole tensile test, a tensile strength test, a notched tensile test, a compression test, or a torsion test; however, other types of tests in which the specimen 12 undergoes material failure can also be used. In some embodiments, the specimen 12 under test may undergo material failure that is not visible. The acoustic emission is emitted by the specimen 12 under test before the material fractures. Specifically, the acoustic emission occurs when the load 38 causes deformation of the specimen 12 under test, before the specimen 12 fractures under the load 38. The microphone 20 converts the acoustic emission emitted by the specimen 12 under test into electrical signals 40 (see Fig. 2). The electrical signals 40 generated by the microphone 20 indicate the sound amplitude. The control module 22 is electronically connected to the microphone 20 and monitors the electrical signals 40 generated by the microphone 20. As explained below, the control module 22 determines a trigger amplitude of the electrical signals 40, indicating that the specimen 12 under test is about to fracture, caused by the load 38 exerted by the device 14. In other words, the trigger amplitude indicates a precursor to material fracture in the specimen 12 under test. Figure 2 is a block diagram illustrating the control module 22, which comprises a bandpass filter 50, a Fast Fourier Transform (FFT) module 52, a Kernel Density Estimation (KDE) function module 54, and an analyzer module 56. The control module 22 can refer to or be part of an electronic circuit, a combinational logic circuit, a field-programmable gate array (FPGA), a processor (shared, dedicated, or as a group) that executes code, or a combination of some or all of the above, for example, in a system-on-a-chip. Furthermore, the control module 22 can be microprocessor-based, such as a computer with at least one processor, memory (RAM and / or ROM), and associated input and output buses. The processor can operate under the control of an operating system residing in memory.The operating system can manage computer resources so that computer program code, embodied as one or more computer software applications, such as an application residing in memory, can be executed by the processor. Alternatively, the processor can execute the application directly; in this case, the operating system can be omitted. The bandpass filter 50 of the control module 22 can be implemented using analog components such as resistors, inductors, and capacitors, or as a digital filter (see Fig. 1 and Fig. 2). The bandpass filter 50 of the control module 22 filters the electrical signals 40 generated by the microphone 20 (see Fig. 1). In particular, the bandpass filter 50 allows frequencies within a range of interest to pass through, while attenuating frequencies outside of this range. The frequencies within the range of interest include the acoustic emissions emitted by the sample 12 under test, as well as the sound emitted by the sample 12 during material failure. The frequencies outside the range of interest represent background noise that can be detected by the microphone 20. It should be noted that the specific values ​​of the frequencies within the range of interest depend on variables such as...depend on the material of the test specimen 12 and the specific type of test to which the test specimen 12 is subjected. Fig. 3 is a diagram 60 showing an exemplary FFT 62 of the electrical signals 40 received by the bandpass filter 50. The diagram 60 includes an x-axis representing the frequency in Hertz and a y-axis representing the sound amplitude. As can be seen from Fig. 2 and Fig. 3, the FFT module 52 of the control module 22 converts the electrical signals 40 from the bandpass filter 50 into individual frequency components 66. As can be seen in Fig. 3, each of the individual frequency components 66 contains a peak amplitude 68, which represents audible sound. The individual frequency components 66 of the FFT 62 define an amplitude track 70. As explained below, the trigger amplitude is determined based on the peak amplitude 68 of the individual frequency components 66 of the FFT 62. The KDE function module 54 of the control module 22 analyzes the amplitude track 70 of the FFT 62 using a KDE function. Fig. 4 is a graph 80 that represents a smoothed estimate of the amplitude track 82 of the FFT 62 (Fig. 3). The graph 80 contains an x-axis representing the sound amplitude and a y-axis indicating the number of samples. Referring to Figs. 2, 3, and 4, the KDE function module 54 executes a KDE function that generates the smoothed estimate of the amplitude track 82 of the FFT 62 (Fig. 3). The smoothed estimate of the amplitude track 82 contains a plurality of peaks 84. Each peak 84 of the smoothed estimate of the amplitude track 82 represents the number of occurrences of the peak amplitude 68 in the FFT 62 (Fig. 3). The analyzer module 56 of the control module 22 determines the trigger amplitude based on the smoothed estimate of the amplitude track 82. Specifically, the analyzer module 56 identifies a highest peak 84A of the smoothed estimate of the amplitude track 82. The highest peak 84A represents the peak amplitude 68 that occurs most frequently in the FFT 62 (Fig. 3). The highest peak 84A of the smoothed estimate of the amplitude track 82, shown in Fig. 4, represents audible sound generated by background noise. The analyzer module 56 identifies a second-highest peak 84B of the smoothed estimate of the amplitude track 82. The second-highest peak 84B represents the peak amplitude 68 that occurs most frequently in the FFT 62 after the highest peak amplitude 68A. The second highest peak 84B represents the acoustic emission emitted by the sample 12 ( Fig. 1) under test when it is deformed before material fracture. As shown in Fig. 4, the highest peak 84A and the second-highest peak 84B are divided into two distinct groups, and a valley 90 exists between the two peaks 84A and 84B. The analysis module 56 of the control module 22 determines the trigger amplitude by first identifying the valley 90 between the highest peak 84A and the second-highest peak 84B and then selecting an amplitude value corresponding to the valley 90 as the trigger amplitude. If there is no clear distance between the two peaks 84A and 84B, it can be assumed that the acoustic emission has not yet occurred. Fig. 5 is a process flow diagram illustrating an exemplary method 100 for determining the trigger amplitude using the system 10 shown in Fig. 1. Referring to Figs. 1-5, the method 100 begins in block 102. In block 102, the control module 22 (Fig. 2) monitors the electrical signals 40 generated by the microphone 20. The method 100 can then proceed to block 104. In block 104, the bandpass filter 50 of the control module 22 (Fig. 2) filters the electrical signals 40 generated by the microphone 20. The process 100 can then proceed to block 106. In block 106, the FFT module 52 of the control module 22 converts the electrical signals 40 from the bandpass filter 50 into individual frequency components 66 (as shown in Fig. 3). The process 100 can then proceed to block 108. In block 108, the KDE function module 54 of the control module 22 analyzes the amplitude track 70 of the FFT 62 (Fig. 3) using the KDE function. The procedure 100 then proceeds to block 110. In block 110, the analysis module 56 of the control module 22 identifies the highest peak 84A of the smoothed estimate of the amplitude track 82 (as seen in Fig. 4). The procedure 100 then proceeds to block 112. In block 112, the analysis module 56 of the control module 22 identifies the second-highest peak 84B of the smoothed estimate of the amplitude track 82. The procedure 100 can then proceed to block 114. In block 114, the analysis module 56 of the control module 22 determines the trigger amplitude by identifying the trough 90 between the highest peak 84A and the second highest peak 84B. The procedure 100 can then proceed to block 116. In block 116, the analysis module 56 of the control module 22 selects an amplitude value corresponding to valley 90 as the trigger amplitude. The procedure 100 can then be terminated. Fig. 6 shows a system 200 for capturing images of a sample 212 under inspection using a camera 204. In one embodiment, the camera 204 can be a stereo camera configured to capture three-dimensional images of the sample 212 under inspection; however, two-dimensional images can also be used. As explained below, the system 200 instructs the camera 204 to capture images at a faster frame rate when it detects that the trigger amplitude has occurred. Thus, the trigger amplitude detected by the system 10 (Fig. 1) can be used to initiate a hardware trigger event that triggers an external action (i.e., a change in the frame rate of the camera 204). Although Fig. 6 shows that the system 200 uses the trigger amplitude to trigger a change in the frame rate of the camera 204, the trigger amplitude can also be used to trigger other events. In the embodiment shown in Fig. 6, the system 200 comprises the camera 204, a specimen 212 to be tested, a microphone 220, and a control module 222. The control module 222 is electronically connected to both the camera 204 and the microphone 220. Similar to the embodiment shown in Fig. 1, a load 238 is applied to the specimen 212 by a device 214, the load 238 causing the specimen 212 to deform before material failure. Camera 204 captures images of the sample 212 under inspection. Specifically, camera 204 captures images at a first frame rate and a second frame rate, the first being lower than the second. It is understood that camera 204 captures images at the first frame rate before system 200 detects that the trigger amplitude has occurred. Since the images captured by camera 204 are not of great interest, the first frame rate can be significantly slower than the second. Once control module 222 detects that the trigger amplitude has occurred, it instructs camera 204 to capture images at the second frame rate. The second frame rate is higher than the first because the images captured immediately after the acoustic emission of the sample 212 under inspection are of greatest interest.Capturing images at a faster frame rate as soon as the acoustic emission occurs results in a reduction of the images that are captured, processed and stored by the control module 222. As shown in Fig. 6, the control module 222 monitors the camera 204, which captures images at the first frame rate. The control module 222 also monitors the electrical signals 40 generated by the microphone 220 (as shown in Fig. 2) for the trigger amplitude. During the test, the control module 222 detects that the electrical signals 40 generated by the microphone 220 indicate that the trigger amplitude has occurred. In response to the detection that the trigger amplitude has occurred, the control module 222 instructs the camera 204 to capture images of the sample 212 under test at the second frame rate. In a non-restrictive example, the first frame rate is approximately five frames per second (fps) and the second frame rate is approximately fifty fps. In this example, the total time the camera captures images is approximately fifty-four seconds, and camera 204 is instructed to switch to the second frame rate after approximately fifty seconds during the test, as soon as the trigger amplitude is detected. Therefore, camera 204 captures images at the second frame rate only during the last four seconds of the test. If the second frame rate were used for the entire duration of the test, the control module 222 would capture 2700 images. However, because camera 204 switches to the second frame rate 50 seconds after the test begins, only 450 images are captured.Therefore, in the present example, the instruction to camera 204 to switch to the second frame rate leads to an approximately 83 percent reduction in the number of images that are captured, processed, and stored by control module 222. Accordingly, the described system 200 offers an approach to reducing the amount of image data that is stored and processed by control module 222.

Claims

System (10) for determining a trigger amplitude indicating a precursor to material failure in a test specimen (12), the system (10) comprising: a microphone (20) that converts the sound emission emitted by the test specimen (12) into electrical signals (40), wherein a load (38) is applied to the test specimen (12) and the sound emission is emitted when the load (38) causes deformation of the test specimen (12) prior to material failure; a control module (22) in electrical connection with the microphone (20), the control module (22) being configured to execute instructions for: monitoring the electrical signals (40) generated by the microphone (20); filtering the electrical signals (40) generated by the microphone (20) to allow frequencies within a range of interest and to attenuate frequencies outside the range of interest;Converting the electrical signals (40) generated by the microphone (20) into individual frequency components (66) based on a fast Fourier transform (FFT) (62), wherein each of the individual frequency components (66) contains a peak intensity that represents audible sound; and determining the trigger amplitude based on the peak intensity of the individual frequency components (66) of the FFT (62); wherein the individual frequency components (66) of the FFT (62) define an amplitude track (82); wherein the control module (22) is executed to determine the trigger amplitude by: analyzing the amplitude track (82) using a kernel density estimator (KDE) function, wherein the KDE function produces a smoothed estimate of the amplitude track (82), and wherein the smoothed estimate of the amplitude track (82) contains a plurality of peaks (84), each representing a number of times that a peak amplitude (68) occurs in the FFT (62);and identifying a highest peak (84A) of the smoothed estimate of the amplitude trace (82), wherein the highest peak (84A) represents the peak amplitude (68) that occurs most frequently in the FFT (62). System (10) according to claim 1, wherein the highest peak (84A) represents the audible sound generated by background noise. System (10) according to claim 1, wherein the control module (22) is configured to determine the trigger amplitude by: identifying a second-highest peak (84B) of the smoothed estimate of the amplitude track (82), wherein the second-highest peak (84B) represents the peak amplitude (68) that occurs most frequently in the FFT (62) after the highest peak (84A). System (10) according to claim 3, wherein the second highest peak (84B) represents the acoustic emission emitted by the sample (12) under test when it is deformed prior to material fracture. System (10) according to claim 3, wherein the control module (22) is configured to determine the trigger amplitude by: determining a valley (90) between the highest peak (84A) and the second highest peak (84B); and selecting an amplitude value corresponding to the valley (90) as the trigger amplitude. System (10) according to claim 1, wherein the frequencies outside the range of interest represent background noise. Method (100) for determining a trigger amplitude indicating a precursor to material failure in a test specimen (12), the method (100) comprising: converting the acoustic emissions emitted by the test specimen (12) into electrical signals (40) by a microphone (20), wherein a load (38) is applied to the test specimen (12) and the acoustic emissions are emitted when the load (38) causes deformation of the test specimen (12) prior to material failure; monitoring electrical signals (40) generated by a microphone (20) by a control module (22); filtering the electrical signals (40) generated by the microphone (20) to allow frequencies within a range of interest and to attenuate frequencies outside the range of interest;Converting the electrical signals (40) generated by the microphone (20) into individual frequency components (66) based on a fast Fourier transform (FFT) (62), wherein each individual frequency component (66) contains a peak intensity representing audible sound; and determining the trigger amplitude based on the peak intensity of each frequency component (66) of the FFT (62); wherein each frequency component (66) of the FFT (62) defines an amplitude track (82); determining the trigger amplitude via the control module (22) by: analyzing the amplitude track (82) using a kernel density estimator (KDE) function, wherein the KDE function produces a smoothed estimate of the amplitude track (82), and wherein the smoothed estimate of the amplitude track (82) contains a plurality of peaks (84), each representing a number of times a peak amplitude (68) occurs in the FFT (62);and identifying a highest peak (84A) of the smoothed estimate of the amplitude trace, (82) wherein the highest peak (84A) represents the peak amplitude (68) that occurs most frequently in the FFT (62).

Images