Study on Acoustic Emission Monitoring Technology during Fatigue Test of a Certain Aircraft

Research on Acoustic Emission Monitoring Technology during Fatigue Test of a Certain Aircraft - Acoustic Emission Monitoring during Landing Gear Retraction and Extension Control Test

Abstract: The landing gear retraction and extension control system test is part of the fatigue test of a certain aircraft. The areas that require acoustic emission (AE) monitoring include key parts such as the upper lock and actuator lug of the main landing gear and the front landing gear, and the lower lock of the main landing gear. In view of the strong impact and noise interference during the test process, the trend analysis and correlation analysis methods of acoustic emission signal parameters are used to achieve real-time monitoring of multiple targets and dynamic objects, and successfully monitor faults such as actuator leakage and hinge wear. The method used has fast analysis speed and good real-time performance, which can be used as a reference for acoustic emission monitoring technology in subsequent fatigue tests. Keywords
: aircraft; fatigue test; fatigue crack; acoustic emission (AE); trend analysis
0 Introduction
A certain type of aircraft is the third-generation main combat aircraft of the Air Force. The main purpose of the full-aircraft fatigue test and non-destructive testing during the test is to understand the fatigue damage and fatigue crack propagation laws of key structural parts (especially some inaccessible structural parts), which can provide important basis for fatigue life determination, repair cycle determination, and fatigue detail design improvement of future improved models (Type B) of this type of aircraft. Therefore, this test and research are of great significance. The
fatigue test of the landing gear retraction and extension control system is part of the full-aircraft fatigue test of the aircraft. The main purpose is to determine the fatigue life and fatigue damage status of the landing gear auxiliary parts (such as the main landing gear upper lock, lower lock and actuator support and lug) during the retraction and extension process. In order to ensure that no sudden failure of the landing gear auxiliary parts occurs during the test and to obtain useful fatigue damage data, AE technology is used to track and monitor the test process in real time. In the past, AE monitoring of landing gear fatigue tests was performed on the landing gear itself alone, that is, the landing gear was only regarded as an independent component and was not installed on the aircraft [1]. The landing gear in the test described in this article is installed on the aircraft and is carried out as a component of the aircraft. It can more realistically and objectively reflect the damage process of the landing gear, and its AE monitoring is also more difficult. The characteristics of landing gear AE monitoring are to monitor multiple target objects at the same time, and the target objects are moving parts, which brings great difficulties to real-time monitoring.
1 Monitoring principles and strategies
There are many parts that need to be monitored by acoustic emission during the test, and the workload of signal acquisition and processing is too large. Therefore, it is very important to optimize the monitoring points and determine the key monitoring parts, so as to ensure the smooth progress of the monitoring work. Another important issue that must be considered is the use of appropriate signal processing methods. Since fatigue tests are carried out continuously and the noise interference during the test is particularly serious, the amount of data in the test is large. It is particularly important to use appropriate signal processing methods to use AE technology to obtain information related to fatigue damage and fatigue cracks. Although under certain conditions, waveform analysis (including modal acoustic emission technology) can effectively identify fatigue cracks and noise interference signals, considering the real-time nature of monitoring work and the timeliness of data processing work, this method will only be used in a local area when it is considered necessary.
The judgment of whether damage has occurred is based on the following principle: the monitoring instrument system and the simulated loading system constitute a complete fatigue test monitoring system for aircraft landing gear accessory parts. For the loading system and monitoring instrument system that are in stable operation, if the signal of a certain monitoring target object changes significantly, it means that the stable environment in a certain area where the target object is located may change, that is, damage may have occurred. When this happens, other methods (such as amplitude distribution, correlation analysis, etc.) are used for analysis and verification, and other non-destructive testing methods are used for detailed testing to further verify.
The processing and analysis of AE signals involves a variety of methods [2-5]. In order to adapt to the characteristics of real-time monitoring, multiple target objects, large signal data volume, and the use of resonant sensors (large distortion of the source waveform), the classic acoustic emission signal characteristic parameter analysis method is selected. This method is simple, intuitive, fast in analysis speed, good in real time, and suitable for engineering monitoring. By performing trend analysis on the changes of specific parameters (number of hits) with the number of landing gear retraction and extension times [2], the state changes of the relevant area can be judged. A prerequisite for the successful application of trend analysis technology is that the acoustic emission instrument used should have low noise, high stability, sufficiently wide dynamic range and high data transmission rate.
2 Monitoring objects, instrument settings and sensor installation
There are 11 target objects that need to be monitored in real time, which are the upper locks (3) of the left and right main landing gear and the front landing gear, the lower locks (2) of the left and right main landing gear, and the ears at both ends of the left and right main landing gear and the front landing gear actuator (3 ears with the landing gear and 3 ears with the body). They are all key parts that need to be paid attention to in the landing gear retraction and extension test. Figures 1 and 2 are the upper locks of the right main landing gear and the ears of the left main landing gear actuator and the landing gear, respectively.

Figure 1 Right main landing gear uplock Figure 2 Left main landing gear actuator ear
monitoring instrument is the upgraded PAC company's full digital acoustic emission system MISTRAS 2001, and then replaced with PAC's full digital 20-channel DiSP acoustic emission system. The sensor is a resonant R15 type, equipped with a 1801A preamplifier with a gain of 40dB, a 1MHz sampling rate, and a threshold value of 40dB.
In order to ensure the smooth implementation of real-time monitoring, the sensor must be installed and fixed in an appropriate position close to the target object. Due to the limited operating space (see Figures 1 and 2) and the strong vibration during the fatigue test, according to previous experience, 704 silicone rubber is used as a coupling agent to bond the sensor to the target object. The coupling agent can provide very good sound transmission function (i.e., small sound wave attenuation) and strong bonding ability. During the fatigue test, the sensor will not fall or peel off from the bonding surface. At the same time, when necessary, the sensor can be removed more conveniently without damaging it. It is important to fix the preamplifier and the signal line to avoid interference with the components in motion, causing unnecessary damage and affecting normal monitoring.
3 Results and discussion
A total of 7500 cycles of landing gear retraction and extension tests were carried out (each cycle is about 45s), and routine non-destructive monitoring and maintenance were performed every 500 cycles. The first 1000 cycles were used for trial production and monitoring instrument adjustment, and the actual monitoring and data collection started from 1260 cycles. Consider the following factors: a. Fatigue damage has a development process, and the damage will generally not disappear suddenly within a few cycles after it occurs; b. Acoustic emission sensors are very sensitive, and changes in the state of the target object will cause obvious changes in the collected signal; c. Although the amount of signal data is large, if there is no sudden change in the state of the local area of ​​the target object, the data collected and processed are repetitive data, and there is no need to collect and process all of them. Therefore, in real-time monitoring, 10 cycles of data are collected for every 30-40 cycles of landing gear retraction and extension for processing and analysis. Figures 3 and 4 show that under normal conditions, the received acoustic emission signals have strong regularity and their variation range is within a reasonable range. This is the basis for using trend analysis to make a preliminary judgment on the system status. The
trend analysis of data is carried out as follows: Each time data is collected, firstly, roughly determine whether the relative amount of the number of hits (AE hits) of each target object has changed significantly (the number of hits of each target object is quite different, but the relative difference is relatively stable). If there is a change (see Figures 5 and 6), then determine whether the number of hits of the target object has increased significantly. If the value of the relative increase β (=△hits/hits) exceeds 30%, other methods should be used for detailed analysis, otherwise there is no need for further analysis.

X-axis: time/s Y-axis: hit number
Figure 3 Relationship between the number of acoustic emission hits and time of the main right upper lock in a cycle (45s) of landing gear retraction and extension

Figure 4 Changes in the AE impact number of the main lift left actuator within 10 loading cycles (1257-1266) x: t/s; y: AE impact number

After the landing gear was retracted and extended for 4500 cycles, the number of impacts of the two sensors installed at the ear (Figure 2) and the middle of the piston cylinder (not shown) on the left main landing gear actuator of the target object began to increase significantly. Figure 7 shows the trend of the number of impacts of the sensor at the middle of the piston cylinder (a, the signal was stable before 4500 cycles, and there was a small mutation in the signal at 3020 cycles, which was caused by the wear of the actuator ball hinge; b, the signal changed dramatically from 4500 to 5540 cycles, which was caused by the wear of the actuator sealing rubber ring and the leakage of hydraulic oil; c, the actuator was disassembled at 5540 cycles, confirming that the sealing ring was severely worn (Figure 8). After replacing the sealing rubber ring and reinstalling the sensor, the signal returned to normal (low AE impact number). After 7250 cycles, the signal changed dramatically, which was also caused by the wear of the actuator sealing rubber ring and the leakage of hydraulic oil. Since the test was close to the end, there was no shutdown at this time until the 7500 test tasks were completed.

Figure 7 Relationship curve between the number of impacts and the number of retraction and extension cycles of the left main landing gear actuator
The above results show that although the actuator is located in a hidden position and its leakage cannot be observed visually, it is possible to make a correct judgment by analyzing the AE signal changes of the target object. Take the signal at the left main landing gear ear (see Figure 2) as an example for analysis. The retraction and extension process of the landing gear is as follows: retract to the upper lock for 7 seconds → stop for 8 seconds → lower to the lower lock for 5 seconds → stop for 25 seconds. The normal signal monitored is shown in Figure 9: the signal is stable and has strong regularity. If fatigue cracks occur in the monitored area, the AE signal should change significantly during the retraction process of overcoming the external load. The actual fault signal monitored is shown in Figure 10. The signal becomes unstable and irregular, and there are a large number of AE signals when the landing gear is in the lower stop position (there should be no signal), while the signals of other target objects do not change. It can be judged that the actuator has a leak, not a fatigue crack.

After the landing gear retraction and extension test was conducted for 3000 times, the left main starter cylinder made a creaking sound during operation. This was because the ball hinge connecting the cylinder and the fuselage had dry friction, resulting in a certain amount of wear. After the sensor was installed, a lead break test was performed. The AE signal at this position can be received by the sensor installed in the middle of the cylinder. Although the number of collisions did not increase much, from an average of 14 times to 19 times, and its relative increase was just over 30%, the correlation analysis of the AE signal showed a significant change in the signal. The correlation analysis of the AE signal is to obtain the characteristics of the signal by analyzing the relationship between two characteristic parameters to evaluate the state of the monitored target object. Figures 11 and 12 are the correlation distributions of the number of collisions and amplitudes of normal signals and fault signals, respectively. When there is a fault, the number of collisions with an amplitude of 100dB increases from less than 5 to 65, and the low-amplitude signal also increases. Therefore, even if the sound of friction and wear is not heard, the fault can be monitored. After decomposing the ball hinge, it was found that there was obvious wear and damage. After lubrication treatment, the fault phenomenon disappeared. 5 Conclusion During the landing gear retraction and extension control test, acoustic emission technology is an effective method for real-time monitoring of fatigue damage of its auxiliary components. Although no fatigue cracks were generated, relatively weak leakage faults and dry friction and wear faults were monitored. Typical parameters of acoustic emission signals, such as the number of hits, can provide very valuable information about the state of the target object. The trend analysis method and correlation analysis method of acoustic emission signal parameters are simple, intuitive, fast, and have good real-time characteristics. They realize real-time monitoring of multi-target and dynamic objects, which can ensure the smooth progress of the test and avoid sudden accidents. It has guiding significance for subsequent full-aircraft fatigue tests. Of course, it should be pointed out that an important premise of the above analysis is that a fully digital acoustic emission detector with stable and reliable performance and extremely low background noise is an indispensable material guarantee condition to ensure the reliability of the results. The research conducted by the author shows that acoustic emission technology has a good application prospect in the real-time monitoring of engineering fatigue damage.