Internet of Things Acoustic Emission for Unattended Quantitative Leakage Monitoring

 
 
ABSTRACT
 
This paper analyzes the theoretical basis of acoustic emission monitoring of leakage in principle. Based on the actual field leakage experiments, the following quantitative relationships are obtained: the quantitative relationship between the amount of leakage and the acoustic emission parameters under the same pressure; the quantitative relationship between different pressures, acoustic emission parameters and leakage rate under the condition of the same valve opening; the characteristic relationship between different leakage openings and acoustic emission parameters under the same pressure difference (tank wall plug). An online experimental demonstration system of leakage rate of water pipes is established. The acoustic emission collector automatically detects the leakage and leakage rates, and transmits the data to the cloud server. After the alarm conditions are set, the alarms will be pushed to the mobile phone if the alarm conditions are met. The cloud server is open for readers to view the results. 
 
Keywords: acoustic emission, leakage, quantitative, Internet of Things, unattended, alarm
 
 
1. Introduction

Valves and pipelines are widely used in all walks of life, ranging from aerospace industry, marine industry, petrochemical industry to domestic water supply and gas supply. However, in the long-term effects of erosion and corrosion in valves and pipelines, it leads to the untight seals in valves and the reduced wall thickness, which often occurs leakage accidents. Although conventional detection methods (such as direct observation method, flow balance method, negative pressure wave method, operating pressure method, etc.) have been widely used, their deficiency is that they rely on professionals to go for the site inspection, and need to scan point by point or even need to dig and check the buried pipeline, which involves high labor intensity and low work efficiency.
 
As one of the important branches of the applications of acoustic emission (AE) technology, acoustic emission leakage detection technology has been widely recognized in recent years due to its advantages of dynamic, high sensitivity and wide coverage. However, the current technology and equipment still rely on acoustic emission technical experts to analyze acoustic emission data in order to determine whether and how much leakage there is. Thus it limits the wide range of industrial applications of this technology. This paper analyzes the theoretical basis of acoustic emission monitoring of leakage in principle. Based on the actual field leakage experiments, the following quantitative relationships are obtained: the quantitative relationship between the amount of leakage and the acoustic emission parameters under the same pressure; the quantitative relationship between different pressures, acoustic emission parameters and leakage rate under the condition of the same valve opening; the characteristic relationship between different leakage openings and acoustic emission parameters under the same pressure difference (tank wall plug). An online experimental demonstration system of leakage rate of water pipes is established. The acoustic emission collector automatically detects the leakage and leakage rates, and transmits the data to the cloud server. After the alarm conditions are set, the alarms will be pushed to the mobile phone if the alarm conditions are met. The cloud server is open for readers to view the results.
 
2. Acoustic Emission Principles and Theoretical Basis of Leakage Monitoring
 
The principle of acoustic emission signals produced by leakage is that when the medium is ejected from the gap due to pressure difference to form turbulent flow, the medium in the turbulent flow has impact and friction with the sealing surface of the medium, which stimulates the elastic stress waves. The signal strength and frequency range of the elastic stress waves are closely related to the turbulent velocity of the medium, which is pressure difference, the leakage rate, the valve medium and the structures. The generated leakage signals propagate along the pipe walls and in the medium. When the acoustic emission sensor is coupled on a reasonable position on the surface of the pipe, the signal can be received. The piezoelectric effect of the sensor is used to convert the elastic wave signals into voltage signals, which are then amplified, analyzed and displayed by the acquisition AE equipment.
     
Typical leakage signals have continuous and random non-stationary characteristics and their frequency distribution has an obvious steep peak, so that it has a certain anti-interference ability. It can be seen from the jet leakage model that the leakage signals are mainly generated in the turbulent mixing zone and the transition zone. The high frequency frictional impact signals are mainly generated near the leakage outlet while the low frequency oscillation signals are mainly generated in the position far from the leakage outlet. Generally speaking, the larger the leakage amount, the greater the energy density of the acoustic emission signals generated from the leakage. When the pressure difference increases, the impact sound caused by the leakage blockage will be generated, which is much higher than the turbulence signals, and these signals can be used as meaningful signals for detection.
 
When the gas or liquid leaks from the leak hole under a certain pressure, continuous mechanical waves are stimulated at the leak hole. When the acoustic emission waveform stimulated by the leak is observed by the oscilloscope, its shape is continuous waves with small amplitude fluctuation without following any patterns. The frequency band distribution of leakage acoustic emission wave varies from several Hz to several hundred kHz depending on the size of the leak hole, the leakage speed and the leakage medium. The suitable acoustic emission sensors are chosen to receive the acoustic emission waves from leaking location, then the mechanical waves are transformed into electrical signals followed by amplification and then are transmitted to the acoustic emission host equipment. After signal processing and analysis, the amount of leakage information is obtained. After the appropriate threshold is set, when the AE signal reaches the threshold, it outputs an alarm. Through the Internet of Things (IoT) communication, the amount of leakage and the alarming AE parameters are transmitted to the Internet cloud platform, following by pushing the messages to the administrator user terminals, to achieve the purpose of intelligent alarming for unattended quantitative leakage monitoring.

3. Quantitative Experiments of Acoustic Emission Leakage
 
3.1. Experiment Background Introduction
The experiment was carried out at a submersible manufacturing company in Shenzhen. Thesubmersible, the external valve and the flow meter used in the experiment are showninFigure2and Figure 3. There is enough volume inside the submersible to pump air at a certain pressure, sothat the pressure can be stabilized in a small range for the subsequent leakage tests. Thevalveinspected is a ball valve, whose nominal diameter is 15 mm. There is no leakage whenthevalveis completely shut. A gas flow meter is connected to the rear end of the valve to quantifytheleakage rate.
The SAEU3H-4 digital acoustic emission detector from Qingcheng AE Institute (Guangzhou)was used in the experiment, with the compatible acoustic emission sensors, coaxial cables,acquisition AE cards and the analysis software to form the whole acoustic emissiondetectionsystem. The acoustic emission sensor was the SR40M resonant sensor (the center frequencywas40 kHz and the frequency range was 15 kHz-70 kHz), and the external preamplifier had40dBgain. Details were shown in the figures below. The pencil lead breaking sensitivity test was99dB.
5. Conclusions
The AE equipment of the Internet of Things can automatically control data collection, dataanalysis and automatic alarm pushing through embedded software and hardware to achievelongterm unmanned quantitative leak monitoring. Through automatic data processing, massiveacoustic emission data can be converted into simple and understandable leakage levels, whichcan eliminate the trouble that traditional acoustic emission technology relies on professionalstoanalyse acoustic emission data, so that users who do not understand acoustic emissioncanquickly get started. Moreover, the data processing is centralized in the collector, and onlyasmallamount of data is output/uploaded, which can effectively reduce the requirements of datacommunication speed and equipment cost. Its advantages of long-termstable operation, automatic alarm, quantitative monitoring and low equipment/operation cost are of greatsignificance to industrial applications.
6. References
[1] Shen, G. (2015). Acoustic emission detection technology and its application. Beijing-Science Press.
[2] Kong, D. (2010). Application of acoustic emission technology in on-line valveleakagemonitoring. Beijing - Beijing University of Chemical Technology.
[3] Xie, J., Liu, S. (2019). Internet of Things Acoustic Emission: Systems and Applications.
[4] Shen, G. (2019). Advances in Acoustic Emission Technology. US - Springer, p. 19-31.
[5] Jia, N., Jiang, X., Sun, J. (2014). Research on valve Leakage Monitoring systemandpositioning system of gas supply pipeline based on acoustic signal. Forest Engineer, vol.30, p. 96-98.

More details please click《Internet of Things Acoustic Emission for Unattended Quantitative Leakage Monitoring》