Leakage position detection in a straight pipe by using sound signals and a simple microphone

Read full text
Authors: Feblil Huda, Tengku Hamzir Muhardillah Yahya, Syafri, Meifal Rusli, Ishkrizat Bin Taib
Cited by: 0
References: 18

Abstract

Abstract Piping systems play a crucial role in modern society and industry, serving as a fluid transport system with a high level of integrity. Yet, they also encounter significant problems, including leakages. Several researchers have developed methods for detecting leak positions in pipes. The current detection method uses pressure differences and ultrasonic waves, which are relatively expensive, and the instruments, equipment, and sensors used are very expensive. In this study, a leak detection system for straight pipes using sound signals was developed. For this purpose, a test setup was created using a straight PVC pipe with a length of 1.55 m as the test object, a loudspeaker to provide sound excitation into the pipe, a sound microphone to capture reflected sound from inside the pipe, and a sound card to convert analog data to digital. The loudspeaker provides sound excitation in the form of impulses into the pipe, and the microphone captures the reflected sound inside the pipe. The sound signal is then processed using signal processing software with a wavelet transform to extract the position of the leak. The leak position can be detected very well, which is indicated by a higher peak value of the wavelet coefficient. From the test, the results obtained were that the greater the damage given, the greater the value of the wavelet coefficient produced, with an error in position detection of less than 5%. It has been confirmed that the proposed method works very well in detecting the position of leaks in pipes, promising low-cost damage detection.

Fields of Study & Subjects

Water Systems and Optimization Ultrasonics and Acoustic Wave Propagation Non-Destructive Testing Techniques

Publication details

Journal / source
Journal of Physics Conference Series
Volume
3186
Issue
1
Pages
012024–012024
Publication date
1 March 2026
Language
EN