Health Monitoring of Welded Pipelines with Mechanical Waves and Fuzzy Inference Systems

Morteza Mohammadzaheri, Ahmad Akbarifar, Mojtaba Ghodsi, Issam Bahadur, Farooq Al Jahwari, Badar Al-Amri

Research output: Contribution to conferencePaperpeer-review

10 Downloads (Pure)


This research presents a cost effective vibration-based pipeline integrity identification method for welded pipes using mechanical waves and fuzzy inference systems. At the moment, in terms of instrumentation, there are two approaches towards integrity identification of pipelines: (i) measurement throughout the pipeline (ii) measurement at a limited number of points along the pipeline. The second approach is normally much less expensive. Use of ultrasonic waves is the only widely used method of this approach. This paper introduces an alternative, in the second approach: use of mechanical waves. That is, the pipeline is mechanically excited at a point, and the mechanical response (e.g. acceleration) is measured at another point. Advantageously, mechanical waves are inexpensive to generate and measure. In addition, it is well known that any change in pipe structure affects the measured mechanical response. However, this effect is very complex, so that it is practically impossible for human beings to interpret information of such tests. This research focuses on fault isolation (location) on welded pipelines and employs Fourier series, statistical analysis and fuzzy inference systems to interpret the recorded responses. Preliminary results are promising and show a standard deviation of 3.8 meters. These results can improve significantly with increase of the tests used for data analysis.
Original languageEnglish
Publication statusPublished - Feb 26 2020
EventInternational Gas Union Research Conference, - Muscat, Oman
Duration: Feb 24 2020Feb 26 2020


ConferenceInternational Gas Union Research Conference,


Dive into the research topics of 'Health Monitoring of Welded Pipelines with Mechanical Waves and Fuzzy Inference Systems'. Together they form a unique fingerprint.

Cite this