CD Tesis
Tomografi Korosi Selubung Dalam Pipa Menggunakan Sinyal Suara Berbasis Simple Microphone
Corrosion in pipeline systems is a significant challenge in the industry as it can compromise structural integrity and increase the risk of operational failure. Conventional methods, such as visual inspection and electromagnetic sensors, often fail to provide a comprehensive representation of the internal condition of pipelines. This study proposes the use of acoustic reflectometry to detect and visualize corrosion within pipes by analyzing reflected sound signals using Continuous Wavelet Transform (CWT). Signals were captured from various positions around the pipe's circumference, generating eight distinct signals that were then processed into a 3D image. The study comprised an initial test using a PVC pipe with artificial corrosion made of plasticine and a primary test on pipes with actual corrosion. Results showed that the width and length of the corrosion significantly affected the reflected wave patterns; corrosion spanning the full circumference produced higher signal amplitudes, while longer corrosion resulted in larger waveforms. This method successfully identified corrosion positions longitudinally with a maximum error of 8.97%. However, accuracy in angular position detection remained low due to the mismatch of Wavelet coefficients caused by sound signals not being concentrated on the microphone position. The study was validated through finite element simulation, yielding more stable data than experiments due to the exclusion of external disturbances. The generated 2D and 3D images effectively represented the position and severity of corrosion longitudinally, although further development is needed to improve angular position accuracy.
Keywords: Corrosion, pipe, sound signal, Wavelet transform, tomography
Tidak tersedia versi lain