AN INITIAL STEP IN A BLIND SOURCE SEPARATION METHOD TO DETERMINE THE BASELINE SIGNAL WITH ACOUSTIC EMISSION

• Published in: Journal of Theoretical and Applied Information Technology 2017-01, Vol.95 (1), p.220
• Main Authors: Thenu, Novitha L Th, Arifianto, Dhany, Ariana, I Made, Zubaydi, Achmad
• Format: Article
• Language: English
• ISSN: 1817-3195

The ability to diagnose problems in marine engines quickly and accurately is vital to the safe operation of any ship. Changes sound emanating from the operating system inside the engine room of a ship into vital information for early detection of the engine condition. This paper presents baseline signal measurement technique to detect early cracks that occur on the propeller shaft. Baseline signal measurement is a first stage application of Blind Source Separation method for separating a complex sound. The study was conducted in an anechoic chamber. Test of three shafts have been conducted by varying depth of crack. The first test was shaft without crack (healthy shaft), the second shaft was cracked shaft with the depth of crack is 25% of shaft diameter named cracked shaft 0.25D and the third shaft has 50 % depth of crack (cracked shaft 0.5D). The tested shaft was rotated at shaft speeds from 500 to 1000 rpm by an electric motor. The microphone as Acoustic Emission sensor was placed at various distances as long as of test shaft from 5 to 40 cm. The accelerometer attached on the casing of the electric motor. The sound of data is measured for 5 seconds and then processed through signal processing. Acoustic signal spectrums on each measurement position are different from one another. Note that the spectrums amplitude are approximately equal to the microphone position which are 30 and 40 cm. Locations around the crack are in the measurement position. Another result shows that the amplitude value will increase when the shaft speed improved. These findings indicate that the acoustic emission technique may be a suitable replacement to an accelerometer-based measurement. Furthermore, the specific finding is the initial step of baseline measurements successfully performed as data references to the application of Blind Source Separation method.