DIAGNOSIS OF GEARBOX FAULT USING ACOUSTIC SIGNAL

Transcription

1 International Journal of Mechanical Engineering and Technology (IJMET) Volume 9, Issue 4, April 2018, pp , Article ID: IJMET_09_04_030 Available online at ISSN Print: and ISSN Online: IAEME Publication Scopus Indexed DIAGNOSIS OF GEARBOX FAULT USING ACOUSTIC SIGNAL Dr. G Diwakar Professor, Mechanical Engineering Department, Koneru Lakshmaiah Educational Foundation, Vaddeswaram, Guntur, Andhra Pradesh, India R Dileep, P Satya Nikhit, N H S Venkata sai, A Bala Krishna UG Scholar, Mechanical Engineering, Koneru Lakshmaiah Educational Foundation, Vaddeswaram, Guntur, Andhra Pradesh, India ABSTRACT Gear-Box is an important component in automobile which transfers the power from the engine or motor to wheels by performing necessary speed reductions. Any fault in gearbox leads to reduction in speed and performance. Gearbox maintenance is of utmost importance in monitoring the condition of automobiles. Condition monitoring is one of the predictive maintenance techniques which helps in reducing maintenance costs and is of great importance in the modern industrial applications. In this paper acoustic signal generated by gear-box is used for diagnosing the fault. The sound signal generated by the gear box is analyzed in good and faulty conditions and the results are presented. Keywords: Gear box, Gear tooth, Acoustic signal and Microphone. Cite this Article: Dr. G Diwakar, R Dileep, P Satya Nikhit, N H S Venkata Sai and A Bala Krishna, Diagnosis of Gearbox Fault Using Acoustic Signal, International Journal of Mechanical Engineering and Technology, 9(4), 2018, pp INTRODUCTION Gearbox is a chief component in the power transmission in case of automobiles and other machinery. It is used to increase or decrease speed. In case of automobiles, it takes input from clutch shaft and gives output to the driver shaft, which is later transferred to the wheels. So, it becomes important to periodically monitor and diagnose the faults in gearbox. Condition monitoring is a maintenance technique in which parameters like temperature, vibration or an acoustic signal is monitored to identify a significant amount of change that indicates fault is developed or being developed. In condition monitoring there are different types of condition monitoring techniques like Acoustic signal emission, Infrared thermography, Vibration signature analysis and diagnosis, Analysis of Lubricant, Ultra-sound testing and Motor-Current-Signature-Analysis editor@iaeme.com

2 Dr. G Diwakar, R Dileep, P Satya Nikhit, N H S Venkata Sai and A Bala Krishna Vibration-signature-analysis is a technique which is used to detect any change in condition of vibration of the gearbox in contrast to normal operating condition [1-2]. Acoustic-Signal-Analysis is another method which can be used in a similar way to detect any change in normal operating condition of gearbox by analyzing deviation of sound loudness and occurring peak values at the corresponding frequencies. Acoustic-Signal-Analysis is used to diagnose the fault using frequency domain plots and spectrum analysis [3]. The data acquisition can be done by using an accelerometer which gives vibration as output wave [4]. After acquiring the data, methods like Support Vector Machine also known as SVM, which can be used to draw the hyper plane on the graph and sets the limit of decision parameter above which it is considered to be a fault in equipment. SVM produces great accuracy in classifying fault and its diagnosis [5]. Fault detection and diagnosis can also be done using statistical methods such as analyzing the RMS value, shape-factor, amplitude peak value, peak to peak, crest-factor and kurtosis etc. With the advent of SCADA (Supervisory-Control- And-Data-Acquisition) systems the monitoring became easy. SCADA systems use parameters like statistical methods, Time-domain analysis, Cepstrum analysis, Time synchronous averaging, Fast-Fourier Transform and Wavelet transforms etc. [6]. Development of powerful and efficient microprocessors has made online condition-monitoring easy and efficient. A computer application is necessary in carrying out online condition-monitoring [7]. Amplitude and phase modulation is a commonly used technique in fault diagnosis of gears [8]. In this paper, microphone is used for acquiring sound signals from gearbox. The acoustic signal received from microphone is used to detect and analyze gear box faults. The audio signal is recorded from the gear box before and after artificially introducing gear tooth fault. The signal is loaded into the MATLAB and spectrum is developed with the help of Fast Fourier Transform (FFT). 2. CONSTRUCTION OF EXPERIMENTAL WORK 2.1. Gearbox specifications: The splendor bike gearbox is selected for performing the experiment. Its specifications are given below. Gear No. Table 1 Gearbox details Teeth on gear of input shaft Teeth on gear of output shaft Gearratios Gear I Gear II Gear III Gear IV Electric motor specifications: AC type, single phase and 50 Hz operated electric induction motor is chosen for our requirement and its specifications are as follows. Speed Table 2 Specifications of electric motor power Operating voltage Operating current 1440 rpm 0.25 HP V A editor@iaeme.com

3 Diagnosis of Gearbox Fault Using Acoustic Signal Output shaft of above specified electric induction motor is coupled to input shaft of above specified gearbox using rigid coupling and whole assembly is fitted on a wooden board using bolts, nuts, washers and dampers. Dampers are provided to pacify the unnecessary vibrations or noise and to receive the readings as accurate as possible. The whole assembly is covered to prevent external disturbances Assembly line-diagram: After assembly the whole setup is: Figure 1 Final Assembly Figure 2 Experimental setup 3. EXPERIMENTAL PROCEDURE Motor is connected to normal household power supply (AC). The equipment is operated in peaceful environment, by placing the microphone inside the gearbox. Audio recordings for each gear before and after tooth breakage are taken. After converting all the recorded files into the mp3 format, waveforms are loaded to Matlab in.mp3 format since Matlab accepts only.mp3 and.wav type audio files. In the Matlab time and frequency domain plots (using FFT) are obtained. Graphs are plotted between loudness (i.e. decibels) and frequency (in Hz), for which single sided magnitude spectrum is chosen. Figure 3a Gears assembly inside gearbox before gear tooth breakage editor@iaeme.com

4 Dr. G Diwakar, R Dileep, P Satya Nikhit, N H S Venkata Sai and A Bala Krishna 3.1. Matlab source code: Time domain plot: [y fs] = audioread('x.mp3'); N = length(y); samples = 0:N-1; t = samples/fs; plot(t,y) xlabel('time - seconds') ylabel(loudness - decibels') title('time-domain-plot') Figure 3b Gears in gearbox after breaking tooth of 3 rd and 4 th gears Frequency domain plot: [y fs] = audioread( X.mp3 ); N = length(y); Y_mag = abs(fft(y)); bin_values = [0:N-1]; freq_hz = bin_values*fs/n; M_2 = ceil(n/2); figure( ); Plot(freq_Hz(1:M_2), 10*log10(Y_mag(1:M_2))) xlabel( Frequency in Hertz ) ylabel( Magnitude in decibels ); title( magnitude spectrum (single sided) ); editor@iaeme.com

5 Diagnosis of Gearbox Fault Using Acoustic Signal axis tight 4. PLOTS OF RECORDED AUDIO SIGNALS 4.1. Time domain plot (Healthy gearbox): Figure 4 Gear I Figure 5 Gear II Figure 6 Gear III Figure 7 Gear IV 4.2. Time domain plot (Faulty-After breaking one tooth of 3 rd and 4 th gears): Figure 8 Gear I Figure 9 Gear II editor@iaeme.com

6 Dr. G Diwakar, R Dileep, P Satya Nikhit, N H S Venkata Sai and A Bala Krishna Figure 10 Gear III Figure 11 Gear IV 4.3. Frequency domain plot (FFT of healthy gearbox): Figure 12 Gear I Figure 13 Gear II editor@iaeme.com

7 Diagnosis of Gearbox Fault Using Acoustic Signal Figure 14 Gear III Figure 15 Gear IV 4.4. Frequency domain plot (Faulty-gearbox): Figure 16 Gear I editor@iaeme.com

8 Dr. G Diwakar, R Dileep, P Satya Nikhit, N H S Venkata Sai and A Bala Krishna Figure 17 Gear II Figure 18 Gear III 5. RESULTS AND DISCUSSION Gear NO. Healthy Figure 19 Gear IV Table 3 Gearbox details Faulty Frequency peaks (Hz) Loudness (decibels) Frequency peaks (Hz) Loudness (decibels) Gear I Gear II Gear III Gear IV editor@iaeme.com

9 Diagnosis of Gearbox Fault Using Acoustic Signal For, healthy gearbox, the peak value is obtained at the frequency of 73.5 (averagely) with amplitude of around 40.8 decibels. For faulty gearbox, the peak value is obtained at the frequency of (averagely) with amplitude value of around decibels. Through analysis of audio signal, fault detection and diagnosis is done and the equipment can be monitored using acoustic signal analysis. 6. CONCLUSION The acoustic signals are acquired for healthy and faulty conditions of the gearbox at different gears. Analysis of frequency-domain is performed on the acquired acoustic signals which are in time-domain form, with the help of Fast-Fourier-Transform function in Matlab. It is observed that loudness amplitudes and frequency peaks differ for the healthy and faulty gearbox conditions. Hence, from the deviation in loudness amplitudes and frequency peak values from the normal values of healthy gearbox, it can be said that the condition of gearbox at which unusual frequency peaks and amplitudes occur should be diagnosed with fault. REFERENCES [1] G Diwakar, MRS Satyanarayana and P Ravi Kumar. Detection of Gear fault using vibration analysis. International Journal of Emerging Technology and Advanced Engineering, 2, 2012, pp [2] V Ranjith Kumar, P Venkata Vara Prasad and G Diwakar. Detection of Gear Fault Using Vibration Analysis. International Journal of Research in Engineering and Science, 3, 2015, pp [3] Zijun Zhang, Anoop Verma and Andrew Kusiak. Fault analysis and condition monitoring of the wind turbine gearbox. IEEE transactions of energy conversion, 27, 2012, pp [4] P Vecer, M Kreidl and R Smid. Condition indicators for gearbox condition monitoring systems. Acta polytechnica, 45, 2005, pp [5] Achmad Widodo and Bo-Suk Yang. Support vector machine in machine condition monitoring and fault diagnosis. Mechanical systems and signal processing, 21, 2007, pp [6] Fausto Pedro Garcia Marquez, Andrew Mark Tobias, Jesus Maria Pinar Perez and Mayorkinos Papaelias. Condition monitoring of wind turbines: Techniques and methods. Renewable energy, 46, 2012, pp [7] T Praveen Kumar, B Sabhrish, M Saimurugan and K I Ramachandran. Pattern recognition based on-line condition monitoring system for fault diagnosis of automobile gearbox. Measurement, 114, 2018, pp [8] P D McFadden. Detecting fatigue cracks in gears by amplitude and phase demodulation of the meshing vibration. Journal of Vibration, Acoustics, Stress, and Reliability in Design, 108, 1986, pp editor@iaeme.com