Data Acquisition System and Signal Processing Technique for Bearing Fault Analysis

Abstract

Background/Objectives: In rotating machinery bearings plays an important role to reduce rotational friction and support radial and axial loads. Abnormal vibration affects the lifetime of machinery. This abnormality may causes by bearing Fault. In order to monitor the machines health and it is needed to focus the functionality of bearing components like ball, inner race, outer race and cage component. Methods/Statistical Analysis: There are many methodologies available to measure and analyze the fault of bearing but they utilize expensive resource to identify the bearing system. The proposed system explains and utilizes a cost effective, less time consuming system, is used for bearing faults identification. The novel system provides real-time data acquisition and signal analysis for Fault detection. Low cost Micro Electro Mechanical System (MEMS) Accelerometer sensor is used for vibration measurement, LABJACK U3 Data acquisition system connected with Raspberry Pi-2 CPU which acquires and process with high performance of acquired data by using python application software. Application/Improvements: The online signal processing technique explains the abnormality of bearing time domain data's amplitude easily classify the faulty bearing. Fast Fourier Algorithm converts time domain to frequency domain which easily represents faulty components frequency with high amplitude. Short Time FFT (STFT) shows the color map with high color intensity of vibration amplitude of faulty components and its time instant. These data representation technique is utilized to easily identify fault of bearing components.