Petrol Engine Fault Detection Using Mechanical Vibration Analysis

Abstract

Vibration monitoring is the most widely used condition monitoring technique, where the machine's vibrations are analyzed to determine incipient faults. The early detection of a fault has many advantages as detecting a simple fault before causes too complicated problems. Every machine contains mechanical parts that emit a vibration. These vibrations can be monitored and recorded as a reference signal, displaying it using different strategies such as the time domain, frequency domain, time-frequency domain simultaneously. The root mean square (RMS) method is used to analyze the signal to obtain the signal code.

This paper presents a monitoring method, displaying, analyzing and coding of the vibration signal using a newly invented device called Immediate Diagnosis Device (IDD) presented by the same authors to detect the Daewoo Lanos engine faults using a mobile device. A vibration sensor (model: 333B32) was installed on the engine body to monitor the engine vibrations at different engine speeds 900 and 1500 rpm. The measured signal is fed to IDD and passes through a series of operations and processing in its components to assess the engine fault. Through IDD, the vibration signals were sent to the processing unit microcontroller 32BIT- 72MHZ-12BIT ADC, Serial Interface (STM32F103C8T6) to store and process it. Then compares it with the reference data by using Microsoft Visual Studio and the C ++ language program, the final result will be displayed a message on the device screen LCD (16X2 8BIT) to explain the fault reason.