SIMULATING GAS TURBINE BEARING FAILURE TOWARDS EFFECTIVE CONDITION MONITORING

Abstract

Recently, the requirements for the reliability of machines that perform different technological processes, their vibrational effects, and the failure of their bearings pose a great problem to gas turbine manufacturers and equipment owners. On most gas turbine plants, heavily-loaded slow rotating bearings are used which are in constant movement and vibration. A step-by-step method to the reduction of these vibrational forces led to this research. Thus, vibration readings were obtained from Afam power station to enunciate step towards resolving the problem. The data collected from the plant was hence used in the simulation of likely defects that could lead to down time resulting from vibration. A computer program code in Visual C#.Net was used to actualize the simulation. Results obtained show that early detection of imminent faults could help to prevent downtime. The results also show that the defect meted to the gas turbine depends on the amplitude of vibration measured at the bearing housing. Since the bearing at the turbine end carries more load than others, it should be constructed rigidly. Consequently, the results obtained further indicated that a speed of 3043 rpm is to be avoided due to the risk of resonance. Also, high vibrational amplitude was observed which can be attributed to the effect of imbalance. Generally, it is evident from the work that vibration analysis is very effective in condition monitoring of the health of a gas turbine and that its implementation is most efficient for detection and identification of incipient defects to its bearing.