Detection of ball bearings defects by vibration analysis and implementation of predictive maintenance on ship's machinery

Vibration, very simply put, is the motion of an object forth and back from its beginning position. Vibration is generally undesirable in machinery or its parts. It causes wasted energy in the machine, unwanted sound and noise, high tension, equipment wear and fatigue, and even can be very destructive and can lead catastrophic failure within the entire system. Bearings are vital and widely-used components of the rotating machinery. They are preferred in almost every type of machinery. Inadequate and improper lubrication, temperature extremes, manufacturing defects, the presence of corrosive particles, contamination, humid working environment, overloading, improper mounting and operating errors may occur in bearings. Defects on inner race, outer race or rolling elements cause the sudden changes of deformation contact between the bearing rings and the balls, during the running period. These deformations depending on the external load and the element of the bearing where the defect found cause changes in the resultant force produced by the bearing. With these changes, specific vibrations occur on the shaft supported by bearings. Rotating machines produce vibrations that are function of machine dynamics, such as the mechanical looseness, misalignment and unbalance of the rotating machinery. Measuring the amplitude of vibrations at certain frequencies can provide important information about the accuracy of shaft unbalance, looseness or misalignment. The bearings used in machinery should be taken into consideration the minimum load specified by the manufacturer, and must be done in the adequate and proper lubrication. It is possible with the proper running conditions to achieve the expected life time of bearing. However, only the %10-20 of the bearings can achieve the planned service life. So, it is important to develop proper methods for predicting failure of bearing early enough to allow preventive maintenance. The failure of bearings is one of the foremost causes of break down in rotating machinery. Bearings are an important element and bearing failure is one of the primary causes in rotating machinery on ships. Because of a defect on a bearing, the machine which the bearing running on will be affected and serious problems will arise on the ship's operation. For example, during the sail, an unexpected and sudden downtime of a critical machine will cause high maintenance costs, unnecessary element replacement, difficulty of supplying spare parts, loss of time, stress and pressure on staff and work accidents and, even cancellation of the sail at advance stage. Various maintenance methods are implemented in ship's machinery and equipment to encounter problems. With the periodic change of oil and equipment, some faults can be prevented, but the emergence of unexpected failures cannot be prevented fully. Bearings failures can be detected by predictive maintenance using vibration measurements method while the machine's operating period. With the change of defective bearing and doing planned maintenance in early failure phase, the unexpected downtime can be prevented. The aim of predictive maintenance is firstly to predict when a machine failure might occur, and secondly, to prevent occurrence of the failure by performing any required maintenance procedure. The task of monitoring for future failure allows maintenance to be planned before the failure occurs. Several predictive maintenance techniques are used but the dominant technique is vibration analysis. When predictive maintenance include vibration analysis is implementing effectively as a maintenance method, maintenance is only performed on machines when it is required. That is, just before failure is likely to occur. This brings several advantages such as, increasing equipment lifetime, optimizing spare parts handling, fewer accidents with negative impact on environment, minimizing cost of spare parts and supplies, minimizing the production hours lost to maintenance, reduction the total time spent maintaining equipment. So implementation of predictive maintenance include vibration analysis on ships is an important subject for ships' machine health. With the usage of predictive maintenance include vibration analysis on rotating machinery; bearing failures will be detected during the operation period, before reaching these errors to the dangerous phase. By knowing which bearing needs maintenance, maintenance work can be better planned and what would have been "unplanned stops" are transformed to shorter and fewer "planned stops". In this thesis, firstly; mechanical vibrations and some basis subjects were introduced to provide theory to allow the concepts of vibration and its analysis to be understood. Issues related to vibration analysis and data collection and vibration standards for ships mentioned. Then, information on maintenance and its implementation in ships were given. Preventive maintenance, breakdown maintenance and proactive maintenance were introduced and preventive maintenance was detailed. And finally; after bearings and bearings components, defects and characteristic defects frequencies were mentioned, and passed the experiment phase. To investigate the effects of a bearing failure of the fatigue induced defects, which has the highest percentage of failure with the rate of 34%, via vibration analysis, an experimental setup designed. Firstly; faultless bearing was operated at 1750 RPM and 2250 RPM, and vibration data were recorded. Then inner and outer race running surface defective bearings (the defect sizes are 0.15 mm, 0.45 mm and 0.90 mm) were operated at the same velocity, and data were recorded. These records are presented in a detailed manner with graphs and tables. As a result of detailed analysis of the frequency spectrum of defected and faultless bearings, it is revealed that manufactured shaft for use in experimental setup is unbalanced and the experimental setup has a mechanical looseness because the bearing support is not fully grasp the bearings. The read frequencies from spectrum which belongs to inner and outer race running surface defective bearing, and the frequencies calculated from numerical equations were observed to be in good agreement. The size of defects increases the amplitude of characteristic defect frequencies. In addition, mechanical unbalance and looseness can be detected via vibration analysis. As a result, bearings defects and defect sizes, unbalance and mechanical looseness can be detected by predictive maintenance include vibration analysis. Because bearings are important elements on rotating machinery on ships, implementation of predictive maintenance using vibration measurements method in ships for early failure detection becomes more important.