Early warning of bearing failures
Monitoring for early indications of bearing damage enables corrective action to be taken well ahead of bearing failure. Noise, high temperature running and excessive vibration are all indicative of imminent failure, allowing repair or replacements to be planned proactively rather than reactively, as Phil Burge, marketing and communications manager at SKF, explains
Bearings that are worn or damaged usually exhibit easily identifiable symptoms. While an experienced maintenance technician will be able to assess these by touch, hearing or sight, at this stage the damage may already be extensive and the time to failure that much shorter. Achieving early detection is thus advantageous and this is where condition monitoring trumps the abilities of even the most experienced of maintenance engineers.
The advantage of employing objective instrument-based technologies, as opposed to relying on the subjective assessment of an individual during a routine inspection, is that damage is detected at the earliest stage of development before it becomes problematic. By using professional condition monitoring instruments and advanced analytical tools such as enveloped acceleration, the all-important pre-warning time can be maximised, allowing replacements to be made as part of planned maintenance activity, rather than during unexpected downtime.
Early warning
Monitoring the condition of bearings on a regular or continuous basis will enable the early signs of bearing damage to be detected. It will also allow trend analyses to be produced, to ensure that the underlying causes of bearing wear are identified. Vibration may be the result of one or a combination of factors, including contaminants or particulates entering the bearing housing or raceways, which causes corrosion or accelerated wear; damaged bearing housing surfaces; shaft misalignment; unbalanced, poorly distributed or excessive loads; uneven wear causing eccentricity in the rotating parts, and so on.
Bearings that are well maintained and which are in good condition may emit a consistent but virtually inaudible sound signature. Any audible noise, however, such as grinding, squeaking and other irregular sounds, will indicate that the bearings are in poor condition, that shaft alignment has been compromised, or that operating conditions have changed adversely. Although all machines vibrate to some extent, it should be noted that any deviation from the normal vibration signature is likely to indicate the onset of a mechanical problem. By analysing the characteristics of the vibration signature it is often possible to pinpoint the actual source and nature of the fault.
Monitoring the temperature of bearings is also important. An overheated bearing may be due to loss of lubrication, over-lubrication or a breakdown in the structure of the lubricant because it was incorrectly specified for the conditions under which it is operating. Badly located, poorly sized or deteriorating seals will also contribute to an overall rise in bearing temperature. So long as all other operating conditions remain unchanged, a temperature rise in a bearing is often an indication that it is becoming damaged in some way.
One of the most common and reliable methods of monitoring the condition of bearings is to measure their vibration levels, either on a continuous basis, in real time, or at predetermined intervals. In the case of both temperature and vibration monitoring, measurement can be carried out periodically using handheld instruments fitted with suitable probes. These instruments can be extremely sophisticated, enabling the storage and time-stamping of measurements and providing detailed analysis of operating conditions. An alternative approach is to modify bearing housings to incorporate wireless or hard-wired vibration sensors, which provide continuous, real-time information.
Whichever option is chosen it is important to standardise on the method of measurement, and to monitor conditions at set intervals, to build up a consistent picture of machine and bearing conditions over time. This will allow any deviation to be identified at an early stage and for appropriate maintenance measures to be planned well in advance of any shutdown, creating an environment where uptime and machine availability can safely be predicted.
IPE publishes a weekly eNewsletter, delivering a carefully chosen selection of the latest stories straight to your inbox.
Subscribe here

