Slow-Speed Oven Motor Bearing: Failure Detection Using Ultrasound

Vibration analysis has long been the method of choice for bearings and other rotating equipment. More commonly, ultrasound is used in conjunction with vibration analysis to help technicians confirm the condition of mechanical assets.

Because of the versatility of ultrasound, if a facility does not have a robust vibration analysis program in place, ultrasound can be implemented to detect earlystage bearing failures, as well as other issues. If the vibration analysis is performed by an outside service provider quarterly or monthly, ultrasound can be used during the interim. This will help the facility to know the condition of some of the more critical assets prior to the service provider entering the facility; therefore, the service provider's time can be used more efficiently because the plant knows if there are any prominent problems with the assets that are being monitored by ultrasound. The service provider can then prioritise its activity based on the ultrasound findings.

Another scenario where ultrasound may be used first over vibration analysis is monitoring slow-speed bearings. Slow-speed bearing monitoring with ultrasound is easier than you might think. Because most high-end ultrasound instruments have a wide sensitivity range and frequency tuning, it is possible to listen to the bearing's acoustic quality, especially at slower speeds. In extremely slowspeed bearing applications (usually less than 25rpm), the bearing will produce little to no ultrasonic noise.

In that case, it is important to not only listen to the sound of the bearing but, more importantly, to analyse the recorded ultrasound sound file in a spectrum analysis software, focusing on the time waveform to see if any anomalies are present. If 'crackling' or 'popping' sounds are present, then there is some indication of a deformity occurring. In bearing speeds above 25rpm, it is possible to set a baseline decibel level and trend the associated decibel level readings over time.

USING ULTRASOUND TO IDENTIFY OVEN MOTOR BEARING FAILURE

An inspection with an ultrasound instrument was carried out on a site with a newly installed oven dryer. This large drum oven was about 20m long by 5m wide. Four large motors rotated it, each of them having two large sets of bearings. These motors rotate the oven and are rotating at a speed of around 7-10rpm. This is an example of extremely slowspeed bearings, which are usually challenging to inspect.

An ultrasonic instrument was used to inspect all bearings – almost all of them presented a smooth sound and a 0dB reading, except for one. On one of the bearings from this set, the ultrasonic instrument displayed 2dB instead of 0. Also, the sound heard from the headphones was different: it was not as smooth as in the other bearings and presented a repetitive 'knocking' sound. This gave the inspector an indication that something might have been wrong with this specific bearing.

After the results from the ultrasonic inspection, a grease sample was taken to confirm if there was any damage to the bearing – in which case, the grease sample would show metal contamination. The grease analysis results showed the presence of metal particles, confirming the damage indicated by the ultrasound instrument.

The next step was scheduling an outage to replace the bearing, which was in poor condition. Part of the outer race came away as it was opened. It was also noticeable that one of the rollers had moved 90°. The cage had been damaged as well.

THE METHOD

Ultrasound technology is useful when monitoring the condition of slow-speed bearings. An ultrasonic instrument/sensor can warn maintenance personnel of failure, even at extremely slow speeds, as in this case.

Bearings rotating at normal speeds can be inspected using ultrasonic technology by comparing changes in dB values. This establishes whether a bearing with a certain value above a decibel baseline needs lubrication or is already in a failure state, depending on how many decibels it is above the baseline.

However, with slow speed bearings, comparing dB levels and establishing alarms is not enough: in many situations, the difference in the dB levels will not be significant or may even be non-existent, in which case the inspector might think there is nothing wrong with it.

For slow-speed bearings, one must rely on the sound quality and the sound pattern. For this, it is necessary to use an ultrasonic instrument with sound recording capabilities, like UE Systems' Ultraprobe 15000, and then analyse the sound file on sound spectrum analysis software such as Spectralyzer. With this approach, maintenance professionals can simply record the sound produced by a slow-speed bearing, load the file into the software and analyse it in the Time Series view.

The spectrum analysis of this oven motor bearing showed clearly where the roller at 90° hits the crack as the knock stops briefly. Thus, the sound pattern indicated an existing problem, being the most reliable source of information when determining the condition of a slow-speed bearing using ultrasound.

On the other hand, the spectrum of a recorded sound from one of the good bearings showed a very different picture: a very uniform spectrum with almost no changes in the amplitude.

This find has saved a significant amount of money, as it was necessary to get cranes in to replace such a big bearing, a job that took up to six hours. Luckily, this was done during a planned outage, avoiding the costs of unplanned downtime.