While bearing failure is a common fear in industry, it only happens in a tiny proportion of cases. A staggering 90% of bearings outlast the machine in which they are installed.
Only an estimated 0.5% actually fail in service. The main reasons are fatigue and poor lubrication, though contamination and poor installation can also play a part. The remaining 9.5% of bearings are replaced in service prior to failure, for safety reasons.
When failures do occur, it’s critical to understand what went wrong. If a bearing continues to fail in service there is likely to be an underlying cause. Careful analysis of the damaged bearing can reveal crucial details about the underlying cause of failure, in order to correct it in future.
Using the methodology detailed in the international standard ISO 15243:2004 – which categorises six main damage/failure modes, broken into 14 sub-categories – it is possible to pinpoint the cause of failure, and change machine conditions accordingly in order to overcome the problem.
Bearing inspection standard
The standard relies on a visible inspection of rolling element contact and other functional surfaces – such as raceways – which will suggest the mechanisms involved in each type of damage or failure. The main causes of bearing damage can be linked to six main damage/failure modes:
- electrical erosion;
- plastic deformation; and,
Further investigation will help to determine which sub-mode was responsible for the failure, which will ultimately reveal the likely cause – and its solution.
While damage can be divided into six modes, the ways in which this comes about can be varied.
For instance, ineffective seals allow contamination to enter the bearing. These contaminant particles can then be over-rolled by the rolling elements, creating indentations in the raceways. Hard particles can cause indentations with sharp edges.
Once the area around the indentation is subjected to cyclic stress from the rolling elements, surface fatigue can set in – causing metal to break away from the raceway, in a process called spalling. Once spalling has begun, the damage will worsen until the bearing eventually fails.
Fatigue failure might be seen in the form of sub-surface micro-cracking – which is usually as a result of long-term abnormal loading – or as spalling of raceway surfaces, which can indicate lubrication problems. Here, fatigue is the prime failure mode, subdivided into sub-surface or surface manifestations of fatigue – either of which will identify the underlying cause of failure.
Damage to cages and raceways is usually caused by contaminants that have entered the bearing, either by way of the lubricant or as a result of poor or inadequate sealing against adverse operating environments.
Inadequate lubrication can also cause ‘adhesive’ wear: material transfers from one mating surface to another as a result of heightened friction between them. This is sometimes accompanied by tempering or re-hardening of the surfaces, producing localised stress concentrations and potential spalling of the contact areas.
Bearing damage can be light or severe, and close inspection can identify the cause of the problem. The key is to look for ‘patterns’ – especially path patterns – on the raceway. Some patterns are normal, but others can be indicative of a problem.
Different types of bearings will exhibit different ‘problem’ patterns. However, a knowledgeable engineer can identify a failure mode by inspecting the surface damage. For instance, in a radial bearing working under normal conditions, the wear pattern will be restricted to the centre of the raceways and will be evenly distributed. If an axial load is applied to the bearing, the wear pattern becomes eccentric, or off-centre. As problems become more complex, so do the surface wear patterns.
Finding Probable bearing failure cause
In a failure mode, a variety of causes can lead to the same effect. However, identifying the most likely causes – or combination of causes – is the aim of an investigative process called root cause analysis (RCA).
RCA works on the theory that every failure stems from at least one of three basic causes: physical or technical; organisational failures related to systems, procedures and decision-making processes; and human errors of omission or commission.
After a bearing has failed, RCA must first identify one or more of the three basic causes that initiated the problem. The root cause is then defined by matching the failure mode and its sub-modes with the operational conditions from the time of failure – based on detailed visual inspections of the bearing surfaces and sub-surfaces. The investigating engineer must then develop and apply a solution, and monitor its effectiveness.
Identifying the root cause of the failure is not easy – as it is often hidden among a host of ‘secondary effects’. Examining a bearing that has run for some time will reveal a number of changes: dull or even shiny areas on raceways and rolling elements; discoloured inner and outer ring seats; evidence of cage wear; and fretting corrosion on the inner ring bore or outer ring external surface. By carefully analysing the evidence – in the form of damage ‘patterns’ on the surface – an experienced engineer can track down the root cause.
It is not always a straightforward diagnosis. When searching for an answer, investigating engineers need all the relevant data – which may involve reviewing equipment maintenance history, and the actions taken in response to similar failures in the past.
Understanding the maintenance procedures in place at the time of a failure – and the precise conditions that led to it – is also a crucial part of solving the puzzle.
In one case, a variable speed electrical motor – used in the reel section of a tissue paper machine in a paper mill – was experiencing bearing failure after only 1-2 months. There was severe damage to the cylindrical roller bearing (but none to the ball bearing). The machine was stopped due to high vibration levels.
There was heavy, irregular wear on the raceway of the inner ring. Some flats were observed, and it was dull grey in colour. The outer ring had a similar appearance, and also included marks similar to those caused by vibration.
The initial diagnoses suggested two possible causes: excessive vibration; or current passing through the bearings – though the customer insisted the bearings were insulated.
The bearing was cut open and inspected microscopically, which revealed micro-craters on the raceway surface. These were caused by current leakage. This then led to vibration, which created a tell-tale ‘washboarding’ pattern.
Because current leakage was identified as the underlying problem, the customer inspected the electrical system, which showed that a grounding (earth) cable had been disconnected from the motor during repairs, and not reconnected. Once this was fixed, the problems did not return.
While inspecting the bearing and identifying its failure modes is essential, the ultimate aim of RCA is to prolong service life. In the above example, a fault in the electrical system was corrected. In other cases, the long-term solution may require other actions, such as overhauling maintenance procedures, introducing automated lubrication or installing a condition monitoring system.
In fact, successful diagnosis of the problem is really only the beginning of the solution. Once the underlying cause has been identified, it is vital to introduce some kind of monitoring system in order to ensure that problems are not repeated – because replacing a bearing ahead of time is preferable to replacing it after a shutdown.