Bearing diagnostics in machinery and production lines

A bearing is a small component, but if it fails, it can bring the entire machine or a section of the production line to a standstill. That is why bearing diagnostics should not only begin when a distinct noise or a rise in temperature becomes apparent. The point of it is to detect changes at an early stage, assess their significance, and distinguish between the symptom and the actual cause of the problem.

In a production environment, bearing failure rarely occurs without warning. Usually, there is a prior increase in vibration, a change in the nature of the signal, a deterioration in lubrication conditions, or the machine begins to run less smoothly. However, simply detecting an anomaly is not enough. Similar symptoms can stem from various causes: overload, contamination, incorrect assembly, poor fit or inadequate lubrication. A good diagnosis must therefore answer two questions simultaneously: is the condition of the bearing deteriorating, and why is this happening?

Why is bearing diagnostics so important for maintaining operational continuity?

Bearing failure very often does not stop at the bearing itself. If the problem is overlooked, it can result in damage to the shaft, housing, seals, coupling or other drive components. On a production line, this means not only the cost of replacement parts, but also downtime, loss of productivity and the risk of secondary damage throughout the entire system.

This is precisely what distinguishes a reactive approach from condition-based diagnostics. In the first scenario, a bearing is replaced after a failure or just before it occurs. In the second, trends are monitored and action is taken in advance, before the damage spreads to other components. This does not mean that every machine requires extensive online monitoring. It simply means that the method of bearing monitoring should be determined by the criticality of the equipment and the cost of its downtime.

What are the most common symptoms?

The most characteristic signs of a bearing’s deteriorating condition are increased vibration, a rise in temperature, a change in the nature of the noise, a deterioration in the condition of the lubricant, and signs of unstable operation of the shaft or bearing assembly. However, one must be careful, as the order in which these signs appear is not always the same. In some cases, changes in the vibration signal are the first to appear; in others, temperature is the first indicator; and sometimes the problem is first evident from the condition of the lubricant.

This is precisely why bearing diagnostics should not rely on a single parameter. If the assessment is based solely on noise or temperature alone, it is easy to miss the early stages of damage or misinterpret its nature. A multi-parameter approach provides a much better picture: the vibration signal shows the development of a defect, temperature may indicate a problem with friction or lubrication, and the condition of the lubricant helps to assess whether the cause lies outside the bearing itself.

Table of symptoms

SymptomWhat could this mean?How to spot it, as usualWhat to check next
increase in vibrationdamage to the raceway, rolling elements, play, misalignmentvibration analysis, spectrum, envelopefrequency of faults, installation condition, load
elevated temperatureinadequate lubrication, excessive friction, overloadingtemperature sensor, thermography, trendquantity and condition of lubricant, fit, resistance to movement
an unusual noisecontamination, wear and tear, surface damagelistening, ultrasound, sound analysiscondition of the running surface and rolling elements; cleanliness of the system
contaminated greaseparticle ingress, lubricant degradationlubricant monitoring, particle analysisseals, working environment, lubrication schedule
uneven shaft operationexcessive play, damage to the housing, installation errorsvibrations, operational monitoring, alignment checksfit, alignment, condition of the mounting

This table illustrates one simple point: the same symptom can have several different causes. A rise in temperature alone does not necessarily mean the treadmill is faulty, and an increase in vibration does not automatically indicate whether the problem lies with the bearings, lubrication, alignment or load. Diagnostics begins with the detection of an anomaly, but does not end with an alarm.

It is only by comparing the symptom with the operating conditions that a clear picture emerges. If the temperature rises and lubrication issues are observed at the same time, the approach to identifying the cause will differ from that taken when the vibration envelope increases whilst the temperature remains stable and the lubricant is in good condition.

Vibration as a fundamental method of bearing diagnosis

Vibration analysis remains the most important method for detecting bearing damage whilst the bearing is in operation. This is due to a simple fact: damage to the raceways, rolling elements and cage generates characteristic changes in the signal even before the problem becomes clearly evident in terms of temperature or noise audible to the operator. A proper assessment of a bearing therefore requires looking not only at the overall vibration level, but also at the spectrum, the time signal and the characteristic frequencies of the defects.

This is important because a simple overall indicator merely suggests that something is changing. It does not yet indicate exactly what is happening to the bearing. Only an analysis of the envelope, characteristic frequencies and changes over time allows us to distinguish an early raceway defect from other sources of vibration, such as imbalance, misalignment or mechanical problems outside the bearing itself.

Temperature, lubrication and contamination

Measuring temperature is simpler than advanced vibration analysis, but it still has significant diagnostic value. It is particularly effective for problems related to lubrication. A rise in temperature is often one of the first signs of excessive friction, an insufficient lubricating film or excessive operating resistance. This does not mean that every higher temperature is evidence of bearing damage, but it is a sign that should not be ignored.

The condition of the lubricant is equally important. Contamination, degradation, incorrect viscosity, insufficient lubricant, or an inappropriate relubrication schedule can be both a cause of damage and an indicator of it. Some faults begin precisely with friction and the lubricating film, and only later progress to damage to the raceway and rolling elements. Therefore, bearing diagnostics without an assessment of lubrication is simply incomplete.

Diagnosing symptoms versus diagnosing causes

This distinction is crucial. It is possible to detect very early on that something is wrong with a bearing, yet misdiagnose the cause. In such cases, replacing the bearing only solves the problem temporarily. To make a correct diagnosis, data on load, speed, temperature, lubrication method, sealing, fits and operating history are required. Without this, one assesses the effect rather than the source of the problem.

The most common causes tend to be quite similar: inadequate lubrication, contamination, overloading, incorrect installation, a fit that is too tight or too loose, electrical currents, misalignment and classic material fatigue. These are not competing hypotheses, but phenomena that often occur in combination. A bearing may operate under poor lubrication conditions, causing the temperature to rise and accelerating wear, whilst the entire system may also be overloaded or geometrically misaligned.

SymptomPossible causeWhat can confirm the diagnosisCorrective action
rising vibration envelopedamage to the raceway or rolling elementsfailure frequency analysisbearing replacement and investigation of the root cause
a higher temperature than usualinsufficient lubrication, poor-quality lubricant, excessive frictionchecking the lubricant, load and playlubrication adjustment, assembly check
an irregular metallic noisecontamination, surface damagechecking the lubricant, dismantling after a standstillimproving the sealing and cleanliness of the system
frequent re-damagean installation error or an unresolved issueanalysis of repair history and working conditionschange to the assembly procedure, adjustment of fits
load-dependent signalsoverload or misalignmentanalysis of performance under various load conditionsadjustment of settings and operating parameters

The key point to take away from this comparison is that a symptom is not yet a diagnosis. The same symptom can lead to completely different repair procedures if it is caused by a different underlying issue. Jumping to the conclusion that ‘the bearing is faulty’ all too often results in the part being replaced without addressing the root cause of the problem.

Effective diagnostics therefore have two stages. The first is detecting deterioration. The second is identifying the conditions that caused this deterioration. Without the second stage, maintenance falls into a cycle of repetitive repairs that do not address the problem at its root.

Periodic monitoring or online monitoring

Not every machine requires constant online monitoring. In many facilities, periodic rounds using portable measuring devices work well, particularly for less critical machines or those operating under stable conditions. This approach allows for effective monitoring of equipment condition without the need for a costly expansion of the measurement system across the entire site.

Continuous online monitoring is particularly useful where a machine is critical to production continuity, difficult to access, operates non-stop, or where faults can develop rapidly. Such a system provides trend continuity, faster alerts and a better correlation between bearing condition and process parameters. When applied to the right equipment, it can also pay for itself quickly, as it reduces the risk of sudden downtime.

The most common errors in bearing diagnosis

The first mistake is to assess the condition of a bearing based on a single parameter. A rise in temperature, overall vibration levels or noise alone are not enough to identify the type of damage with any degree of certainty. The second mistake is to ignore the operating conditions. A signal from the machine is meaningless without knowledge of the load, speed, lubrication and system geometry.

The third mistake is confusing the symptom with the cause. Replacing the bearing may be necessary, but if the damage is due to misalignment, incorrect installation or contamination, the new bearing will soon develop the same problem. The fourth mistake is the lack of historical trends. A single measurement provides only a snapshot of the moment. Only a series of data points reveals whether we are dealing with gradual wear, accelerating degradation or a temporary deviation.

What does this mean for maintaining production line operations?

Proper bearing diagnostics is not an end in itself. Its purpose is to enable better maintenance decisions: whether the machine can still be operated safely, when to schedule a shutdown, whether an adjustment to the lubrication is sufficient, whether a replacement is needed, or whether the alignment, fit or load on the system needs to be checked. On a production line, such knowledge is worth far more than the cost of the bearing itself, as it determines the continuity of the process.

In a well-structured maintenance system, bearing diagnostics combines monitoring data with knowledge of the machine. In this context, vibrations, temperature and lubricant condition are not merely sensor readings, but form part of a logical assessment of the machine’s technical condition. This is precisely where predictive maintenance comes into play, as it genuinely reduces unplanned downtime rather than merely detecting it sooner.

Summary

Bearing diagnostics in machinery and production lines is effective when it combines vibration monitoring with temperature monitoring, lubrication assessment, knowledge of operating conditions and root cause analysis. An alarm from the monitoring system is just the start. Value is only realised when the data can be used to make the right decision: when to intervene, what to replace and what to eliminate as the source of the problem.

FAQ – Bearing diagnostics in machinery and production lines

How can you tell if a bearing in a machine is faulty?

Most commonly through increased vibration, a change in noise, elevated temperature or deterioration of the lubricant. A single symptom is usually not enough, so it is best to assess the condition of the bearing based on several signs at the same time.

Does a rise in temperature always indicate a faulty bearing?

No. It may also indicate a lubrication problem, excessive friction, overload or an overly tight fit. Temperature is an important indicator, but it needs to be considered alongside other data.

What types of vibration indicate a bearing problem?

The most valuable information comes from analysing the spectrum, envelopes and characteristic frequencies of defects, rather than the overall vibration level itself. This allows bearing damage to be distinguished from other problems in the rotating system.

Which is more likely to cause bearing failure: lubrication or overload?

Both causes are common, but a very large proportion of problems stem from improper lubrication or contamination of the lubricant. Overload, misalignment and assembly errors are also often part of the same chain of failures.

Is online monitoring worthwhile for every production line?

No. It makes the most sense for critical machines, those that are difficult to access, or those where a failure results in high downtime costs. For less critical equipment, regular periodic measurements are often sufficient.

0 comments
Oldest
Newest
Inline Feedbacks
View all comments