Inspection Checklist for Roller Bearing Condition Monitoring

Why condition monitoring matters for roller bearings

Roller bearings fail in predictable ways, but the warning signs often appear well before total failure. A rolling element bearing can run quietly for months, then develop a slight hum, followed by knocking, heat, and eventual separation. Condition monitoring is the practice of checking these signs before they turn into a stoppage.

For maintenance teams, the value is simple. Catching a defect early lets you plan a replacement during a planned shutdown. Ignoring it forces an emergency swap, often with more damage to the housing, shaft, and adjacent components.

This checklist is built for field use. It is not a lab test procedure. It is a practical list of checks you can perform during routine rounds, pre-start inspections, or when a machine sounds “off.”

Section 1: Visual inspection and lubrication condition

Start with what you can see and touch. Many early faults leave visible marks before they show up in sensor data.

1. Check grease color and consistency. Healthy grease usually looks clean, slightly stiff, and evenly distributed. Discoloration to brown, black, or bright red can indicate heat damage or contamination.
2. Look for metal particles. Fine, shiny flakes in the grease or on the housing surface suggest rolling element wear. Coarse chips indicate spalling or impact damage.
3. Inspect grease quantity. Too little grease leads to overheating. Too much grease can cause churning, overheating, and seal failure.
4. Check for leaks. Grease weeping out of the seals is normal in small amounts. A steady stream, dried grease crusts, or a wet housing area means the seal is compromised.
5. Look for housing corrosion or paint damage. Rust on the housing often starts at the seal interface. This can trap moisture and accelerate bearing corrosion.

Red flags:

• Grease that is burnt, smelly, or has a gritty texture.
• Metal shavings visible without a loupe.
• Grease packed too high around the shaft or seal.
• Housing paint blistering near the bearing seat.

Section 2: Noise and sound checks

Ears are the oldest condition monitoring tool, and they still work. Roller bearings make distinct sounds when they are unhealthy.

1. Listen for a hum or growl. A low, steady hum can indicate misalignment or early surface defects.
2. Listen for knocking. Intermittent knocking is a classic sign of spalling or pitting on the race or roller. The knock often gets louder as speed increases.
3. Listen for grinding. A gritty grinding noise suggests contamination, such as dirt or metal particles in the lubricant.
4. Listen for whistling or squealing. This can point to insufficient lubrication or a damaged seal allowing air in.
5. Compare against a known good bearing. If you have a spare assembly or a healthy unit on the same line, listen to both side by side. Your ear calibrates quickly.

Red flags:

• A new sound that was not there during the last inspection.
• Noise that changes with load but not with speed.
• A knocking pattern that matches the rolling element count, if you can identify it.
• Noise that gets louder when the bearing is cold and fades when warm, or vice versa.

Tip: Use a mechanic stethoscope or a solid metal rod held to your ear. This amplifies bearing sounds and blocks ambient machine noise.

Section 3: Vibration analysis

Vibration is the most data-rich part of condition monitoring. A handheld vibration meter gives you a number you can trend over time.

1. Measure baseline vibration. Record RPM, speed, and ambient temperature. A healthy bearing has a low, stable vibration level.
2. Check vibration at multiple positions. Measure on the bearing housing, not the shaft, to isolate bearing noise. Use the X, Y, and Z axes if your meter supports it.
3. Look for trend changes, not absolute values. A bearing might run at 3.5 mm/s RMS today and 4.0 mm/s next month. That small increase can be the first sign of a defect.
4. Compare against similar units. If one bearing on a parallel shaft reads consistently higher, it is likely the outlier.
5. Record the speed at which vibration peaks. Some faults show a specific frequency related to the rolling element count or defect location.

Red flags:

• Vibration that rises with speed in a non-linear way.
• A sudden step increase from the baseline.
• High vibration in one direction but not others, suggesting a localized defect.
• Vibration that correlates with temperature rise during the same inspection.

Section 4: Temperature checks

Heat is a direct result of friction, insufficient lubrication, misalignment, or overloading. Temperature is easy to check and hard to ignore.

1. Measure bearing housing temperature. Use a contact thermometer or infrared gun. Compare the reading to the same bearing on a healthy machine.
2. Check temperature rise during operation. A bearing that runs 5 degrees cooler when idle but 15 degrees hotter after an hour may be developing a fault.
3. Look for temperature gradients. One side of the housing being warmer than the other can indicate uneven load or misalignment.
4. Check for hot spots. A single warm area on the housing, even if the overall temperature is normal, warrants a closer look.
5. Correlate temperature with lubrication. A hot housing with dry or burnt grease points to a lubrication failure. A hot housing with fresh grease points to mechanical issues.

Red flags:

• Housing temperature exceeding the lubricant manufacturer limit.
• A temperature rise of more than 10 degrees Celsius during a short test run.
• A hot spot that moves or changes location over time.
• A bearing that is hotter than its neighbor under identical load.

Section 5: Mechanical fit and alignment

A roller bearing can be healthy internally but fail quickly if the housing or shaft is misaligned. Condition monitoring includes checking the mechanical environment around the bearing.

1. Check shaft end play. Excessive play can cause impact loading and spalling.
2. Check housing concentricity. If the bearing seat is out of round or out of square, the bearing will not seat evenly.
3. Check for shaft deflection. Under load, a long or thin shaft can bend, changing the bearing contact pattern.
4. Inspect mounting surfaces. Chips, cracks, or corrosion in the housing bore change the fit and can damage the bearing race.
5. Check for loose fasteners. A loose bearing cap or housing bolt allows movement that creates impact stress.

Red flags:

• Visible gap between the bearing and housing.
• Scratches or scoring on the shaft where the bearing sits.
• Housing bolt heads that are loose or cross-threaded.
• Shaft deflection visible under load, especially on long cantilevered supports.

Section 6: Lubrication system health

The bearing is only as good as its lubrication. A failed lubrication system is a common cause of premature roller bearing failure.

1. Check grease pump or lube line. Look for blockages, leaks, or air in the line.
2. Verify grease type. If the machine uses a specific grease, confirm it is still in the system. Mixing incompatible greases causes separation and overheating.
3. Check lube interval logs. Grease has a service life. Even if it looks clean, old grease breaks down and loses its protective properties.
4. Inspect seals and seals. A failed seal allows contamination in and grease out. Both conditions destroy the bearing.
5. Check for water ingress. Water in the grease causes rust and emulsification. Look for white, milky grease or rust on the shaft.

Red flags:

• Grease that separates into oil and solid particles.
• Water droplets or milky appearance in the grease.
• A lube line that runs dry or cycles irregularly.
• Grease that has changed color from its original state.

Quick reference table: common red flags and likely causes

Red flag	Likely cause	First action
Intermittent knocking at speed	Spalling or pitting on race or roller	Increase inspection frequency, check vibration trend
Burnt, dark grease	Overheating, low lube, or contamination	Check lube system, measure housing temperature
Metal shavings in grease	Rolling element or race damage	Inspect for further wear, plan replacement
High vibration at one bearing only	Misalignment, loose fit, or early defect	Check fit, alignment, and compare with neighbor
Housing hotter than neighbor	Uneven load, misalignment, or lube issue	Measure temperature gradient, check load distribution
Grease leaking out of seals	Seal failure or over-packing	Replace seal, re-lube to correct quantity

Building a simple condition monitoring routine

You do not need a full sensor network to benefit from this checklist. A basic routine works like this:

1. During a daily or weekly walk, listen to each bearing with your ear or a stethoscope.
2. Check visible grease condition and look for leaks or metal particles.
3. Use a contact thermometer on the housing. Record the number.
4. Use a handheld vibration meter if available. Record the RMS value.
5. Log the findings in a simple table or maintenance software.
6. Flag any item that differs from the baseline by a set threshold, such as 10 percent in vibration or 5 degrees in temperature.
7. If a flag appears, re-inspect within 24 to 48 hours. If the trend continues, plan a replacement.

The goal is trend monitoring, not a single perfect measurement. A bearing that is getting worse slowly is easier to manage than one that fails suddenly.

Common mistakes to avoid

• Relying on one check. Noise is normal, but vibration is normal, and temperature is normal. None of them tell the whole story alone.
• Ignoring baseline. Without a baseline, you cannot know what “normal” is for that specific machine.
• Waiting for failure. The whole point of condition monitoring is to catch the problem before it stops the line.
• Over-tightening fasteners. A tight fit can distort the housing and change the bearing preload.
• Mixing greases. Different grease chemistries can separate and lose their protective effect.
• Not documenting. If you do not record the findings, you lose the trend data and the ability to prove a trend to your maintenance manager.

This checklist is a starting point. Adapt it to your machines, your environment, and your risk tolerance. The more consistent you are, the more useful the data becomes.