How to Check Bearing Temperature for Early Fault Detection

Why Temperature Reveals Bearing Faults Early

Bearing temperature rises when friction, misalignment, or lubrication problems increase energy loss inside the housing. A small increase above normal operating levels often appears hours or days before visible damage. Maintenance teams can use this signal to schedule work while the part still has service life.

The heat generated inside a bearing is a direct result of work being done. As the inner and outer races rotate, the lubricant must separate the surfaces and reduce contact friction. When the lubricant film breaks down, or when metal surfaces touch, the mechanical energy that would otherwise be dissipated as heat is concentrated in the rolling elements and the raceways. This conversion of mechanical energy into thermal energy is what a temperature sensor detects.

Early warning is possible because thermal mass is slow to respond but steady in its behavior. A bearing does not heat up instantly when a defect appears. The defect must generate enough excess energy to raise the temperature of the entire housing assembly. This lag is useful. It gives technicians time to observe a shift in the baseline before catastrophic failure occurs. If a bearing is failing due to spalling, the temperature may creep upward over several days. If it is failing due to a sudden lubrication leak, the temperature may spike within hours. Understanding this difference in response time allows for targeted intervention rather than emergency shutdowns.

Temperature is also a leading indicator of lubrication degradation. Over time, lubricants break down due to heat and oxidation. The viscosity of the oil decreases, and the additive packages that protect the metal surfaces become less effective. As the lubricant thins, the film thickness reduces, increasing friction. This creates a feedback loop where higher friction generates more heat, which further degrades the lubricant. Monitoring temperature allows teams to catch this cycle before the lubricant reaches a critical state where metal-to-metal contact becomes likely.

Prerequisites Before Starting Temperature Checks

Before collecting data, confirm the equipment is running at its typical load and speed. Record ambient temperature, bearing type, and lubricant condition in the maintenance log. Ensure the surface is clean. Oil, dust, or paint can block infrared readings and distort contact sensor results. Verify that the sensor is the right size for the bearing housing and that it is rated for the expected temperature.

A temperature reading is only meaningful when it is taken under defined conditions. If a motor is running at 80% load, the bearing temperature will be different than if it is running at 20% load. Establishing these parameters is the first step in any reliable monitoring program. The equipment must be in a steady state for the measurement to be valid. Transient states, such as during ramp-up or ramp-down, involve dynamic friction and thermal gradients that do not represent the normal operating condition.

Cleaning the measurement surface is a simple but frequently skipped step. Infrared sensors rely on the emissivity of the surface to reflect thermal radiation. Paint, grease, and dust alter the emissivity, causing the sensor to read incorrectly. A shiny metal surface might reflect the heat from nearby components, giving a reading that is higher than the actual bearing temperature. A dark, oily surface might absorb heat and reflect less, giving a lower reading. Wiping the surface with a lint-free cloth and a suitable solvent ensures that the sensor is reading the metal housing itself, not the debris on top of it.

Sensor selection matters just as much as placement. Contact thermocouples must be firmly attached to the housing. If the thermocouple bead is not in intimate contact with the metal, it measures the air temperature rather than the housing temperature. Infrared sensors require a clear line of sight and a distance within the manufacturer’s specified range. The sensor housing must be rated for the ambient conditions of the plant. If the plant floor is hot and humid, a standard sensor may drift or fail prematurely. Checking the sensor specifications before installation prevents false data that leads to unnecessary maintenance or missed failures.

Step-by-Step Bearing Temperature Monitoring

1. Establish the baseline temperature. Run the machine at normal load for a set time, such as one hour. Record the temperature at the bearing housing. This value is your reference point. Without a baseline, a reading of 60 degrees Celsius means nothing. You need to know whether 60 is normal or abnormal for that specific unit.

The baseline is the most critical piece of data in temperature monitoring. It represents the thermal equilibrium of the bearing under specific operating conditions. To establish it, the equipment must run continuously at the same speed and load for a sufficient duration. One hour is a common minimum, but for large motors with high thermal mass, it may take longer for the temperature to stabilize. Record the ambient temperature at the same time. The baseline is not a single number but a pair of values: the bearing temperature and the ambient temperature. This pairing allows you to calculate the temperature rise, which is the most accurate indicator of friction and energy loss.

2. Select the right sensor placement. For contact thermocouples, attach the sensor to the bearing housing near the outer race. For infrared thermography, aim at the outer race area, not the seal or the shaft. The outer race is usually the hottest point during normal operation. Avoid measuring the seal, which often runs cooler and gives misleading data.

Placement determines the accuracy of the reading. The outer race is the preferred location because it experiences the highest friction loads and has the most direct thermal connection to the rolling elements. The inner race moves with the shaft, and its temperature is influenced by shaft heat, which can be misleading. The seal area is often cooler because it is exposed to ambient air and has a different thermal profile. If you measure the seal, you may miss a bearing fault that is already developing in the race. For contact sensors, ensure the thermocouple is placed on the housing metal, not on a painted area or a flange that is thermally isolated from the bearing.

3. Use multiple measurement points if possible. Measure both the driving side and the non-driving side of the bearing. In some cases, measure the housing top and bottom. A temperature difference between sides can indicate uneven load or misalignment. A difference between top and bottom may suggest oil starvation or cooling issues.

Multi-point monitoring provides a diagnostic picture that a single sensor cannot. If the driving side is significantly hotter than the non-driving side, it may indicate that the bearing is taking an uneven load. This can happen if the shaft is misaligned or if the coupling is out of adjustment. If the top of the housing is hotter than the bottom, it might be due to radiation heat from above or, more critically, oil starvation where the oil film is not circulating properly to the upper race. Comparing these readings helps isolate the location of the problem and guides the next step in troubleshooting.

4. Record readings at regular intervals. Check the temperature during startup, steady state, and shutdown. Startup temperatures are often the highest due to initial friction. Steady state readings show the normal operating range. Shutdown readings help identify heat retention and cooling issues. Log each reading with the date, time, ambient temperature, and load condition.

Capturing data across the operating cycle is essential for understanding thermal behavior. Startup friction is higher because the lubricant is not yet distributed evenly, and the metal surfaces are colder. This causes a temporary spike in temperature that should drop as the machine reaches steady state. If the startup spike is unusually high or does not drop as expected, it may indicate a seal issue or a lubricant problem. Shutdown readings are often overlooked but are valuable. If the temperature remains high after the machine stops, it suggests poor heat dissipation or a blocked cooling fan. This heat retention can stress the bearing even when the machine is idle, accelerating wear.

5. Compare readings against alarm thresholds. Set warning thresholds above the baseline. A common rule is to trigger a warning when the temperature exceeds the baseline by a set margin or when it reaches a maximum safe limit for the bearing type. The exact margin depends on the bearing material and lubricant. Consult the bearing manufacturer for maximum operating temperature limits.

Thresholds must be set based on the specific bearing and lubricant combination. Different bearing materials, such as steel or stainless steel, have different thermal properties. Lubricants also have maximum operating temperatures above which they begin to break down rapidly. Setting a threshold too low leads to frequent false alarms and alarm fatigue. Setting it too high risks missing early warning signs. A practical approach is to set a warning level at a certain degree above the baseline and a critical level at the manufacturer’s maximum limit. This two-tier system allows for early intervention before the bearing reaches a dangerous state.

6. Investigate any temperature increase over the baseline. If a reading is higher than expected, check for common causes. Look for low lubricant level, wrong lubricant viscosity, misalignment, excessive load, or contamination. Use vibration analysis to check for spalling or damage. Inspect the seal for leaks. Do not assume the bearing is faulty until other causes are ruled out.

A temperature rise is a symptom, not a diagnosis. The most common causes of increased bearing temperature are external to the bearing itself. Low lubricant level is a frequent culprit, especially in machines where the lubrication system is manually filled. If the oil level is below the minimum mark, the oil film cannot support the load, leading to metal-to-metal contact and heat. Incorrect lubricant viscosity can also cause problems. If the oil is too thin, it breaks down easily. If it is too thick, it does not flow properly at low temperatures or high loads. Misalignment forces the bearing to operate at an angle, creating uneven contact pressures. Contamination, such as dust or metal particles, acts as an abrasive, increasing friction. Always check these external factors before considering the bearing itself as the source of the fault.

7. Track trends over time. A single high reading may be a false alarm. A slow upward trend over weeks is a strong warning sign. Plot the temperature data on a chart. Look for gradual increases that suggest lubricant breakdown or wear. A sudden spike usually points to a sudden event, such as a seal failure or a shock load.

Trend analysis is the key to distinguishing between normal variation and developing failure. A single high reading could be due to a temporary load change or a measurement error. However, a slow upward trend over weeks is a clear indicator of progressive degradation. Plotting the data on a chart makes this trend visible. If the temperature is rising gradually, it suggests that the lubricant is breaking down or that wear is accumulating. If there is a sudden spike, it indicates a discrete event, such as a seal failure that allowed lubricant to leak out or a shock load that damaged the bearing. Understanding the shape of the trend helps determine the appropriate response.

8. Take corrective action based on the trend. If the temperature is rising slowly, plan lubrication service or alignment checks. If there is a sudden spike, stop the equipment and inspect the bearing. If the trend stabilizes after maintenance, continue monitoring. If the temperature continues to rise, replace the bearing before it fails.

Corrective action should match the severity of the temperature trend. A slow rise warrants scheduled maintenance, such as lubrication service or alignment adjustment. This allows the team to address the root cause without disrupting production. A sudden spike requires immediate action. The equipment should be stopped, and the bearing should be inspected for damage. If the temperature stabilizes after maintenance, the issue is resolved, and monitoring can continue. If the temperature continues to rise despite corrective action, the bearing is likely damaged and should be replaced. Acting based on the trend prevents both unnecessary replacement and catastrophic failure.

Common Mistakes in Bearing Temperature Monitoring

• Measuring the wrong surface. The seal and the shaft often run cooler than the outer race. Measuring the wrong spot gives false confidence. Always aim for the outer race area.
• Ignoring ambient temperature. A bearing running at 55 degrees Celsius in a cold plant is normal. The same temperature in a hot plant may indicate a problem. Always record ambient conditions.
• Using one reading to make a decision. A single high reading may be caused by a temporary load spike. Look at the trend over time.
• Forgetting to clean the surface. Dust and oil block infrared readings. Wipe the surface before measuring.
• Not updating the baseline. If the load changes or the lubricant is replaced, the baseline changes. Update the reference value after any major maintenance.
• Relying only on temperature. Temperature is one signal. Combine it with vibration and acoustic analysis for a fuller picture.

These mistakes are common because temperature monitoring is often treated as a simple task. Measuring the wrong surface is a frequent error, particularly when using infrared thermography. The seal and shaft are often more accessible, leading technicians to measure them instead of the outer race. This results in readings that do not reflect the actual bearing condition. Ignoring ambient temperature is another critical error. A bearing at 55 degrees Celsius in a cold plant is operating normally, but the same temperature in a hot plant may indicate a lubrication problem. Always record ambient conditions to ensure accurate comparisons. Using a single reading to make a decision is dangerous. A temporary load spike can cause a high reading that does not reflect a bearing fault. Looking at the trend over time provides a more reliable picture. Forgetting to clean the surface is a simple but costly mistake. Dust and oil distort infrared readings, leading to false data. Not updating the baseline after maintenance is a logical error. If the load changes or the lubricant is replaced, the baseline changes. Failing to update it leads to incorrect alarm thresholds. Finally, relying only on temperature is a limitation. Temperature is a useful signal, but it is not a complete diagnosis. Combining it with vibration and acoustic analysis provides a more comprehensive view of bearing health.

Table of Temperature Ranges and Actions

Condition	Typical Behavior	Recommended Action
Normal operation	Stable within baseline range	Continue routine monitoring
Slight increase	Slow rise over days or weeks	Check lubricant level and viscosity
Moderate increase	Rise above warning threshold	Inspect for misalignment or contamination
High increase	Reading near or above maximum limit	Stop equipment and inspect bearing
Sudden spike	Sharp jump in minutes or hours	Immediate shutdown and detailed inspection
Cooling issue	Temperature stays high during shutdown	Check cooling system and ventilation

Final Verification Step

After any maintenance action, run the equipment and take new temperature readings. Compare them to the original baseline. If the temperature has returned to normal and the trend is flat, the issue is resolved. If the temperature is still elevated, continue monitoring and investigate further. Document the action taken and the result in the maintenance log. This creates a record that helps future troubleshooting and supports continuous improvement.

Verification is the final step in the temperature monitoring process. After performing corrective action, such as lubrication service or alignment adjustment, the equipment must be run again to confirm that the temperature has returned to normal. This step is critical because some issues may not be fully resolved by a single maintenance action. If the temperature is still elevated, the root cause may be deeper, such as bearing damage or a persistent contamination issue. In this case, continue monitoring and investigate further. Documenting the action taken and the result in the maintenance log is essential. This record helps future troubleshooting by providing a clear history of the problem and the actions taken. It also supports continuous improvement by identifying patterns in failures and maintenance effectiveness. A well-documented history allows maintenance teams to refine their monitoring thresholds and predictive maintenance strategies over time.

When to Replace the Bearing

Temperature is a warning, not a diagnosis. A bearing that shows a rising temperature trend may still have months of life left if the cause is lubrication or alignment. Replace the bearing when the temperature remains high after correcting the cause, or when vibration analysis confirms damage such as spalling or brinelling. Do not replace a bearing based on temperature alone. Use temperature as one part of a broader condition assessment.

The decision to replace a bearing should be based on a combination of factors, not temperature alone. A rising temperature trend may be caused by external factors, such as lubrication or alignment, which can be corrected without replacing the bearing. In such cases, the bearing may still have significant life remaining. However, if the temperature remains high after correcting the cause, it indicates that the bearing is damaged and should be replaced. Vibration analysis is a key tool for confirming bearing damage. It can detect defects such as spalling, where the raceway is pitted, or brinelling, where the raceway is indented by heavy loads. These defects cause increased friction and heat, leading to a rising temperature trend. By combining temperature data with vibration analysis, maintenance teams can make a more accurate decision about when to replace the bearing. This approach prevents unnecessary replacement and avoids catastrophic failure.

How Temperature Data Fits Into a Broader Maintenance Program

Bearing temperature monitoring works best as part of a condition-based maintenance program. Combine it with vibration analysis, oil analysis, and visual inspection. Each method highlights different failure modes. Vibration shows mechanical damage. Oil analysis shows contamination and lubricant condition. Temperature shows energy loss and heat buildup. Together, they give a complete picture of bearing health and help maintenance teams act before failure occurs.

Temperature monitoring is a powerful tool, but it is most effective when used as part of a broader maintenance program. Condition-based maintenance relies on multiple data sources to assess the health of critical components. Vibration analysis detects mechanical damage, such as spalling and misalignment. Oil analysis identifies contamination and lubricant degradation. Temperature monitoring reveals energy loss and heat buildup. Each method provides a different perspective on the bearing’s condition. By combining these data sources, maintenance teams can create a complete picture of bearing health and act before failure occurs. This integrated approach reduces unplanned downtime and extends the life of critical equipment. It also allows for more precise maintenance planning, where actions are taken based on actual condition rather than fixed time intervals.