• Article date:

Detecting failure in railway bearings


Every engineer knows the importance of detecting bearing failure before it happens.  It’s important in every industry, and it’s absolutely critical in railways since the potential damage of a derailed train could be devastating.  There are a few ways bearing failure can be detected in railways and each has there advantages and disadvantages.

Hot box detector  

Hot box detectors are widely used in the railway industry to detect bearing failure.  A hot box is place by the side of the rails and uses infrared technology to detect heat emitted from defective bearings.  Since a consistent symptom of a failed bearing is overheating, this is considered a reasonably reliable method of condition monitoring.  However, hot boxes can only detect failure when it is in the line of sight, so inward bearings are not monitored and when a bearing reaches a level where it begins to overheat it is well on the way to failure.  A system which can detect failure much earlier will allow more efficient and safer maintenance. 


Accelerometers are an electromechanical device which measure acceleration forces and have a host of different applications.  When it comes to bearings, accelerometers are used to monitor vibration.  An accelerometer produces an electrical signal that is proportional to the acceleration of the vibrating bearing; excessive vibration signals that a bearing is close to failure.  The accelerometer has to be attached as close as possible to the bearing in order to detect failing.  For rail applications they’re placed on the rails.

Acoustic bearing monitoring

Another method of bearing monitoring which was used in the 1980s and has been recently revived is acoustic monitoring.  A recent innovation in acoustic detectio is said to detect failing sooner than other techniques, thereby improving the overall reliability of rail transport.  RailBAM monitors the sound of wheelset bearings as a train passes a set of sensors.  Using a number of sensors improves spatial discrimination and directionality and software uses geometric wheel measurements and acoustic characteristics to reduce crosstalk when detecting sound so that a fault on a large axle will not disrupt monitoring of a smaller axle.      

These are the most common ways failure is detected in railway bearings though innovations are occurring all the time as rail travel develops and high speed trains become more widespread.