Fatigue Failure in Bolted Joints

By Sam McNair, CMRP, of Life Cycle Engineering
As appeared in Plant Engineering

Fatigue failures can be particularly hazardous because they often occur with no visible warning signs and the failure is often sudden and total. While most maintenance technicians understand how to torque fasteners properly, very few understand the reason why it is so critical to torque properly in applications subject to fatigue. Even fewer understand how a fatigue failure really occurs – especially in what they think is a static joint. We often unknowingly avoid fatigue failures in gasketed joints simply because the required crush for the gasket often dictates a torque or bolt tension that minimizes the risk of a fatigue failure.  But, at a later time, changing to a new gasket type requiring less crush may set the stage for bolt fatigue failure. Maintenance technicians who don’t understand the basic how and why of fatigue can unknowingly set up the conditions for serious failures.  

It is not unusual for technicians to assume that a bolt has failed due to overload when it has in fact failed from fatigue. Where there is no obvious cause for overload they may assume that the last person to install the bolt had over-tightened it.  This can set the stage for a cycle for of ever shortening times to failure if they “fix” the problem by reducing the torque. If they get caught up in a cycle of ever decreasing MTBF they may also fail to replace all of the bolts in a joint where only one fastener has failed.  This adds to the problem.  Some simple rules to help avoid fatigue failures:

• Always replace all fasteners in a joint set when one fastener has failed.
• Never re-use fasteners where the service manual specifies new fasteners when re-assembled.
• Never use fasteners with any evidence of damage like nicks or bad threads.
• Learn to “read” any failed fastener and recognize the signs of a possible fatigue failure.
• If you suspect fatigue, a rule of thumb is to calculate the number of cycles a ferrous metal part has been exposed to before it failed.
  • Failure at less than 10,000,000 cycles:  suspect fatigue.
  • Failure at over 100,000,000 cycles: fatigue is not likely.
  • It only takes 3600 rpm motor only 46 hours to produce 10,000,000 cycles.
  • A 75 Rpm mixer shaft coupling bolt failing after 2 ½ years has over 100,000,000 cycles.
• Get engineering support to analyse the loads when a fatigue situation is suspected; especially where danger to people, environment, or property is a possibility.
• Always torque fasteners to specified values using the proper tightening sequence and procedure.
  • Use the high end of the torque range given, but never over without engineering review.
  • Pay attention to dry / lubricated thread torque values.
  • If a certain type of thread lubricant is specified, use only that type.
• Don’t substitute a different grade or type of fastener without review. You may set the stage for another, more serious failure.
• Don’t substitute a standard grade washer where a hardened washer is specified, and NEVER use lock washers in a joint subject to cycling loads.
• Don’t substitute gasket material types (especially non metallic for metallic types) without a design review.

© Life Cycle Engineering, Inc.

For More Information

843.744.7110 | info@LCE.com