Simplifying Root Cause Analysis – Case Study Two

Apr 15, 2019 11:33:03 AM

An example of why retaining and analysing failed parts is essential to getting to the root cause of a failure.


In our recent article on identifying failure modes, we talked about the importance of retaining failed parts for analysis.  We said that in most cases, the maintenance execution team can determine the failure mode.  However, occasionally a site will need to send the failed component for forensic analysis.  In the case study below, we provide an example of one failure investigation where an off-site forensic analysis gave us the information we needed to get to the root cause very quickly and easily.

We were asked to investigate the failure of the chain master link (Figure 1) on an apron feeder (Figure 2).


simplifying root cause analysis case study 2 pic 1

Figure 1- Master Link


failure location pic 2

Figure 2 – Failure Location


A preliminary inspection found that one of the studs on the non-drive side master link sheared (Figure 3), which led to the subsequent failure of the second stud and eventually the link itself.


failed studs pic root cause

Figure 3 Failed Studs


The maintenance execution team examined the failed components and determined that, due to the amount of corrosion on the failed stud (Figure 4), the failure had likely been developing for some time.  It was considered likely that the failure was due to insufficient pre-tension of the stud; however, there was nothing to substantiate the theory, and the maintainers were obviously reluctant to admit to anything relating to maintenance execution quality.


failed stud showing corrosion on failure surface

Figure 4 - Failed Stud Showing Corrosion on Failure Surface


Given the state of the failed components, we decided to send the master links off-site for metallurgical analysis.  When the report came back, the cross-section of the failed master link (Figure 5) was quite clear-cut: a “plug” of thread lubricant had been pushed down to the bottom of the stud, preventing a complete tensioning.  Close-up views of the failure location (Figure 6) showed the ingress of contamination into the thread, which eventually resulted in corrosion fatigue.


cross section from failure analysis

Figure 5 Cross-Section from Failure Analysis


close up of failure area

Figure 6 Close-up of the failure area


With the failure mode now known, we could focus on our investigation on the indirect and root causes.  The maintainers had applied a substantial amount of anti-seize compound onto the studs to make them easy to remove.  Their actions were consistent with the task instructions, however, they also admitted to using an impact gun rather than a torque wrench to tension the studs (in contravention of the procedure).  Had they used a torque wrench, they might have noticed something was wrong.

Therefore, the conclusion was twofold – an incorrect maintenance strategy (calling for the use of anti-seize), and inadequate maintenance execution quality.  Drilling further, we identified that there was a general lack of knowledge of this failure mode across the business, and that there was a belief that using anti-seize was always good because it allowed you to remove bolts more easily so you could re-use the master link.  Given that the master link was a relatively inexpensive component, it made more sense to simply replace them each time we rebuild the apron feeder than to risk this failure recurring.  One solution therefore was to change the maintenance task instructions, along with conducting a general education campaign for the maintainers.

However, changing the task instruction would be insufficient if the maintainers still didn’t follow them.  A number of leadership and cultural issues also had to be addressed to improve the maintenance execution quality.

The failure investigation performed in this case study is an example of when it can become necessary to have a forensic failure analysis performed.  Without a sound technical explanation, we would never have reached the root causes and may have focussed our defect elimination projects on the wrong factors.