One of the more unpleasant aspects of engineering is having to occasionally perform a failure analysis on a returned part and then write a Failure Analysis Report (FAR) about it.
When a customer returns a part, it means it was non-compliant in some way. Maybe it didn’t meet the mechanical specifications, maybe it shorted out electrically, or maybe the cause for return was cosmetic in nature. Maybe even there was nothing wrong with the part – it met all the specs – but didn’t work properly in the customer’s application.
From an engineering point of view, a returned part means the customer is probably unhappy, or concerned about additional failures you may or may not be aware of. If it left your factory, then the part failed in your customer’s system, or their customer’s system, or somewhere ‘in the field’. Many times there is more than one part, maybe they’re stockpiled somewhere, or the customer has a large inventory on hand.
Often there is a lot of pressure to get it right in determining the cause of failure, and chances are, you might be feeling overwhelmed. Many times there is a lack of information about operating conditions in the field, how the part was being used, or even how it failed. Unknown conditions like voltage transients can be particularly tricky, as can any intermittent condition. Trying to figure all that out is part of writing a FAR.
Keeping in mind the humorous aspects of human nature, not all engineers approach the analysis in the same way. Some are determined to prove the design is good, and the customer just misused or broke the device. Others assume the design or manufacture was inherently flawed. Most of the time, the truth is somewhere in the middle, and in the end, has to be substantiated by ‘the facts’.
Either way, the goal is to keep the customer happy. Sometimes that means writing Failure Analysis Reports and making sure the parts never break again. Painful as they may be, once in a while, good things come out of FARs. Sometimes they spur a new revision on an old design, or a new design, and sometimes a re-examination of the manufacturing process or raw materials. And that can be a good thing!
What was your worst ever Failure Analysis experience?