One aspect of my job that I really love is Failure Analysis.  This is the process of investigating why a piece of equipment failed prematurely. One of the first things they teach you as a mechanical engineer, is that you can’t design anything to last forever. It would simply be too expensive. (I can see the Greenies eye-balling me right now as I say this). So we design everything – a pump, a shaft, a ceramic mug, to have a certain lifespan, measured in years, number of times used or number of vibrations/ rotations. Something like a rotating shaft will be designed to rotate millions of times before failing (in the region of 10^7 rotations), but other things are designed to break after just one use (opening a coke can). 

So everything is designed to break predictably, which helps engineers plan the maintenance schedule and life-cycle costs of a particular piece of equipment. But what happens when something breaks prematurely?  Do we just go out and get a new one to replace it?  No, that would be naïve. Engineers are problem solvers, we crave solving puzzles – so figuring out why that part/ piece of equipment broke (failed) in order to fix the problem and prevent future failures is a field called failure analysis.

Now this is a really broad field, as in order to understand why a particular part broke, you have to understand the entire system as a whole. Then you need to do a thorough investigation of the ‘crime scene’ which includes taking interviews of everyone involved. The rule of thumb though, is that ‘everyone lies’, so investigating the actual material is the only way to really put the pieces together. I love it because it’s a little like CSI. Actually, it is EXACTLY like CSI!

So what you do next is collect all the fragments of the thing that broke and try and put them back together again, laying them out in their correct position. This is sometimes easy, in the case of a split drive shaft where the two pieces obviously fit in a certain orientation, but may become very complex when you have something literally blown to bits, like an airplane fuselage or any other pressure vessel.

Next you investigate the material properties of the item that broke and search for tell-tale signs to indicate where the failure originated. Different materials behave differently upon failure and it’s a learned skill to identify the origin of a failure.

Sometimes, the origin is impossible to find, so you would need to dig a little deeper and use other clues to figure out what went wrong. You may do some non-destructive testing (x-ray/ dye-penetrant) or perhaps take a sample of the material and test that its properties are what the manufacturer said it was. You may also want to look at the way the item was used or its history – this sometimes is the key to solving the problem. The other obvious cause could be an error in the design, so you will need to find all the drawings of the object and do a few calculations to check that it was sound for both fracture (brittle failure) and ductile failure.

There is a certain thrill that one gets from figuring out what went wrong – and every failure is different so it’s quite exciting.  If you’re interested in failure analysis, I suggest you take a fracture mechanics course to supplement the mechanics of solids courses standard in a mechanical engineering degree; any materials science courses including a manufacturing with materials and manufacturing processes course will also help.  Finally, you may want to look into systems analysis.  A sound understanding of electrical systems will also help a lot, and from my experience, a bit of civil/ structural design knowledge is very critical!

Learn more about FAILURE ANALYSIS here :  http://failure-analysis.info/2010/05/analyzing-material-fatigue/