The designer's lot is not often a happy one. The poor fellow is expected to 'get it right first time', and this means planning for the unexpected as well as the expected. Two not-so-recent examples that illustrate the designer's difficulties spring to mind.
The first case concerned the failure of a crane manufacturers test weight. The designer had made a selection of weights by filling steel boxes with scrap and concrete of differing proportions to achieve the weights required. He described the problems as a brittle fracture of one of the lifting lugs on a particular weight, and was concerned about the safety of the others. The 'brittle fracture' diagnosis was based on the fracture surface appearance (no fatigue) and the fact that the failure occurred during the first lift on a very cold morning. Inspection of the fracture confirmed the absence of fatigue, but revealed clear evidence of ductile tearing more akin to static overload than brittle fracture. Some quick calculations confirmed the designer's 'factor of safety' of about two, but how could you overload a test weight? The answer was forthcoming from the test driver, once someone had the sense to ask: 'the thing had been left in a puddle overnight and was frozen solid, so I gave it a good hard tug. and bang!'
The second case concerned some mysterious fatigue cracking in an off-road dump truck. The truck body was fabricated from plate stiffened by lengths of channel welded on to form closed box sections. This particular crack ran right cross the face of the channel and down each side, terminating at the weld between channel and plate. Now, fatigue cracking in this sort of machinery is nothing new, but cracking initiating in plain material away from welds is a little peculiar to say the least. Closer inspection of the face of the channel revealed some irregularities beneath the yellow paint - a vital clue. Removal of the damaged section confirmed our suspicions: a large hole had been cut in the channel in error during manufacture of the skip, and carefully filled up with weld metal from the outside only. Skilful grinding of the 'repair' followed by a thick coating of paint had cleverly disguised the mistake, but the globular weld deposits on the inner surface were best likened to a bunch of grapes! A better fatigue crack was hard to imagine.
Both of these above failures arose from circumstances that could hardly have been foreseen by the designer, but the moral is clear: expect your product to be abused in every way possible and design against or around it wherever you can.
For information about TWI's capabilities please contact us.