Monday, July 14, 2008

Accelerated Aging - Getting Bad Data Even Faster - 1st in a Series

Accelerated aging is a laboratory technique for simulating long periods of outdoor exposure in a much shorter time period. This is typically done by exposing samples to various UV lights with or without occasional water sprays, and even some dark periods where the light is shut off. Since the UV lights can run 24/7 while the sun shines for considerably less time (most of which is not a peak intenisty) just running a sample under a bulb as luminous as the sun could give faster results, while having an even more luminous bulb could give even faster results.

I say "could" for good reason. This testing is far more complicated than it appears (my employer does lots work for people that don't do their accelerate aging properly). All the pitfalls are too numerous to mention, but one for starters:

1) Acceleration of the wrong failure mechanism When a part fails in the outdoors, there is a specific chemical or physical mechanism that leds to this result. It is critical that this specific mechanism is accelerated in the testing more than any other mechanism. If this isn't done, your results will be meaningless as you've accelerated a mechanism that is not of interest.

No comments: