Long time readers of this blog know that one mistake way too many researchers and product developers make is to run poorly-designed simulations of weathering conditions in order to (hopefully) obtained faster results. But in the same way that a computer can give you a wrong answer faster, so too can accelerated aging tests give erroneous results when poorly designed and carried out.
That's why this report about the tarps used by international aid workers is so refreshing. The developer of the tarp, kept trying and failing in his design until he came up with a winner. He didn't do any of his testing in a lab however, he did it all in the real world. He needed to know how well the tarps would stand up to wind so he tied some prototypes to poles and stuck them in the beach along the coast. UV exposure? He let the tarps sit out in the sun.
Why is this so difficult for so many researchers to understand? The exposure conditions aren't as consistent as what we can create in a lab, but they are the real world and the real world is where the ultimate truth about product durability lies. I suspect in this case that since the developer was not an engineer with prior experience on weathering that he did what was logical. And the non-budget busting aspects of his testing were no doubt appealing too!
While some may think that we get so wrapped up in our instrumentation and mathematical analysis that we forget what we are ultimately concerned about, I think the real reason for missing the simple approach is ego. Since you don't need to a four-year college degree to run these simple outdoor-exposure tests, people may not look so kindly at an engineer that suggests such an approach.
For instance, when I was back at Aspen Research, we had a client that was concerned about the internal temperatures of an electronics enclosure. We offered to them 2-D equations for heat transfer that were simple to use and could predict the temperature after countless changes, but no, they wanted a bunch of 3-D FEA color images that had been run on a large computer (said images were generated in part by the same equations we had offered up). The images were useless for predicting what would occur after additional changes were made, but the client (and in particular, the management team at that client) loved the pretty pictures.
That client is no longer in business. Surprised?