Oblivious to the risk. (Editorial).

"I called it 'Chronicle of a Failure Foretold,'" says Prof. Tim Osswald, referring to a speech he gave recently about an example of faulty design of plastics parts. Osswald, director of the Polymer Engineering Center at the University of Wisconsin--Madison, pointed out an a]l-too-common error committed by designers inadequately trained in the ways of plastics. "I looked at the creep data for the material in that part and I knew it wouldn't last five months." Unfortunately, the designer was more familiar with mechanical design for metals--which don't creep-and overlooked the stress-relaxation behavior of plastics. "It's scary. People are designing with plastics using the wrong data. I get calls on lawsuits about failure of plastics products. The causes are often things that should have been fixed on day one."

Poorly designed plastics parts are, directly or indirectly, a liability to everyone who makes their livelihood in plastics manufacturing. Nowadays, plastics-processing firms are likely to perform some design services, so they take on even more responsibility for the performance of their products. That's why Prof. Osswald's new research project should matter to you.

"I want to conduct a survey to find out what property data people use, or don't use, to design plastics parts," Prof. Osswald says. At the last SPE ANTEC meeting, he polled 80 plastics designers and found that 80% of them used just short-term, single-point mechanical-property data like flexural modulus, and only 20% used longer-term creep data that add the element of time to stress-strain curves. Prof. Osswald found those results "appalling." The reason, he says, is that there aren't enough good educational programs in designing with plastics. Out of some 200 mechanical engineering programs in the U.S., at most 25 give solid grounding in plastics, he claims.

That's why he wants to conduct a larger-scale survey of plastics designers to document the problem. If you, or someone else at your firm, wants to help or participate, call Dr. Osswald at [phone] or e-mail [e-mail].