Boeing

Shear Folly

By Tom Anderson

March 2016

I took a job with Boeing in 1961 after I received my master's degree in structural engineering. I worked in the 727 stress analysis group, a bunch of 120 engineers, many from Canada, and was assigned to the controls and empennage section. I loved the work and the insight I could apply from the summers I had spent at several National Science Foundation summer science structural institutes during my graduate teaching years.

Shear flow, q, in pounds/inch, I knew well. The formula is q = T/2A. T is the applied torque to an open, thin-walled, irregular section like the vertical fin, and A is the entire cross-sectional area of the tube-like section. This calculation is the basis for selecting the thickness of the skin.

When I first started in the group, they gave me only jobs checking the stress analyses done by other, more experienced engineers. Within the first few weeks I was checking the "calcs" that had been done to determine the skin thickness for the vertical tail of the 727.

But what to my wondering eyes should appear as I pored over the calcs but the formula q = T/A! No. That can't be right, but sure enough, the stress analyst screwed up. Needless to say, those calculations resulted in the vertical tail skin thickness being two times what it needed to be. And the 727 was tail-heavy, big time, with its three engines aft in the fuselage.

I immediately brought this error to the attention of my supervisor. He almost had a heart attack, had somebody else check my work and then finalized my finding. We went to the weights group and told them they could cut the vertical tail skin thickness in half, and they were delighted. They got bonuses for every pound that they saved.

I guess that was my initiation, which I passed with flying colors. Now they knew that I knew what I was doing, so I was given some really challenging stress analysis assignments to now do the original work, including the flap, aileron, spoiler control box in the wing and the stress analysis of the main spar of the horizontal tail. It was a forged, 30-foot-long (9-meter), bent I-beam under six degrees of freedom forces, supported at its pivot, which did not pass through the spar web -- eccentric. And this was the age before finite element analysis had been developed. I had to use the Least Work Method. That is another long story, but I finally nailed it after weeks of computer calcs.

Needless to say, I had the privilege to work with a good many experienced stress analysts, control system designers and computer-savvy geeks. All of which convinced me that I needed to go back to school to find out where the "edge" was of what we knew in structural theory and analysis. So that launched me into my PhD.

It was a wonderful year in Renton, Wash.