We rarely give thought to all the things that work right. The wheels on the car that don't fall off. The glass door that slides open as we approach. The elevator that moves predictably between floors.
These are only noteworthy when they fail. Lucky for us, engineers think about them.
So you know those towering light fixtures that illuminate the Seward Highway, the Glenn Highway and Minnesota Drive? We have 128 of them in Southcentral Alaska that scratch the sky at 120-170 feet, dwarfing their more ordinary 25- to 40-foot cousins.
Well, these tall and elegant light poles, so useful during our long, dark winters, have a problem: bolts that persistently loosen. Alaska highway inspectors have repeatedly noticed that the long bolts attaching the poles to their foundation plates come loose. They tighten them, come back later and find them loose again. Not good.
To make sure a serious problem doesn't develop, the Alaska Department of Transportation stepped up inspections, which turned out to be a substantial labor investment. Each pole has either 12 or 16 bolts and each bolt has to be hand-inspected. If it's loose, the tightening process can take four steps per bolt. We could be talking 6,144 moves or more just to make sure 128 light poles stay where they belong. And repeat.
To be fair, not every bolt loosens, and the poles have never even been close to failure. But it's a mysterious, nagging problem and a labor cost that any engineer worth his salt would rather solve. Looking for a permanent fix, state highway engineer Charles Wagner conjured some design modifications, like adding an additional plate for the bolts to attach to, trying bolts with finer threads and considering whether the plates themselves are flat enough.
But Wagner's not alone in his quest for a solution. A UAA civil engineering team is also on the case, and the eventual solution will become one graduate student's engineering master's thesis.
Matchmakers in this collaboration are UAA professor Scott Hamel and the Alaska University Transportation Center in Fairbanks, a coordinator for federal and state research aimed at solving Alaska transportation problems. The center began in 2005 with state and federal support. Billy Connor, a retired state highway engineer, runs it.
As in this case, solutions often entail connecting a University of Alaska team with a problem that needs solving. Hamel listened to state engineers describe the problem and made a pitch for research dollars to solve it. Sixty thousand in federal money came through from AUTC, matched by the state. Hamel enlisted two graduate students, David Hoisington and Daniel King, and the engineering forensics began.
In February, they installed monitoring gear on an existing pole along the Glenn Highway so they could gather data on temperature swings, wind direction and speed, and strain on the light pole bolts. They'll outfit another pole near Peters Creek later this summer.
Bolt stretching is particularly interesting to them. One hypothesis is that the way the bolts get tightened may lead to the problem. The process goes like this: Workers tighten the first bolt to "snug tight," then move across to an opposite bolt and tighten that one to "snug tight," and so on, much as you do when you change a tire. Next come three rounds of 20-degree turns to each bolt.
It's that "snug tight" business that could be the problem. The definition is "the full effort of a workman on a 22-inch lever arm wrench," Hoisington said, hardly a precise measurement. Inadvertently, workers may tighten bolts past their "yield point," leading them to "stretch" too much and eventually loosen.
It's hard to imagine steel bolts stretching. Hoisington offers the rubber band analogy. If you stretch it shy of its yield point and let go, it snaps back into shape. But if you stretch it past that point, it's weaker and gives way more easily to additional stress. Any more pulling results in a stretched out, useless rubber band. I have old bungee cords like that.
Steel bolts are not that different. When you tighten one into the light pole's foundation plate, the force you apply clamps the plate down to the nut by stretching the bolt up. Tighten the bolt too much and it loses structural integrity like the rubber band. Bolts that have "yielded" could be the problem.
If the bolts are indeed too tight, the plates they secure could also be affected, Hamel said. Picture Jell-O. If you squeeze it lightly, the Jell-O rebounds. If you squeeze harder, the Jell-O deforms and won't spring back. That could be happening to the plates, again leading to loose bolts.
Still ahead: more data monitoring and analysis, and some computer modeling. Factors like temperature change, or wind speed and direction, are still under consideration.
Although Hoisington says he won't relax until his master's thesis is done next spring, preliminary results are encouraging and "this engineering detective work is kind of fun."
DOT's Wagner says he appreciates fresh eyes on the problem.
Kathleen McCoy is an electronic media specialist at UAA, where she highlights campus life through social and online media.