OAKLAND -- When very large, high-strength steel rods on the new Bay Bridge snapped in early spring, worried commuters suddenly saw bolts everywhere and wondered why some hold and others fail.
And rightfully so. Steel fasteners clamp together everything from office furniture to spaceships, and questions about their strength are natural.
Metallurgists have been studying the same thing for more than 50 years.
But as the broken bolts on the Bay Bridge demonstrate, not every industry pays the same level of attention to the answers.
Where weight-conscious aerospace and auto designers write highly customized specifications for critical components that absolutely cannot fail, bridge engineers often bulk up on parts and rely on safety in numbers.
As a result, Bay Bridge designers used "off-the-shelf" industry standards for the bolts -- including 32 that later snapped -- which is the equivalent of "going down to Home Depot," said Detroit auto materials engineer Cory Padfield, who is writing a textbook for the American Society for Metals that will feature the busted Bay Bridge bolts as a case study.
"I feel bad for Caltrans because they did get bad steel," Padfield said. "But Caltrans didn't put in enough effort into the initial specifications. The automobile and aerospace people learned this lesson a long time ago, but the construction industry is just now starting to catch up."
Rather than write their own specifications for the myriad pieces and parts on a project, Caltrans engineers and the design consultants they hire rely heavily on standards set by American Society for Testing and Materials International, which has 144 technical committees made up of public and private industry representatives who are experts in everything from cement to steel alloys to electrical conduits.
But the guidelines are voluntary, and registered professional engineers must exercise their best judgment given the specific project.
"ASTM is the minimum, and specifications can be written to modify or increase those levels," said Ron Bianchetti, a corrosion engineer with Russell Corrosion Consultants. "But it is up to the professional engineer to evaluate every element of a project's safety and consider the well-being of the public; it is what we are tasked to do under the law."
One-of-a-kind projects such as bridges are purposefully overdesigned, meaning they contain duplicate and heftier components that give the structures 150-year life spans, experts say. Space and weight are lesser factors in the design, and if a few bolts should crack, it won't -- usually -- doom the entire span.
On the other end of the spectrum, aerospace, airplane and auto manufacturers don't need products to last 150 years, but they do need them to be relatively compact and light. Engineers for those products operate within tight weight and space constraints, where a bigger fastener or an extra bolt is not an option.
When an O-ring seal failed on the space shuttle Challenger in 1986 shortly after takeoff, for example, it broke apart and all seven astronauts died. Seven years earlier, when key bolts securing the engine on a DC-10 airliner sheared, the engine fell off over Chicago and the plane crashed, killing all 258 passengers and 13 crew members.
These industries treat fasteners and other pieces deemed too important to fail as "safety critical" and demand customized fabrication specifications and testing designed to significantly reduce the odds that a bad part will end up on a car, airplane or rocket.
"A rocket launch vehicle is a flying bomb," said aerospace engineer Leon McKinney, whose St. Louis-based firm consults on NASA projects and private satellite systems. "If a part fails and the launch blows up, a lot of bad things happen. At minimum, you lose a $500 million satellite. At worst, it could fall into a populated area and kill people."
The Bay Bridge design team at Caltrans, TY Lin and Moffatt & Nichol, could have written tougher specifications that might have averted the bolt failure.
As it was, the engineers included about 500 pages of special provisions, plus hundreds more in a series of amendments to the eastern span contract. They put the heaviest testing and inspection requirements in the areas where they anticipated the biggest construction challenges for what would become the largest bridge of its kind in the world. But when it came to the bolts, like the vast majority of government transportation departments, they opted for the industry standard and chose a superstrong grade of galvanized steel for its combination of strength and flexibility. The steel was used for 2,306 fasteners on the new self-anchored suspension span, including 96 very large anchor rods embedded in seismic stabilizers.
On the flip side, high strength steel is susceptible to a well-known phenomenon in which hydrogen atoms invade the spaces between the alloy's crystalline structure, causing it to become brittle and fracture. ASTM standards warn that galvanizing increases the chances of embrittlement.
But the bridge team went ahead, relying on ASTM's recommended fabrication practices for "safeguarding against embrittlement of hot-dip galvanized" fasteners.
And as the world now knows, despite the preventive measures, a third of the 96 anchor rods produced in 2008 by Dyson Corp. in Ohio snapped in March within days after contractors tightened them down.
The bolt failure triggered a retrofit estimated to cost as much as $20 million, and cast doubt on the integrity of the span's other fasteners made from the same grade of steel. The problem also delayed the bridge's planned Sept. 3 opening date to at least December.
Caltrans Director Malcolm Dougherty repeatedly has reminded policymakers that the broken fasteners met all ASTM specifications and passed all tests in place at the time. But with critical fasteners broken and unusable and the rest of the steel undergoing additional testing, Caltrans in April backtracked and tightened its specifications.
It's still not enough, countered materials engineer Padfield.
"On safety-critical parts, Caltrans needs to write custom specifications just like Boeing does," he said.
If the bridge design team had seriously consulted metallurgists who fully understood the reasons behind the ASTM warnings against galvanizing high-strength steel, they wouldn't be in this mess, said UC Berkeley materials science professor Tom Devine.
"It may be that you can design and build a bridge without input from a metallurgist, but if you are going to do it, it is incumbent upon you to rigorously adhere to the industry-accepted specifications and standards," Devine said.
Others say the bridge bolt problem shouldn't be used as a call for writing custom specifications for every fastener. Frieder Seible, chairman of the Seismic Safety Peer Review Panel for the Toll Bridge Program, and John Fisher, a fellow panelist, said the snapped bolts were the inevitable consequence of engineering innovation driven by the demand for public safety in a seismically hazardous region. Engineers chose high-strength galvanized rods to meet the state's stringent earthquake safety standard for the new span, while accounting for the corrosive marine environment.
Unfortunately, the panelists said, the research and experience behind the ASTM standards and testing protocols used by the bridge design team was based on bolts that were half the diameter and used at far lower loads.
The Bay Bridge bolt experience likely will lead to engineers turning to lower-grade alloys, dropping hardness levels and reducing the loads or change designs.
"It is one of those moments in time where engineers, as a profession, learn a little bit more and move the state of the art to the next level," said Seible, who has advised Caltrans since 1990.
Still, the Bay Bridge is the first and only span either man has seen where galvanized high- strength fasteners of this grade, diameter and length are loaded to 70 percent of their capacity.
That pushed the rods into the danger zone on three fronts: The higher the strength and the greater the load, the more susceptible the fasteners are to hydrogen-induced fracture. Slightly less than half of the span's 2,306 strongest fasteners are tightened to more than 50 percent, the level metallurgists consider the threshold for embrittlement.
"I didn't even know they were using (this grade of steel) until the bolts failed," said Fisher, an internationally recognized expert on structural steel failures who joined the peer panel in 2008.
On the whole, the panel never discussed the adequacy of the bolt specifications until the March failure, said Seible.