The way Jim Sawyer figures it, the collapse of a 50-foot length of the Bay Bridge may have begun with a two-cent rivet. Sawyer has made a career of designing and examining bridges, and he is sure that the seismic blast from the earthquake snapped a row of rivets and caused a patch of concrete five lanes wide to give way on the top deck of the Bay Bridge.
The falling slab pierced the traffic lanes on the lower deck, exposing daylight and the chilly bay waters 250 feet below. Two cars wedged by the falling debris teetered on the edge of the damaged roadway. Initial reports said some motorists jumped into the water, but the Coast Guard said there is no evidence of that.
It was a staggering scene, one that framed the disaster for a nation that had tuned in to watch Game 3 of the World Series only to see live pictures of a stricken bridge.
And Sawyer, a Tampa Bay civil engineer watching on television in Florida, could not help but think about the rivets that join the huge concrete panels on the bridge.
"It was probably a connection failure, right about where the rivets sit," Sawyer said. "Chances are it wouldn't happen in a hundred years. Those rivets are routinely inspected."
But Sawyer and other structural experts who watched the quake's flattening power say it is impossible to examine each joint, each beam and each rivet.
"Truth is you can't go over it with a microscope," Sawyer said.
In time, Caltrans officials will eyeball, photograph, measure and note what happened to the failed link of the Bay Bridge and draw conclusions as to the cause. And they will undoubtedly ponder why the Golden Gate Bridge, the more glamorous of the twin spans that link San Francisco with its bayside neighbors, weathered the shake with little more than a superficial crack or two.
One theory advanced by experts focuses on the inability of the Bay Bridge to bend or give when struck sideways by the powerful ground waves generated by the temblor. As evidence, the roadway collapse occurred along one of the bridge's most rigid sections.
"The bridge didn't give. It broke," said one structural engineer, who requested anonymity because he has a contract with the state. "It was more like a stick that snapped than a rubber band that stretched. That jolt may have displaced the bridge as much as a foot before that section went."
State highway officials have declined to speculate, saying it is too soon to know what caused the bridge failure. Caltrans has spent $58 million since the 1971 Sylmar earthquake to strengthen and upgrade 1,500 bridges in quake-prone areas, including the Bay Area. Among the retrofitted structures was the Bay Bridge, where crews retied the bridge decks to column supports, Caltrans spokeswoman Lisa Covington said. A second $64-million phase to reinforce more than 700 bridge columns and supports statewide should be completed by 1992.
"Some retrofitting of the Bay Bridge has been done, but what sections had been worked on is unclear," Covington said.
Jim McCarty believes the busted bridge may have had more to do with geography than concrete patch jobs. The Bay Bridge, built during the Depression, runs east and west, tying Oakland with San Francisco. When the quake hit, it sent a powerful ground swell north from its epicenter near Santa Cruz toward the Bay Bridge, said McCarty, Oakland's public works director for two decades.
The Bay Bridge incorporates three engineering styles, with the Oakland side of the bridge supported by metal trusses. Unlike the Golden Gate Bridge or even the San Francisco side of the Bay Bridge, both of which are supported by cables that give when rocked by a quake, metal trusses tend to be more resistant to movement, McCarty said.
Thus, when the quake's powerful groundswell broadsided the bridge, a chunk of the upper deck collapsed, McCarty said.
The Golden Gate Bridge, which opened a year after the Bay Bridge, was spared because it was perpendicular to ground motion. The Golden Gate's roadbed, resting in a massive sling of cables and span, swayed with the 15-second shake but survived intact.
"A truss bridge is much more rigid, much less forgiving," McCarty said. "The TV pictures showed us that."
The section of bridge that failed is over a pier support, and McCarty said he believes that the columns supporting the roadway "separated ever so slightly, and boom, the concrete gave way."
Moments after it happened, Thomas Kelly was driving his camper on the bridge's upper deck, trying to return to San Francisco. Then, as he videotaped the confusion among rattled motorists around him, he witnessed a car falling through the crevice in the roadway.
"It's a moment frozen in my mind," said Kelly, an Oklahoman who was vacationing in the Bay Area with his wife, Debbie. "One minute the car was there. The next it was gone. I really thought we were next. I thought the entire bridge was going."
The Golden Gate remained opened, and after several closures to inspect joints and seams, the San Mateo Bridge was declared safe and passable for motorists fewer than 24 hours after the quake.
Reopening the Bay Bridge, traveled daily by half a million motorists, is less certain.
A large marine crane moved into place Thursday below the wounded structure to lift the downed section bridge in an attempt to assess the damage. Officials also considered spanning the gap with steel girders called Bailey Bridges, a World War II innovation to move tanks and heavy equipment across waterways.
It is a temporary fix, at best, and one Caltrans official said rather than putting a "Band-Aid on the hurt, we'd like to fix it right." But that could mean it will be several weeks before the toll bridge is open again for business.
"That bridge has done a damn good job for 50 years," said McCarty, a Bay Area resident for most of his life. "Until it's reopened, life just won't be the same."