John Bennett went to West Virginia in January 1968 to look at pieces of a bridge.

By then the bridge itself no longer existed in any useful sense. The Silver Bridge had fallen into the Ohio River more than a month earlier, scattering steel across the river, its banks and whatever happened to be underneath it when it came down. Forty-six people were dead.

Bennett worked for the National Bureau of Standards, predecessor of NIST. He examined a particular fracture through an eyebar stamped 330. One side of the break looked relatively flat. The other showed much more deformation. His early conclusion was that the first side had fractured progressively and the second had torn afterward under the sudden excess load.

That judgment turned out to be essentially right. The harder question was how something small enough to fit under a thumbnail had begun the sequence.

A bridge held up like a chain

The Silver Bridge opened in 1928 between Point Pleasant, West Virginia, and the Ohio side of the river. Its suspension system did not use the familiar bundles of steel wire. It hung from chains of enormous steel eyebars: flat links roughly twelve inches wide and two inches thick, joined through circular eyes by large pins.

The bicycle-chain analogy is useful up to a point. A bicycle chain can lose a link and stop working. The Silver Bridge had a comparable problem at a much more consequential scale. Its suspension chains used pairs of eyebars rather than a larger redundant group capable of carrying the load after one critical link disappeared.

The steel itself mattered too. The eyebars used high-strength, heat-treated carbon steel. Later NBS work found a material that could behave normally in an ordinary tensile test while having comparatively low resistance to crack propagation under the wrong conditions.

Nine days

The bridge had not simply been forgotten. West Virginia records inspections in 1959, 1963 and 1964. A 1965 inspection produced recommended repairs that were carried out. Two more inspections occurred in the summer of 1967. A state area maintenance engineer visited again on December 6.

Nine days later the bridge collapsed.

It is tempting to stop there and write a story about negligent inspection. The federal investigation does not let us do that. The flaw that mattered was inside an eyebar joint, at the surface of the eye hole where the steel fitted around the connecting pin. NTSB concluded that the location was inaccessible to visual inspection and that no inspection technique then available could have detected the defect without dismantling the joint.

People had looked at the bridge. The bridge contained a place they could not meaningfully look into.

December 15

At about five on Friday afternoon, thirty-seven vehicles carrying sixty-seven people were on the bridge. Thirty-one vehicles went down with it. Twenty-four fell into the Ohio River and seven onto the Ohio shore. Forty-six people died; nine were injured.

The exact visible duration is less neat than later retellings make it. NTSB described the progressive structural sequence as unfolding within about a minute. West Virginia historical accounts preserve eyewitness descriptions of less than twenty seconds. I would leave both where they belong. Nobody on the bridge was running a stopwatch.

The structural sequence is clearer. Eyebar 330 fractured at joint C13N in the Ohio side span. Once it separated, its sister eyebar slipped from the connecting pin. The north suspension chain opened at that joint. The bridge did not have a redundant path capable of redistributing what the chain had been carrying.

One man got out of a truck

Later West Virginia accounts preserve truck driver William Needham’s experience. His truck went into the river. His partner, asleep in the rear of the cab, did not escape. Needham forced open a window as the truck sank and used a box to help himself reach the surface.

I would not build the entire disaster around one survivor. There were sixty-seven people on the bridge. But his experience restores something structural diagrams remove. “Progressive collapse” eventually became a truck filling with river water.

They put the bridge back together

The federal investigation treated the wreckage somewhat like an aircraft accident. The Bureau of Public Roads arranged a twenty-five-acre reassembly site downstream where recovered members could be identified, laid out and examined while the Corps of Engineers cleared the navigation channel.

FHWA engineers also built an eighteen-foot model for the structural working group. That decision says something about how uncertain everyone initially was. If you already know which piece caused a bridge to collapse, you examine that piece. If you do not, you collect the bridge.

0.12 inches

Laboratory examination found the pre-existing crack at the surface of the eye hole. NIST’s later reconstruction gives it at roughly 0.12 inches deep and 0.28 inches long. FHWA histories often round the depth to about one-tenth of an inch.

The bridge did not collapse merely because a tiny scratch existed. Several conditions had to occupy the same small region of steel: material vulnerable to stress-corrosion cracking, corrosion in a difficult-to-protect joint, high local stress and fracture toughness low enough that a strategically located crack could become catastrophic once it reached critical size. NTSB described stress corrosion and corrosion fatigue acting together over the bridge’s forty-year life.

The twin bridge closed

About seventy-five miles upstream stood another 1928 bridge built around essentially the same concept, later known as the Hi Carpenter Bridge at St. Marys. It shared the same engineer, contractor, fabricator and shop drawings.

The St. Marys bridge closed immediately. The problem was almost painfully logical: if one bridge could contain a hidden stress-corrosion crack inside an eyebar joint, how could anyone prove the other did not? It was never returned to service and its suspension spans were later removed.

America did not know how many bridges it had

The collapse widened beyond Point Pleasant quickly. The United States did not yet have a nationally coordinated system for inventorying and systematically inspecting highway bridges. Federal officials began a broad examination of older bridges, and Congress included national bridge-inspection requirements in the Federal-Aid Highway Act of 1968. The first National Bridge Inspection Standards took effect in 1971.

The modern descendant is more specific still. Current federal rules identify nonredundant steel tension members whose failure could leave a bridge unable to redistribute the load. FHWA guidance calls for hands-on inspection: the inspector must get close enough to find small cracks and other defects, with nondestructive testing used where appropriate.

I would resist saying every modern bridge rule came from Silver Bridge. It did not. Bridge safety accumulated lessons from many failures and many technologies. But FHWA itself traces the national inspection program directly to December 15, 1967.

That institutional thread is why this story belongs beside They Checked the Dam That Morning. In both cases, the uncomfortable part is not simply that somebody inspected a structure before disaster. It is what the available inspection could actually tell them.

Inside the joint

The photograph of the wreck is dramatic. The fracture itself is less satisfying. There is no enormous missing bolt or obvious hole announcing, from across the room, that forty-six people died because of it.

Before the collapse, the crack had been about an eighth of an inch deep. The bridge had been inspected. Repairs had been made. A maintenance engineer had visited nine days earlier.

None of those facts makes the investigation contradictory.

The important surface was inside the joint.

Sources & Further Reading

• National Transportation Safety Board, Collapse of U.S. 35 Highway Bridge — definitive federal finding on eyebar 330, the collapse sequence and the inaccessible flaw.
• National Institute of Standards and Technology, investigation of the Point Pleasant bridge collapse — metallurgical work and measured pre-existing crack.
• Federal Highway Administration, National Bridge Inspection Standards 50th anniversary history — the investigation, scale model and development of the national bridge-inspection program.
• Bureau of Public Roads, Reassembly of Point Pleasant Bridge — documentation of structural recovery and reassembly.
• Federal Highway Administration, bridge inspection program guidance — current inspection framework for critical steel members.