Fatigue Could Loom Large in Bridge Collapses

DN Staff

November 19, 2007

4 Min Read
Fatigue Could Loom Large in Bridge Collapses

A spate of news stories about structural deficiencies in bridges is leading engineers to examine an age-old culprit: metal fatigue.

Discussions of fatigue have gained momentum in the wake of the bridge collapse in Minneapolis in August and following news that the Federal Highway Administration has designated countless bridges across the country as "structurally deficient." Experts and forensic engineers appearing on CNN and elsewhere in the national media have demonstrated the phenomenon by bending paper clips until they pass their elastic limits. The underlying message: Fatigue is causing plastic deformation in the supporting members of the bridges prior to failure.

"If you get sufficient bending, things can start moving atomically," says Richard Weyers, professor of civil engineering at Virginia Tech. "After repeated loading, the stresses can go into the plastic region (of the stress-strain curve). That's called fatigue."

Overloading and Corrosion Key

Determining the causes of that fatigue has been difficult, however, especially in the case of the Minnesota bridge collapse. Most discussions have focused on two issues: overloading and corrosion.

The problem of overloading could be problematic on older bridges, particularly those built in the 1960s and '70s. "The number and magnitude of the loads have an influence on fatigue life," Weyers says. "If your actual weights have increased over time to a point that's greater than anticipated or if you have overloads crossing the bridge on a daily basis without your knowledge, you will get fatigue damage."

Engineers worry about overloaded trucks and construction equipment passing across bridges. Over the course of 40 years, with trucks getting larger and the number of overloaded trucks growing, damage becomes irreversible, they say. Moreover, big pieces of construction equipment may drop heavy loads on a bridge, eliciting a dynamic effect that designers never anticipated. Some dynamic loads can be twice as much as a static load of the same mass, they say.

At the same time, fatigue can be aggravated by corrosion. Corrosion fatigue, which reduces the designated fatigue strength of steels, can be caused by chemicals, such as salt, or by something as seemingly innocuous as bird droppings.

"Corrosion fatigue is much more subtle than overloads," says William O'Donnell, president of O'Donnell Consulting Engineers and chairman of American Society of Mechanical Engineers' technical committee on fatigue for the past 30 years. "In many cases, a bridge is overloaded and corrosion fatigue is related to that overload."

Whether fatigue is caused by overloading or corrosion, the end result is micro-cracking of the support steel, engineers say. Such cracks, almost invisible to the human eye, can propagate under normal loads that occur later. Even then, however, there's no guarantee bridge inspectors will be able to see the cracks.

"Typically, today's inspection technology isn't going to find microcracks," O'Donnell says. "And even though the crack growth rate may be very small per loading cycle, if you put a million cycles on it, that crack can grow to be quite large."

O'Donnell says consulting engineers can gather crack growth rate data by outfitting bridges with dynamic instrumentation that will provide data on dynamic loads on the bridge. That, however, can't be done by bridge inspectors, who must climb over miles of structural steel each year in their search for telltale problems.

"Usually, the bridge inspector is stuck with the limitation of just looking to see if the cracks are already there," O'Donnell says. "The technology it takes to do crack growth rate calculations is beyond the bridge inspector's task."

Search for the Smoking Gun

The question is whether knowledge gained during decades of material science and structural analysis has left structures more susceptible to miscalculation and unexpected overloading. As engineers have migrated from old-fashioned working stress designs to ultimate strength design methods and from slide rules to computers, structural members have been sized less conservatively than they were in the 1920s and '30s, when many American bridges were built.

"As you get more confidence in the reliability of your design methods and more confidence in the reliability of materials, you're able to design closer to the true limit characteristics," says Weyers of Virginia Tech. "But when you do that, you have to be sure of your estimates of traffic growth, as well as the number and magnitude of loads. If the number and magnitude of loads is beyond what you anticipate, it will have an impact."

As engineers continue to search for the cause of the Minnesota collapse, many are hinting that overloads could be having an insidious effect on some bridges, most likely causing a form of metal fatigue that isn't easily spotted.

"That's what you're dealing with," O'Donnell says. "And it probably explains why investigators haven't turned up a smoking gun."

Senior editor Chuck Murray started his career 29 years ago as a bridge designer for the Chicago & Northwestern Railroad.

Sign up for the Design News Daily newsletter.

You May Also Like