Broken Bridges - Pushing the Limits: New Adventures in Engineering - page 74

are too massive to move intact) until engineers are satisfied that they have collected all of the in-
formation they want? Pressures dictate that the debris be broken up and removed as quickly as pos-
sible so that the heavily traveled highways can be reopened with a minimum of delay.
Civil-engineering structures differ from aeronautical creations in that the former can never be so
readily tested under controlled conditions as the latter. Because aircraft are maneuverable by design,
it is economical and feasible, not to mention prudent, for new designs to be subjected to extens-
ive test-flights. Furthermore, because airplanes are built in numbers and operated in fleets, having
one out of commission (because of an accident or for maintenance or repair or testing) presents
no significant interruption in service. A major bridge, on the other hand, can practically be fully
shaken only with the force of actual wind or a real earthquake, which is of course an uncontrolled
and unplanned experiment, and the most ambitious bridges are usually built as unique examples.
(Some rare exceptions are the parallel pair of virtually identical suspension bridges, built seventeen
years apart, carrying I-295 across the Delaware River, and the twin drawbridges that will replace
the Woodrow Wilson Bridge, which carries I-95 across the Potomac.) Computer and scale-model
tests are possible and can be performed on bridge designs, of course, but these are always subject to
the limitations of being tests of models of the structure rather than of the real structure itself.
Highway-bridge failures have seldom claimed as many lives as an airplane crash, but there have
been exceptions. The collapse of the double-deck highway in Oakland, California, during the 1989
Loma Prieta earthquake claimed forty-two lives, and the number of dead and injured during the
1994 Los Angeles earthquake might have been considerably greater had the incident not taken place
so early in the morning.
The greatest losses of life in bridge accidents in America have typically arisen not from natural
disasters but from faulty or poor design or construction. More than eighty people died when the
Ashtabula Bridge in Ohio collapsed in 1876 while a railroad train was crossing it in a snowstorm.
Although the cause of the accident is still the subject of discussion, it is generally believed to have
been due principally to faulty cast-iron construction of the truss bridge. In 1967, the Point Pleas-
ant Bridge across the Ohio River collapsed suddenly, killing forty-six of the people caught in rush-
hour traffic between Gallipolis, Ohio, and Point Pleasant, West Virginia. The failure was traced to
corrosion-assisted cracking that had progressed undetected in the eyebar chain links, in a location
that was largely uninspectable. Some of the most famous bridge failures, such as that of the Tacoma
Narrows, claimed no human lives.
Can bridges be made perfectly safe? In particular, can they be designed to be totally immune to
major earthquakes? Although such an ideal state is possible in theory, it is highly unlikely for sev-
eral reasons, including economical, political, aesthetic, and technical ones. The economics of the
situation are such that highways and bridges that could withstand all earthquakes would cost a great
deal of money and so would have a severe impediment to getting built in a political climate where
there is considerable competition for scarce financial resources. The realities are that bridges are
designed for a certain kind and level of earthquake, which seems at the time to be the maximum
credible one.
The engineering problem, simply stated, is always to design a good-looking bridge that is both
safe and economical. Like all engineering-design problems, this one is fraught with conflicting ob-