Civil Engineering Reference
In-Depth Information
Broken Bridges
Shortly after the 1994 Los Angeles earthquake struck, a veteran engineer called me to talk about
the broken bridges, pictures of which were ubiquitous in television and newspaper reports of the
disaster. In the fallen spans, which had come to epitomize the destructive and disruptive force of the
quake in auto-centric southern California, he saw both classic and confusing examples of the failure
modes that so dominate engineering-design considerations. Since the objective of engineering is to
identify and obviate failure modes in the design of structures such as bridges, when failures arise
unexpectedly they are naturally the object of close scrutiny. What caused the classic and unfamiliar
modes of failure? the engineer wondered. Such questions, which occur to engineers and nonengin-
eers alike in the wake of failures, not only help us to understand the state of the art of engineering
but also help to illustrate its nature and limitations. The questions, however, are often articulated
more easily than are definitive answers.
Among the things the old-time engineer wondered was whether sufficient information about the
failures was going to be collected before the debris was cleared, and how effectively such informa-
tion would be disseminated to the wider engineering community, young and old alike. Because the
design and construction of large civil structures, like the major highway bridges that failed around
Los Angeles, take place in a context of theory and practice that is evolving constantly and that
is never tested completely, any failure presents a rare opportunity for a fortuitous experiment to
test hypotheses on which bridge designs are necessarily based. Unfortunately, just as economic and
physical considerations argue against full-scale tests of large structures prior to their construction,
the same considerations severely limit the time frame in which information about the failures can
be collected.
The engineering ideal would be to leave the broken bridges untouched so that engineers could
pore over old calculations and carry out new ones while they designed and performed tests on
computer models, scale models, and even full-scale models provided by the adjacent but unfailed
bridges. As calculating and testing progressed, naturally revealing new insights and speculations
about what exactly happened during the earthquake, the failure site could be revisited to check de-
veloping theories against reality.
A model approach is followed in the wake of airplane failures, in which case the National Trans-
portation Safety Board (NTSB) has the authority to cordon off a crash site so that every piece of
debris can be collected, arranged, and preserved in a nearby hangar until the investigation is com-
pleted. Even then, recognizing that no matter how careful an investigation and no matter how incon-
trovertible an explanation for a crash may seem, the NTSB declares only a “most probable cause”
for a crash.
Although such a failure-analysis procedure can be imagined for fallen highway bridges, the prac-
tical constraints against it are overwhelming. Who, for example, could argue effectively that the
broken bridges should remain as they fell on the freeways around Los Angeles (because the parts
Search WWH ::
Custom Search