Broken Bridges - Pushing the Limits: New Adventures in Engineering - page 76

amounts of energy are released from slipping at a steep angle to the horizontal. The unique char-
acteristics of such a fault were demonstrated by recordings made by the California Strong Motion
Instrumentation Program at various distances from the epicenter. About five miles away, at Tarzana,
for example, maximum accelerations reached 1.82 and 1.18 times that of gravity in the horizontal
and vertical directions, respectively. Not only are such accelerations more than an order of mag-
nitude greater than those originally assumed in the design of the Bay Bridge, but the strong vertical
component showed the earthquake to be of an unusual type, one in which structures like bridges
would be shaken considerably not just horizontally but vertically.
An unusually strong vertical motion coming from a previously unknown fault is not the kind of
condition that an engineer can readily design against. If highway engineers had in fact proposed
taking into account such extreme conditions when the roads and bridges in the area were being de-
signed, they might reasonably have been expected to justify their assumptions. Without a history
of strong vertical ground motion and without a knowledge of a blind thrust deep below the ground
surface in the area, it is highly unlikely that plans for bridges designed to withstand such incredible-
seeming conditions would have been approved and built.
Even after the Northridge earthquake broke key links in the Los Angeles freeway system, there
remained some open questions about what caused some of the structural failures. A column failure
on the Simi Valley Freeway, for example, revealed that the steel reinforcement deformed into a bas-
ketlike arrangement of severely bent rods. Some reports suggested that this failure mode was caused
by rocking horizontal motion, which caused some compressed bars to bend outward, thereby push-
ing off their concrete cover, which in turn so weakened the column that further rods bent outward
in the subsequent earthquake motion. It is also possible that the failure resulted primarily from the
vertical bouncing of the bridge deck on the column, also causing it to swell outward, again with the
rods pushing out the concrete, but more or less simultaneously and in a more symmetrical fashion.
Establishing exactly how the failure mode originated and progressed was further complicated by
the fact that the overpass in question crossed the streets below at an angle, thus leading to a design
in which some lines of columns were skewed to the roadway. The various combinations of hori-
zontal and vertical ground motion, combined with the relatively complex geometry of the problem,
cause the analysis of the structure to proceed in different directions depending on the assumptions
made at the outset. With the understandable pressures to repair and reopen the freeways as soon as
possible, the physical evidence of the failure modes of the various columns could be retained only
in photographic records. Although these help to confirm or refute the various failure scenarios that
might be offered, the loss of the failed artifact itself always leaves some room for debate about what
exactly happened.
Regardless of the precise sequence of events that causes failures like those of the Simi Valley
Freeway columns, it is possible for engineers to retrofit similar structures. What is clear, even from
photographs, is that the steel reinforcing bars in the concrete columns deflected outward. Even be-
fore the Northridge earthquake, the California Department of Transportation had instituted a pro-
gram of retrofitting bridge columns with steel jackets capable of containing the outward pressures
that improperly reinforced concrete cannot. Bridges that had been so retrofitted appear to have sur-
vived the earthquake intact. However, the Simi Valley Freeway, which was built just after the 1971