Critical Cracks

Why Did So Many Steel Joints Fail in the Northridge Quake? Our Engineers Seek the Answer

by Robert Sanders

Seismic engineers tested a full-scale welded steel joint Feb. 16 in search of clues to why so many steel-frame structures cracked in the Jan. 17, 1994, Northridge earthquake.

The test in Davis Hall strained the joint until it failed in an effort to determine why hundreds of beam-column connections cracked during the 6.8 magnitude Northridge quake.

In the second of a series of tests, a 10-foot horizontal beam was pushed and pulled to simulate the horizontal motion expected in a strong earthquake with extreme displacements up to 6 inches.

In a survey last year of more than 100 steel-frame buildings in the Northridge area, as many as 75 percent were found to have cracks in their steel frames, primarily at the joint between steel beam and vertical steel column.

Several of the buildings remain vacant, dozens have undergone expensive repairs, and one was demolished.

Though welded and riveted joints like these have long been standard building practice in the state, the surprising findings after the Southern California quake are forcing engineers around the world to reassess building codes.

The main reassessment is being conducted by a nationwide consortium directed by Civil Engineering Professor Stephen Mahin.

Funded by $2.3 million from the Federal Emergency Management Agency (FEMA) and the California Office of Emergency Services, it draws upon the expertise of the nation's leading seismic engineers and researchers.

Its formal name is SAC, an acronym for the three organizations in the alliance: the Structural Engineers Association of California, the Applied Technology Council, and the California Universities for Research in Earthquake Engineering.

SAC's goal is not only to find out why these steel joints failed but to come up with an inexpensive way to fix the problem and recommendations for improved building codes for future construction.

Berkeley engineers are performing an initial series of three tests to try to reproduce the kinds of cracks discovered in buildings near the Northridge quake. The first such test produced a crack in a flange of one of the vertical columns that resembled cracks seen in a third of the damaged buildings in Northridge, said graduate student Bozidar Stojadinovic.

The flange cracked with a sharp pop after a 3-inch deformation of the cantilever beam.

The second test is an attempt to reproduce this break.

If all goes well, within a month or two the team of engineers will test various repair options, said Egor Popov, professor emeritus of civil engineering and one of the leaders of the Berkeley project.

The state hopes to release interim guidelines in May that will suggest reliable and cost effective procedures to identify and inspect suspect buildings, to evaluate their safety and expected performance in an earthquake, to repair damaged buildings and to design new steel structures.


Copyright 1995, The Regents of the University of California.
Produced and maintained by the Office of Public Affairs at UC Berkeley.
Comments? E-mail