The Northridge Fractures: Are We Learning the Right Lessons?

After the Northridge earthquake in January 1994, engineers found problems with special moment-resisting frames. There were multiple facture modes, and typical damage involved primary crack initiation at the juncture between the surface of the lower beam flange groove weld root pass and the surface of the column flange. At least 100 buildings with such damage had been discovered when this article was written, and more will possibly be detected as investigations continue. Many engineers and public officials have included poor welds as part of the problem, and the author believes this is an incorrect conclusion. He discusses steel properties, triaxality, strain and stress in connection with the moment resisting frames. He also suggests that current testing methods need to be changed. Welding is always the scapegoat when failures occur, he states, and offers six arguments to refute myths about welding. He says that making welds more complicated and costly will not prevent damage during another earthquake. Changing seismic codes in regard to welding should be done with great care. No code provision can prevent a connection from cracking. Finally, the presence of a code does not imply compliance; that is the responsiblity of the client and the contractor.