Impact of new advances in seismology and geotechnical engineering on structural design since the Northridge earthquake

The Northridge earthquake has brought a decade of change in the practice of engineering for all structures, and for tall buildings in particular. With the earlier Loma Prieta earthquake in Northern California, there was already some impetus and momentum to improve seismic safety and reduce risks from seismic hazards. However, the extensive damage and economic losses from the Northridge earthquake fueled greater public will to meet the challenges and provide programs for the reduction of losses and human casualties. In the geotechnical engineering discipline, many of the advances involved a more consistent application of seismic risk across the United States in dealing with strong ground motions. Initiatives were set in place to develop seismic hazard models of the possible earthquake sources that affect the country and to use these models in developing design criteria for seismic design in all regions of the United States, not just in California. These ground motion estimates are now accounted for in the latest design guidelines and building codes; however, the treatments in these documents is simplified and not always tied to reality. Ground motions are complex and advances are being made to understand the complexities and explain the variations. Great advances in seismology have occurred in the 10 years since Northridge, and these advances will be incorporated into building design and analysis with time. Other advances have been made in the area of defining seismic hazards that result from ground failures in California. The Northridge and earlier earthquakes demonstrated that ground failures (especially related to soil liquefaction and landsliding) can present a significant risk and result in large losses. In the decade since the Northridge earthquake, structural analysis has advanced with leaps and bounds due to advances in computer software and hardware. Complex three-dimensional analyses in the time domain once relegated to only large mainframe computers can now be performed on desktop computers in most structural engineering offices. To model the seismic behavior of structures more accurately, better modeling of the soil–structure interface as well as the soil has become necessary. Models incorporating the appropriate soil behavior provide for better understanding of the forces and displacements of the structures under seismic loading and possibly avoid overconservatism by not accounting for the soil behavior. Copyright © 2004 John Wiley & Sons, Ltd.