Research over the last decade has shown that pulse-type earthquake ground motions that result from forward directivity (FD) effects can result in significant damage to structures. Three typical post-1990 Washington State Department of Transportation (WSDOT) monolithic concrete bridges were chosen to investigate their nonlinear response to FD ground motions (FDGMs) and non-FDGMs. Results showed that significant seismic damage may occur if the structural response is in tune with the period of the velocity pulse of the FDGM. This velocity pulse is a
result of fault propagation effects in the near-fault, and occurs when the direction of slip and rupture propagation coincide. The period of the velocity pulse is proportional to the magnitude of the earthquake. The severity of the demand is controlled by the ratio of the pulse period to bridge fundamental periods. As a consequence of this, damage in a bridge with moderate periods (T=0.1s to 1.0s) may be more significant in smaller magnitude earthquakes where the pulse period is closer to the fundamental period of the structure. This was the case for both the MDOF and SDOF analyses of all three bridges in this research. The results showed also that the occurrence of high PGA and/or PGV is only one of several conditions that can cause high demand on the bridges.
Of the three bridges considered, all typical concrete overpasses ranging from 50 m to 91 m in length, all generally survived the earthquake motions with only minor damage to their columns. However, column flexural failure was predicted for one model when subjected to two of the forward directivity ground motions. SDOF bridge models for preliminary analyses were found to yield slightly unconservative base shears and displacements compared to that of the full bridge models under non-FDGM. For FDGM, the results of a simple SDOF bridge model ranged from very conservative to slightly unconservative. Therefore, nonlinear SDOF analyses are specifically not recommended in the case of FDGM since the results were not consistent. A more detailed MDOF model should be used to assess bridge seismic performance so that SSI and the interaction of the longitudinal and transverse responses of the bridges can be included, particularly if a performance based design or assessment of the bridge is required.