The destructive effects of velocity pulses were first recorded during the 1971 San Fernando, California, earthquake. However, the actual near-field behavior of a rupturing fault remained undefined until the 1994 Northridge, California, and 1995 Kobe, Japan, earthquakes. A large amount of data were gathered from
structural damage and strong ground motion recording stations, allowing engineers to more fully understand the near-field seismic environment. This prompted investigations into the performance of bridges, which often have natural periods close to those of recorded velocity pulses.
For this report, four half-scale beam to column bridge joints were tested with a velocity pulse loading. Two were representative of construction before 1971, and two were typical of joints designed today. After the pulse, the joints were loaded with a typical cyclic displacement history. These joint tests were compared to joints tested previously without a pulse and to the PEER Bridge Performance Levels as outlined in Hose (1999).
This research determined that velocity pulse loading caused considerable damage to the pre-1971 joints, exhibiting Performance Levels 1–5 and loss of strength beginning during the pulse loading. The current desi gn specimens exhibited minimal damage (Performance Levels 1– 3) during the pulse and continued to perform well during the subsequent cyclic loading. Performance Level 5 and degradation began at a displacement ductility of six.
Both designs exhibited higher than predicted strengths during the velocity pulse. By modifying the material properties for the effects of high strain rate, it was shown th at the higher strengths could be accounted for analytically.
Full List of PEER Reports: click here.