Keywords: reinforced concrete bridge, scale model, earthquake, nonlinear, finite element analysis.

A quarter-scale, four-span, 33.5 m long asymmetric conventional reinforced concrete bridge model was subjected to ground motion based on a 1994 Northridge, California earthquake record. The bridge received biaxial horizontal motions at the bases of the three two-column bents from three separate shake tables and was simultaneously subjected to longitudinal motions from actuators attached to the abutment seats. This paper presents the nonlinear computer modelling of the bridge. The analyses include the interaction between the bridge deck and abutments, the in-plane rotation of the bridge, and the corresponding transverse bent displacements.

A three-dimensional finite element model of the bridge was developed in OpenSees. The bridge columns were modelled with nonlinearBeamColumn element with fibre sections composed of Concrete01 and Hysteretic uniaxial materials for concrete and steel, respectively. Column bond-slip moment-rotation behaviour was represented by zeroLength elements with Hysteretic uniaxial material placed at the column ends. The superstructure was assumed to stay within the linear-elastic range; thus, elasticBeamColumn elements were used for the bridge deck.

Seven sets of motions were applied to the bridge with increasing amplitudes. The largest peak target amplitudes (during the last two tests) were 1.0g and 1.2g for the transverse and longitudinal directions, respectively. In the computer model, the biaxial earthquake motions were imposed to the column bases. The longitudinal actuator motions were imposed on both the north and the south abutments. These motions were obtained from a pre-test response history analysis of the bridge that included an abutment spring representing the force-displacement relationship of a typical seat-type abutment.

The abutment-deck interaction was modelled by a set of two contact elements at each abutment. Furthermore, realistic abutment gap openings were included that were based on the start and the end of each test run. Sensitivity analyses were conducted. The results of the friction sensitivity analysis predicted high variability in the calculated transverse residual displacements. Using the average of the residual displacement square root of the sum of the squares (SRSS) of the cases considered, it appears that imposing friction at the NE or the SW or a combination of the NE and SW produced transverse bent displacements that were closer to the experimental results. More abutment contact with the NE and the SW corners are consistent with the observed deformed shape of the bridge deck.

Changing the localized damping at abutments was explored. With 5 percent damping at the abutment, the following changes were observed: (a) the best previous response was improved by about 3 percent; (b) less erratic behaviour was observed in the transverse displacements; and (c) a better pattern was observed in SRSS values with increasing deck corner friction values.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £130 +P&P)