Report No.: CCEER-12-3

Title: Seismic Performance of  Circular and Interlocking Spiral RC Bridge Columns Under Bidirectional Shake Table Loading

Authors: Arias-Acosta, J. and Sanders, D.

Date: August 2012

Sponsoring Agency: National Science Foundation (NSF)

Performing Organization:
Department of Civil Engineering/258
University of Nevada, Reno
Reno, NV 89557


Under seismic excitations reinforced concrete bridge columns (RCC) are subjected to combinations of forces and deformations. These complex actions are caused by spatially-complex variation of earthquake ground motions, the bridge structural configuration, and the interactions between input and response characteristics. The seismic behavior of RCC may be seriously affected by these complex actions, and that in turn influences the performance of bridges as essential components of transportation systems.

To study the impact of bidirectional ground acceleration on the seismic performance of circular and oblong sections (double interlocking spirals), four large-scale cantilever-type RCC specimens were designed and tested on the bidirectional shake table facility at the University of Nevada, Reno (UNR). As part of the study, a unique inertial loading system named the Bidirectional Mass Rig (BMR) was developed to allow shake table testing of single RCC under biaxial ground motions. Pairs of circular and interlocking RC specimens were subjected to different levels of biaxial real time earthquake motions. Within each pair, one specimen had asymmetric distribution of masses on the BMR to induce more torsion. The performance of the specimens was assessed in terms of strength, deformation, ductility and failure mode.

The seismic performance of each pair of specimens was similar and was controlled by the biaxial effect of bending with small influence of shear deformations. The RCC exhibited stable and ductile behavior, and without collapse, under repetitions of earthquakes with spectral amplitude equal to or larger than the design and maximum considered earthquakes in California. For the sections and ground motions used, the biaxial interactions affected mostly the seismic performance of the columns along the direction where the small component of the earthquake was applied, showing reductions in the lateral capacity as predicted by moment-curvature analyses. It was also observed that the asymmetric mass configuration used for specimens C2 and I2 only induced low values of torsion on the columns with measured values of the torque to bending ratio below 20%.

An analytical investigation using OpenSees software was conducted to develop and validate analytical models that can reasonably predict the seismic behavior of RC columns subjected to biaxial earthquake loading. The results show that the modeling of the specimens with a nonlinear beam-column element with fibers (beam-with-hinges element), hysteretic material models with strength degradation (Concrete07 and Reinforcing Steel materials), bond slip and viscous damping (stiffness proportional-only) leads to the best estimation of the measured performance

In order to investigate the impact of biaxial loading on the seismic response of columns, the analytical models were subjected to the combined effects of axial loads, either unidirectional or bidirectional excitations and P-delta effects. The results indicate that circular and interlocking columns designed according to the Caltrans BDS and SDC generally behave well, even under large levels of biaxial earthquake loading. From the analytical results it was observed that for small amplitude earthquakes (before yielding) no major differences are observed in the response of columns under unidirectional or bidirectional excitations. After yielding the biaxial excitations resulted in a reduction of the capacity of the columns, increase of lateral displacements and more accelerated stiffness degradation compared to unidirectional excitation. It was also found that for near-fault earthquakes with forward directivity effects, the peak bidirectional displacements are comparable to the peak unidirectional displacements computed using the strong component of the earthquake, and to the component displacement calculated from the individual uniaxial responses combined using the square root of the sum of squares (SRSS) rule.