1903 1 Doctoral Student, University of California Berkeley, 721 Davis Hall, Berkeley, CA 94720-1710,USA,naito@ce.berkeley.edu 2 Assistant Professor, University of California Berkeley, Department of Civil and Environmental Engineering, mosalam@ce.be 3 Professor, University of California Berkeley, Department of Civil and Environmental Engineering, moehle@eerc.berkeley.ed EVALUATION OF REINFORCED CONCRETE BRIDGE JOINTS Clay J NAITO 1 , Khalid M MOSALAM 2 And Jack P MOEHLE 3 SUMMARY Damage in recent earthquakes has resulted in increasingly conservative design of reinforced concrete beam column bridge joints. Current recommendations produce joint detailing which result in high levels of congestion of steel reinforcement and extreme difficulties in construction. Currently, a research project at the University of California, Berkeley is focusing on the development of a rational model to describe joint response to earthquake loading, a general design procedure for bridge joints, and a method of incorporating headed reinforcement into the design to improve joint constructability. In order to accomplish the project goals, experimental investigations into the response of bridge joints to earthquake loading are being conducted. The investigation consists of quasi-static laboratory testing of eight reduced scale models of bridge joint components. The primary goal of this phase is to improve the understanding of joint behavior and to determine how conventional and headed reinforcement can be better utilized in improving joint response. Results from the first phase of the experimental study have found that California design strategies produce joints that are capable of supporting the formation of a column hinge mechanism, although at the expense of constructability. Headed joint transverse reinforcement proves to be a viable means of reducing construction difficulties without any decrease in joint performance. Some of the secondary issues being investigated are the effectiveness of strut and tie modeling and the effect of controlling slip of joint reinforcement. A parallel computational study of the first phase is being conducted; three- dimensional finite element models are being developed for further understanding of the joint behavior. Preliminary results and techniques for developing an effective three-dimensional finite element model are presented. INTRODUCTION Due to the catastrophic failure of bridge systems in the recent earthquakes of Loma Prieta, Northridge, and Kobe, there has been a great effort directed towards safer civil infrastructure in the United States and Japan. This has taken the form of retrofitting or strengthening existing bridges and increasing the design requirements for new bridge systems. While strengthening techniques and design requirements for beams and columns are well established [Park 1975], designing or evaluating the connection between the two is still in contention. The current methods of joint design are based on either a two-dimensional evaluation of the flow of stresses within the joint or through strut and tie methods which often neglect compatibility in their formulation. In general, reinforced concrete bridges are subjected to multi-directional ground motion. Therefore, response of bridge beam-column joints is predominantly three-dimensional (3D). Accurate evaluation of existing generic bridges and development of general design requirements for beam-column joints may require the use of 3D-models which take into account compatibility, equilibrium and the constitutive properties of the system. Many computational methods exist for modeling systems in three-dimensions, however, how well these models reflect the actual behavior of reinforced concrete bridges, particularly systems subjected to seismic loading, is not clear.