This report documents and presents results of a study to determine time-dependent behavior and relevant design criteria for simple-span precast, prestressed bridge girders made continuous. A questionnaire was used to determine current practice. Creep and shrinkage tests of steam-cured concrete loaded at an early age were made. Computer simulations were used to investigate the effects of time-dependent material behavior and variation in design parameters on the effective continuity for live load plus impact. The findings suggest that positive moment connections in the diaphragms at the piers are not required and provide no structural advantages. The findings also suggest that effective continuity for live load plus impact can vary from 0 to 100% dependent on the design parameters and timing of construction. Computer analyses were also used to determine an upper limit for the amount of negative moment reinforcement over the supports to insure full moment redistribution and attainment of maximum bridge strength. New computer programs were developed for simplified analysis to determine time-dependent effects and service moments. Recommendations for design procedures were presented and design examples given.

Precast prestressed concrete girders have been used in bridge construction for several years in the United States. In general, these types of bridges have performed well in service conditions. Although the connections over the supports provide sufficient continuity for dead loads, which cause negative moments, there has been some concern about the ability of the connections to provide continuity for the positive moments caused by live loads. Cracking has been observed at the bottom of the diaphragm at the connection and this raises concern about the connection's performance. The National Cooperative Highway Research Program developed Project 12-53 to study the effectiveness of these connections. The focus of this research deals with the experimental testing part of project 12-53. Six specimens that represent portions of full-size bridges were tested under monotonic and cyclic loads to evaluate their performance after cracking occurs at the bottom of the diaphragm. Each connection was fatigued until the connection was considered to have failed. The bent strand and bent bar embedded connections performed well under service moments. Under negative moment all of the specimens performed well. Even after failure of the specimens they each were able to resist the negative moment. This was due to the bearing of the diaphragm against the beams. The crack closes and at this point there is no difference between the specimens. Under positive moment the embedded specimens were stiffer and the failure was more gradual than that of the not embedded specimens. The embedment adds to the stiffness of the specimens. The engagement of the diaphragm in the embedded specimens helps to prolong the failure. The warning signs of failure are more prominent in the embedded specimens. The bent bar specimens are stiffer than the bent strand but the failure occurs more rapidly. After all six of the short-length specimens are tested, full-size specimens will be tested. At that point, results from all testing will be compared and discussed providing conclusions about the behavior of positive moment connections. Also, recommendations are expected as to the design and construction of positive moment connections in precast/prestressed bridge girders made continuous.

Continuous precast/prestressed bridges are formed by placing the girders end to end and then pouring a slab and intermediate diaphragms. The girders carry the dead load of the structure as simple spans but are continuous for live loads. These connections provide sufficient negative moment continuity, however their effectiveness in providing positive moment continuity is questionable. Cracking has been observed in bridges of this type due largely to time dependent effects and this has raised questions regarding the bridges performance. The National Cooperative Highway Research Program (NCHRP) Project 12-53 was developed to study the effectiveness of the connections between precast/prestressed bridge girders made continuous. This thesis presents a portion of the experimental work for project 12-53. Six specimens, representing portions of full-scale bridges, were tested and fatigued to evaluate their performance after cracking has occurred. The loading procedure represents extreme loading events occurring after cracking at the diaphragm has occurred and the structure is subjected to normal traffic loads. Data from each specimen was analyzed and the results were compared for each connection type. The specimens were designed to evaluate the effect of the use of bent bars versus bent strands, embedment, and the addition of web bars and/or stirrups to a bent bar connection with regard to performance. It was found that all connection types would provide adequate negative moment continuity. Each detail provided differing levels of crack control; however cracking did not significantly affect the stiffness of the specimen. The best method for providing better crack control and a more efficient section is to embed the girder into the diaphragm. There is little difference between bent bar and bent strand connections and both seem to offer sufficient crack control. Additional stirrups in the diaphragm slightly increase the stiffness of the connection. The main advantage to adding stirrups would be the ductility that is added to the connection. The addition of web bars was very effective, but is not recommended. Although the addition increases the strength of the connection, it causes cracking of the ends of the girders and is difficult to construct.

Introduction and Research Approach -- Findings -- Interpretation, Appraisal, and Application -- Interpretation, Appraisal, and Application -- References -- Appendixes.

At head of title: National Cooperative Highway Research Program.