This paper describes a theoretical and experimental analysis designed to characterize the initial branch of the bond-stress/slipping curves (5-s) for normal-strength and high-strength concretes. The theoretical analysis is used to interpret the results of experimental trials on reinforced concrete ties prepared with class 50 and 100 lMPa concrete mixtures and submitted to tensile forces without inducing any yield in the bar. The purpose of the investigation was to study any changes in bond behavior (over the limited range of slipping values considered) due to the better mechanical features of the 100 MPa concrete, and thus contribute to a better understanding of how high-strength reinforced concrete elements behave in a ser-viceability state.

This paper reports on research in progress conducted at the American University of Beirut to evaluate the effect of silica fume on bond and anchorage of reinforcement in high performance concrete (HPC) structures. The program includes testing the effect of a wide range of variables on the bond strength of beam bar splices and bars anchored in pullout specimens. Results of the first phase of the research program have been analyzed. Ten beam specimens were tested. Each beam was designed to include two bars in tension, spliced at the center of the span. The splice length was selected so that bars would fail in bond, splitting the concrete cover in the splice region, before reaching the yield point. The beams were loaded in positive bending with the splice in a constant moment region. The variables used were the percentage replacement of cement by silica fume and the casting position. Test results indicated that replacement of 5 to 20 percent of the cement by an equal weight of silica fume resulted in an average 8 percent reduction in bond strength regardless of casting position.

The bond between reinforcement and concrete should ensure high structural stiffness and small cracks in the serviceability limit state, generate small splitting forces and allow full utilization of the reinforcement ductility in the ultimate limit state. While bond behavior at service load and splitting behavior has been investigated intensively, bond behavior at large inelastic steel strains is not known very well. Therefore, in this paper the contribution of concrete between cracks at inelastic steel strains is investigated numerically based on a rational mechanical model and using realistic constitutive material laws. The model predictions agree rather well with a large number of test results. According to the results of the parametric study, after steel yielding the ratio of mean steel strain to the steel strain at the crack is mainly influenced by the reinforcement percentage and the shape of the steel stress-strain curve. It is much lower than at service load. Due to this lower ratio of mean steel strain to steel strain at the crack, the rotation capacity of plastic hinges and thus the structural ductility is reduced significantly and may be very low if reinforcement with low ductility is used. Therefore an optimization of bond seems to be necessary. Corresponding extensive numerical and experimental studies are under way in Germany.

Characteristic results of more than 100 cyclic pull-out tests are presented including various load histories (simulating realistic load spectra) like random loading or variable amplitude loading with increasing or decreasing tendencies in addition to the constant amplitude loading with different amplitudes. Slip measurements are compared to acoustic emission measurements. Repeated loading produces a progressive deterioration of bond caused by the propagation of micro-cracks and progress of micro-crushing in concrete. Deterioration of bond may be observed by measuring slip or acoustic emission events. It is quantitatively shown that the actual slip is significantly influenced by the load history: maximum and minimum levels of the repeated load, type of amplitude (constant or variable), frequency, sequence of amplitudes and number of load cycles, respectively.

The results of the first major reevaluation of reinforcing bar geometry in the United States in nearly 50 years is described. The study involves experimental and analytical efforts designed to broaden the understanding of factors that control bond strength, improve the development characteristics of reinforcing bars, and develop practical design expressions that more accurately represent development and splice strength than existing expressions. The research has established that deformation pattern has little effect on the bond strength of uncoated bars that are not confined by transverse reinforcement. Deformation pattern, however, as represented by the relative rib area, does have a major effect on the bond strength of bars that are confined by transverse reinforcement. Increases in relative rib area, obtained with either higher ribs, closer ribs, or a combination thereof, result in an improved bond strength for confined bars. The study has also established limits on how closely ribs can be placed without resulting in a pullout failure. High relative rib area bars provide a reduction in development/splice length of 20 percent for all coated bars, independent of the presence or absence of confining transverse rein-forcement. Based on the experimental work, expressions are developed that accurately characterize development/splice strength. In the development of the expressions, f’,t/4 is shown to be superior to f$2 for characterizing the contribu-tion of concrete strength to bond. The resulting design expressions are accurate for compressive strengths between 2500 and 16,000 psi (17 and 110 MPa). The most accurate representation of the effect of transverse reinforcement on bond strength includes parameters that account for the number of transverse reinforcing bars that cross the developed/spliced bar, the area of the transverse reinforcement, the number of bars developed or spliced at one location and the relative rib area, and size of the developed/spliced bar. The yield strength of the transverse reinforce-ment does not play a measurable role. Practical comparisons illustrate reductions in splice lengths of 20 percent for conventional bars and 30 percent for high relative rib area bars compared to current requirements in AC1 3 18-95.