International Concrete Abstracts Portal

From pull-out and push-in tests on specimens with short embedment length an empirical relation has been derived, which describes the local bond stress as a function of the local slip and steel stress change. With the help of this bond model the transfer length and the bi-linear relationship for the development length of a pretensioned strand (ACI Building Code 1989, CEB-FIP Model Code 1990) is simulated. It is also used to indicate the influence of strand yielding on the development length. For the estimation of the concrete cover and strand spacing required to prevent the occurrence of visible bond splitting cracks the response of the concrete to the radial displacement of the strand-to-concrete interface is analyzed by means of a so-called thick-walled-cylinder model. The radial interface displacement consists of transverse deformation of the strand coupled with steel stress change (Poisson effect) and wedging action caused by the shape of the strand (lack-of-fit effect) and surface roughness. Besides the section geometry, this model takes into account the softening behaviour of concrete loaded in tension. It is used to explain the influence of various parameters such as concrete cover, strand spacing, strand diameter and concrete strength on the bond properties of strand.

In the present paper, the role of the confining forces both on bond capacity and on splitting crack opening under service loads is shown. In par-ticular, theoretical relationships between the transverse reinforcement area, the bond strength, the splitting crack opening and the stirrup stress are pre-sented. The theoretical predictions are contrasted with some experimental results and a discussion on the values provided by building codes is presented. The results show the strong influence of transverse reinforcement whose confining force is expressed by the stirrup index of confinement, which gov-erns bond behavior and is suitable for design. The comparison with several experimental results showed a good agreement between theory and tests. Adequate values of Q are also required to control splitting crack opening under service loads. For common amounts of transverse reinforcement, the splitting crack opening can be larger than one half of the flexural crack, which could be unacceptable for structural durability.

Bond properties are usually described by empirical relations that are based on pull-out tests of bars with short embedment lengths cast in low to medium strength concrete. The limited validity of these formulations is recognized and their applicability in structural analyses is reconsidered. Results of two test series with various confining conditions and concrete strengths provide the basis for the derivation of a new bond model for ribbed ibars. Pull-out bond failure in confined concrete and splitting bond failure in unconfined concrete have been studied. Steel yielding is found to have a iconsiderable influence on bond strength. Significant differences in bond of NSC and HSC are confirmed. An analytical bond model for ribbed bars is developed. It is based on the confining capacity of the concrete surrounding the bar. This is evaluated with a /thick-walled-cylinder model, from which the relation between the radial displacement and the radial compressive stress at the steel-to-concrete interface i s derived. The radial displacement at the interface is linked to the slip of the ribbed bar, distinguishing between the two modes of bond failure: pull-out and cover splitting. The model takes into account concrete toughness and bar contraction, also after yielding. Verification of the model against selected experimental results reflects the potential of the model to be used in a broad range of applications. In its present form the model is used for the analysis of isolated bond problems using a one-dimensional finite difference approach, but its application is also considered to count for three-dimensional deformations in FE codes that treat the bar as a one-dimesional element.

When ribbed bars are anchored in linear structural members, the bond-slip behavior and the anchorage capacity is strongly influenced by splitting I cracks. Many factors influence the formation of the splitting cracks, among others .,the anchorage length, the concrete cover, the bar spacing and arrangements, confinement from stirrups, flexural and shear cracks in the vicinity of the t anchorage region, transverse pressure from support bearings, etc. These ’ parameters often interact in a complex manner, and common design methods for anchorage regions are derived from empirical evaluations of test data and are often strongly simplified. The present study was carried out with the aim of , studying the anchorage behavior of ribbed bars in structural members of high strength concrete and to check the applicability of some common design rn.ethods to these new materials. The influence of concrete type, normal or high-strength concrete, and various detailing of the node regions was examined. The tensile force in the active end of the anchorage zones was evaluated from steel strain measurements and was compared with predictions by means of strut and tie models. These models were found to consider the effect of inclined cracks in an appropriate and consistent way. The observed anchorage capacity was compared to some common design methods. It was found that the methods, to a considerable degree, were unable to reflect the real behavior. Further improvement and development of design and analytical tools is required.

Safety concerns and a lack of test data on bond capacity of deformed reinforcing bars embedded in high-strength concrete are among the reasons for the AC1 318 building code imposing an arbitrary limitation of 10,000 psi (69 MPa) when calculating tension development and splice lengths. This limitation was first introduced in the 1989 revision of the AC1 3 18 building code. In an attempt to evaluate the impact of this limitation and develop provisions for its removal, an investigation was carried out at the University of Nebraska-Lincoln, partial result of which will be presented in this paper. Results of the investigation are used to discuss the differences that exist between normal and high strength concrete, develop hypotheses to explain these observed differences, and suggest alternatives for removal of the current concrete compressive limitations existing in the ACI 3 18 building code for calculating tension development and splice lengths. In this paper high strength concrete is defined as concrete with compressive strength exceeding 10,000 psi (69 MPa).