International Concrete Abstracts Portal

Describes a small scale flexure test for the determination of cement- aggregate bond strength. Cylindrical test specimens were prepared by drilling cores in a perpendicular direction through slabs of rock against which mortar had been cast. A special casting procedure eliminated many sources of experimental variation and allowed the bond strengths of different mortars and rock types to be compared directly. Long term tests were conducted by coring the mortar/aggregate slabs at a number of curing times and coefficients of variation of 5 to 10 percent for bond and mortar strengths were obtained. The effect on cement-aggregate bond strength of partial cement replacement by silica fume was evaluated for a number of aggregate types. For siliceous aggregates (glass, obsidian, and quartzite), bond strength was increased significantly by the addition of silica fume; failure tended to occur away from the interface particularly in long term tests. For carbonate rocks (limestone and dolostone), similar bond strengths were obtained at seven days with and without the addition of silica fume. At later ages, silica fume interfered with strengthening of the cement-carbonate rock interface and lower bond strengths were obtained. For specimens not containing silica fume, bond strength increased more rapidly to glass and obsidian than to quartzite, which showed essentially "inert" behavior. This was tentatively attributed to strengthening of the transition zone by a pozzolanic mechanism involving reactive silica from the aggregate. A marked reduction in bond strength occurred with glass specimens containing boosted alkali content. This was attributed to alkali- silica reaction at the interface and was suppressed by the addition of 15 percent silica fume.

Fracture of the interfacial zone between a fiber and the cementitious matrix plays a significant role in the mechanical behavior of fiber reinforced cementitious composites. For better understanding of debonding characteristics of a fiber in the composites, the behavior of the extension of cracks along the interface was examined under the fluorescence microscope. The correspondence between the features of fracture zones obtained by the microscopic study and the fracture toughness for the interfacial zone is discussed in this paper. Examinations under the microscope revealed that the debonding and the extension of interfacial cracks were not caused by a simple shear failure at the actual interface, but were accompanied by local failures over relatively wide regions surrounding the steel fibers. The incorporation of silica fume resulted in the reduction in areas containing local failures along the interface. Fewer local failures in the interfacial zone in the steel fiber-silica fume-bearing cement composite were reflected by the decrease in the fracture toughness for the interfacial zone.

In many practical engineering applications of fiber reinforced concretes (FRC), fibers are subjected to significant lateral loading. The lateral stress may have a significant effect on fiber debonding and pullout, thus affecting the performance of FRC. In this investigation, a novel experimental set-up was developed to carry out fiber pullout tests under various levels of lateral compression. Interfacial properties for steel fiber in mortar were derived from the measured fiber load versus displacement curves, based on a unified fiber debonding theory. As expected, the interfacial friction at the onset of sliding increases with applied compressive stress. Surprisingly, the pre-sliding interfacial resistance (which can be either the interfacial strength or the interfacial fracture energy, depending on whether interfacial debonding is strength or fracture governed) is also found to increase with lateral compressive stress. Also, with higher compression, there is a more rapid decay of interfacial friction when the fiber is sliding out of its groove. As a result, while lateral compression can significantly increase the peak pullout load, the increase in total energy absorbed during the pullout process is much less drastic.

A new method to predict the debonding behavior of fiber-matrix interface has been proposed by applying the principles of the micromechanics of inclusion and fracture mechanics. The validity of the mathematical model is further verified by uniaxial tension tests carried out on steel fiber reinforced cementitious composite specimens by employing a digitally controlled closed-loop MTS testing machine. It was demonstrated that the debonding occurs before the bend over point; the debonded lengths are largely influenced by the sequence of the occurrence of transverse matrix cracks and the loading stage. A stable growth of debonding was observed in the investigation. The measured debonded lengths were compared with the theoretical prediction of the proposed model. A reasonable agreement was observed.

The bond between reinforcing bars and concrete under impact loading was studied both experimentally and by the finite element method. The experiments consisted of pullout tests and push-in tests, under three different types of loading: static, medium rate, and impact. Different concrete strengths (normal and high), types of fibers (polypropylene and steel), and fiber contents were considered. The study focused on the bond-slip relationships and the fracture energy in bond failure. The experimental results were compared with those obtained by the finite element method, in which a special "bond-link element" that was able to transmit both shear and normal forces was adopted to model the connection between the reinforcing bar and the concrete. It was found that higher loading rates, higher concrete compressive strengths, and the addition of steel fibers had significant effects on the bond resistance, the fracture energy, and the bond stress-slip relationship, especially for the push-in case. Reasonably good correspondence in the results between the two methods was also found, and a bond-stress-slip relationship under high rate loading could be established analytically.