116 B. Williams et al. dynamic behavior of high-strength concretes under triaxial loading conditions. Experiments in this preliminary study were all performed on BBR9, which has an unconfined compressive strength of 130 MPa for cast cylindrical specimens with a diameter of 75 mm and a height of 150 mm [7] and 140 MPa for cored cylindrical specimens with a diameter of 50 mm and a height of 114 mm [8]. Analysis In quasi-static experiments on BBR9, brittle, quasi-brittle and ductile failure modes were observed under confinement pressures of 10 MPa, 50 MPa, and 100 MPa, respectively. Therefore, these pressures were also selected for investigations on the triaxial Kolsky bar. All tests were conducted on cored cylindrical BBR9 concrete specimens with a diameter and height of 25mm. Since the triaxial Kolsky bar experiments presented here utilize pressurized fluid (kerosene) to provide hydrostatic confinement pressure, it is of the utmost importance that the specimen can be sealed from the surrounding fluid. This can be quite challenging since there is a sharp corner where the 25-mm-diameter specimen meets the 50-mm-diameter bar. In a previous study on borosilicate and soda-lime glass, a heat shrink tubing with a coat of J-B WaterWeld epoxy was used to surround the entire specimen, therefore providing the required sealing [5]. However, as the failure mode in concrete becomes more ductile, it was suspected that the rigidity of epoxy could skew experimental results. Therefore, the specimens in this study were prepared in four distinct steps as shown in Fig. 20.1: (1) Specimen is positioned and then seated with an axial prestress of 5 MPa, (2) a thin line of WaterWeld epoxy is applied at the sharp corner of the specimen/bar interfaces and then shaped into a small radius, (3) a translucent latex triaxial membrane is stretched over the specimen while overlapping the initial epoxy application, and (4) an additional layer of epoxy starts at the edge of the bar surface and extends over the latex membrane. This method allows the majority of the specimen to only be covered by the flexible latex layer. Additionally, tests were performed while the epoxy was still malleable (prior to full set). Experimental data were collected using strain gauges with a Wheatstone bridge circuit and recorded on an oscilloscope. An example loading pulse is shown in Fig. 20.2 for a test conducted with a preloaded hydrostatic pressure at 50 MPa. For this experiment, an annular annealed copper pulse shaper was used with a thickness of 1.1 mm and outer and inner diameters of 19.0 mm and 6.3 mm, respectively. An additional annular PTFE pulse shaper with a thickness of 0.26 mm and outer and inner diameters of 11.1 mm and 9.5 mm, respectively, was added to further smooth the loading pulse. Specimen stress was calculated at the front and back face verifying that dynamic stress equilibrium was achieved. However, the strain rate was increasing slightly as the specimen failed. Additional pulse shaping is required for future experiments to ensure that a constant strain rate can be achieved at each specified level of confining pressure. The experimental data was processed using the methods described by Chen and Song [9] for triaxial Kolsky bar experiments where the offset in static preload is shifted back to zero for dynamic analysis. The resulting stress and strain data are presented in terms of principal stress difference and axial strain as shown in Fig. 20.3. For clarification, principal stress difference refers to the difference between axial stress (σ1) and radial stress (σ2)whereσ1 =σ2 during the hydrostatic loading phase. Note that σ2 is assumed to remain constant throughout the duration of dynamic loading. As the confinement pressure Fig. 20.1 Triaxial Kolsky bar specimen sealed with a translucent latex triaxial membrane and layering of WaterWeld epoxy
RkJQdWJsaXNoZXIy MTMzNzEzMQ==