The preceding discussion has shown that considerable differences exist between the ACI shear design provisions and the SMCFT. In particular, the ACI method can not account for either the size effect or the aggregate effect. To investigate these differences, a series of ten large-scale and eight one-fifth scale – model shear-critical reinforced concrete beams were constructed and loaded to failure in the Mark Huggins Laboratory in the Department of Civil Engineering at the University of Toronto.
As summarized in Table 1 and Figure 2, the large beams measured 1510mm tall x 300mm wide x 9000mm long. Nine of the beams were reinforced with five No. 30 rebars in tension at an effective depth of 1400mm and with two No. 20 rebars for compression reinforcement. These beams were constructed without stirrups. The tenth large beam (specimen SB-10-H-S) was reinforced with eight No. 30 rebars in tension, two No. 20 rebars in compression, and approximately the minimum quantity of stirrups required by Equation (8) for the f’c on the day of test. The stirrups were in the form of single legs of 9.5mm (3/8’’) diameter rebars on alternating sides of the beam, spaced at 235mm. The one-fifth scale model beams measured 330mm tall x 122mm wide x 1800mm long. The reinforcement in nine of the small beams consisted of four 9.5mm diameter rebars at an effective depth of 280mm. The tenth scale-model beam (SSB-10-H-S) was reinforced in shear with 5mm smooth bars placed at 160mm on alternating sides.
(1) day of test
(2) Calculated at dv=0.9d from face of loading plate, incl. self-weight
(3) Calculated at d from face of support, incl. self-weight
The N series of beams were constructed with normal strength concrete and with four different maximum aggregate sizes (10, 20, 40 and 50mm). Two duplicate specimens were cast for each aggregate size. The H-series of beams were constructed with high-strength concrete and a maximum aggregate size of 10mm. The concrete used was commercially available from a local ready-mix supplier. The aggregate was strong crushed limestone shipped from a quarry on Manitoulin Island and conforming to CSA grading requirements.
The beams were moist-cured for five days, after which they were removed from their formwork. The beams were loaded to failure in three-point bending at an a/d ratio of 2.89. The large beams were tested in a 4500kN force-controlled Baldwin Test Frame, and the small beams were tested in a 1000kN displacement controlled MTS actuator (Figure 3). Upon reaching 85% of the monotonic failure load of the duplicate specimen, the applied load for beams SB-10-N-2 and SB-20-N-2 was cycled 20 times from 225kN to this load. This process was repeated at 90% and 95% of the monotonic failure load. After Specimen SB-50-N-2 failed on the east side of the beam (SB-50-N-2a), this side was clamped together with a series of externally installed Dywidag bars, and the beam was reloaded until failure occurred on the west end (SB-50-N-2b).
Figure 3: Specimen Test Setup