Fine aggregates (FA) with the grain size of lower than 4.75 mm and coarse aggregates (CA) with the grain size of higher than 4.75 mm were selected. According to ASTM C33 22 guidelines, the collected CA and FA were in the standard range as illustrated in Fig. 2. The coarse aggregates were collected from natural crushed stone with a specific gravity of 2.61 g/cm3 and the maximum size of 19 mm.
Fine aggregates were natural river sand with a specific gravity of 2.56 g/cm3. Further, type II Portland cement was used in this study.Silica fume as an effective pozzolanic material was also used to improve the mechanical properties of the mixtures.
These improvements are caused by pozzolanic reactions between the silica fume and free calcium hydroxide.A waterproofing material (Nitocote CM210) was used to perform the permeability test. This waterproofing material consisted of two components: a liquid polymer and cementitious powder. Two components were mixed in plastic drum using a slow speed drill with an approximate rate of 500 rpm for approximately three minutes and then applied to the four surface of specimens using a brush.
This material provided a cementitious and elastomeric coating and used to resist positive water pressure up to 7 bar. Cylindrical molds were used with the diameter of 3 inches (76.2 mm) and height of 6 inches (152.4 mm) for compression and splitting-tensile tests. Also, cubic molds were used with the dimension of 150 mm for permeability and ultrasonic pulse velocity (UPV) tests. In this investigation, RCC mixtures were designed according to the ACI 207.5R-11 23.
Portland cement, water, and coarse and fine aggregates were blended leading to a relatively homogenous mixture. They were then poured into the mold in three layers. Each layer was compacted with a surcharge until a paste was observed completely between the periphery of the surcharge and the cube mold. A mortar was applied between layers to improve the interface bonding.
After applying the mortar, the mixture left for two hours before performing the next layer. RCC mixtures were cured in the molds covered by a wet sack for 24 hours at the room temperature. Finally, they were demolded and placed in 20 °C water pool for 28 days.For permeability tests, as mentioned before, the four surfaces of the specimens were coated with Nitocote CM210 to prevent water from throwing out sidelong as shown in Fig. 3. Tests were performed on 54 various groups as presented in Table 1. They were named based on R(w/c)-C(SF), where w/c, C, and (SF) denoted the water-to-cement ratio, the cement content, and the percentage of silica fume replaced with the percentage of cement, respectively. To obtain appropriate results, three replicate specimens were tested in each experiment and results were averagedCompression and splitting tension tests were performed on RCC cylindrical specimens according to ASTM C39 24 and ASTM C496 25, respectively.
The splitting tensile strength (T) was calculated using Eq. 1.The ultrasonic pulse velocity instrument was used to determine the properties of RCC specimens according to ASTM C597 26. Longitudinal elastic waves pulses traversed through the specimen, and they were received and converted into electrical energy by a second transducer placed directly on the opposite side. Water permeability apparatus was used to determine the permeability coefficient of RCC specimens as shown in Fig. 4. As mentioned before, the four sides of specimens were waterproofed before setting them under permeability apparatus. The specimens were clamped with the circular gaskets.
In falling head procedure, the upper side of the specimen was sealed with the grease and rubber membrane to ensure one-dimensional flow. The water under five bar (0.5 MPa) pressure was then applied to the surface of the RCC specimens 27.where a is an internal cross-sectional area of the graduated cylinder; A is the average cross-sectional area of the test specimen; L is the length of the specimen; h1 and h2 are the initial and final heads, respectively; and t is the passed the time between h1 and h2.Note that the permeability of RCC is highly direction-dependent, e.g., the permeability coefficient is higher at the direction of flow is parallel to the plane of compaction.
In fact, treatment of joint planes to decrease the permeability is an important design consideration for RCC in hydraulic structures 28. Hence, in this study, RCC specimens were placed under apparatus according to Fig. 5; therefore, the water passed through the layers of the specimen.