*3.4. Compression Test of SIFRCCs*

A universal testing machine (UTM) was used to study the compressive behavior of specimens. The specimens were placed centrally between the two compression plates, such that the center of moving head was vertically above the center of specimen, as shown in Figure 3, then the load was applied on the specimens by moving the movable head. The load and corresponding contraction were measured at different intervals. A review of existing studies on the mechanical properties of high-strength concrete and fiber-reinforced concrete found that there was an applicable strength limit for each prediction equation [8,10–17]. If this strength limit is exceeded, structural design problems can be generated by estimation of the unsafe side. Furthermore, major mechanical properties that define the stress-strain relationship have been found to be determined by the compressive stress of concrete and the fiber volume and shape [8,10–17]. Therefore, this study also used the high fiber volume fraction as a major variable. In addition, an experimental study on the elastic modulus under compressive stress, strain at peak strength, and Poisson's ratio was performed. Figure 3 shows the

experimental setup for compression tests. Compressive strength, modulus of elasticity, and Poisson's ratio were determined according to ASTM (American Society for Testing and Materials) specifications (ASTM C873 [27] and ASTM C469 [28], respectively).

**Figure 3.** Experimental setup for compressive stress and elastic modulus tests.

#### **4. Experiment Results and Analysis**

#### *4.1. Compressive Stress*

Figure 4 shows the compressive stress experiment results with respect to the fiber volume fraction for SIFRCCs. For fiber volume fraction of 6%, the average compressive stress was ~83 MPa. For fiber volume fraction of 5%, the average compressive stress was ~75 MPa; this was ~10% lower than that for the fiber volume fraction of 6%. Furthermore, for the fiber volume fraction of 4%, the average compressive stress was ~66 MPa; this was ~12% and ~21% lower than that for fiber volume fractions of 5% and 6%, respectively. The compressive stress tended to increase in proportion to the fiber volume fraction. It seemed that the increased amount of steel fibers with increasing fiber volume fraction produced a restraining effect on the specimen itself, thereby affecting the increase in compressive stress.

**Figure 4.** Compressive stress experiment results with respect to fiber volume fraction.

Many studies have shown that the mixing of fibers in fiber-reinforced concrete generally does not influence the compressive stress of the matrix itself. The compression experiment results of Shah and Rangan [29] reported that although the compressive strain at fracture increased significantly when fibers were mixed, the compressive stress did not clearly improve. As the fiber volume fraction increased, the compressive stress gradually increased at first; however, above a certain fraction (4% or more), this tendency changed somewhat. Within a certain mixing level of fibers, the compressive stress

increased because the resistance was increased by crack suppression; however, above this level, the compressive stress decreased owing to additional defects that appeared with the improvement effect.

The compressive fracture of conventional fiber-reinforced concrete was caused not by the yield or drawing of fibers but by the fracture of a matrix with relatively low strength before fibers play a structural role. Accordingly, the compressive stress was considered to improve with the fiber volume fraction as sufficient adhesion occurred between the fiber and the high-strength cementitious matrix used in this study. Micro-cracks were caused in conventional concrete and fiber-reinforced concrete by the interface properties of the coarse aggregates and the cement paste. For SIFRCCs, a composite made of fine particles with no coarse aggregates, cracks at the interface could be reduced, and the adhesion performance of the filling slurry and steel fibers could be improved. Therefore, the effect of steel fibers, such as the fiber volume fraction, rather than the effect of the compressive stress caused by matrix fracture was considered to be reflected.
