The axial behavior of circular HPFRCC specimens is presented with axial stress–strain relationships. These relationships are given in
Figure 2 and numerical test results are presented in
Table 4. In
Figure 2, the negative parts of the horizontal axis are used to show lateral strains (ε
h) whereas positive parts are used for axial strains (ε
c). In
Table 4, f′
co and ε
co are the unconfined axial strength and corresponding axial strain, whereas f′
cc and ε
cu are the compressive strength and ultimate axial strain of FRP confined specimens, respectively. f′
cu is the ultimate strength. It should be noted that, in this paper, the ultimate axial strain, which is the strain at sudden rupture of FRP, or the strain corresponding to 70% of the peak strength (whichever is less) is considered for confined specimens as ultimate point. On the other hand, axial strength was considered as the peak strength reached during the test. Stress–strain relationships of FRP confined HPFRCC specimens are different from those of FRP confined conventional concrete. Stress–strain behavior can be represented by a first ascending branch followed by a descending branch resulting with a sudden strength drop and in most cases with a second ascending branch. Although none of them were HPFRCC as the specimens tested in this study, similar behavior was also reported in FRP confined or tube encased variations of high/ultra-high strength/performance concrete specimens tested by Xie and Ozbakkaloglu [
19] and Zohrevand and Mirmiran [
20]. The descending and the second ascending branches are affected by the axial stiffness of FRP jacket. Higher the stiffness of FRP sheets, lower was the strength drop after the first peak of the stress–strain curves (
Table 4). Moreover, initial modulus of elasticity (E
c) is determined as the slope of the first ascending branch between 5% and 40% of the axial strength. As shown in
Table 4 and as expected, average E
c values of confined HPFRCC specimens (38,356 MPa) are similar to those obtained for the unconfined HPFRCC specimens (34,266 MPa) whereas the slopes of the second ascending branches depend on the FRP stiffness. Distinctions on the transition between the first and the second ascending branches demonstrated the effectiveness of the FRP sheets. The effect of FRP confinement on deformability is much more pronounced than that on strength for HPFRCC. As seen in
Table 4, for circular specimens confined with 2, 4, 6, 8 and 10 layers of CFRP, axial strengths were enhanced by 38%, 55%, 61%, 79% and 90%, respectively, while ultimate axial strains were improved by 110%, 332%, 391%, 521% and 580%, respectively. This enhancement is particularly important for HPFRCC members to be subjected to seismic actions. While the unconfined HPFRCC specimens fail in an extremely brittle manner as expected, the FRP jacketed HPFRCC specimens, when the stiffness of the jacket is sufficient, display a limited post-peak strength loss and then an ascending second branch occurs resulting with a more ductile behavior (
Table 4). When the stiffness of the jacket was insufficient, the second peak in stress strain curves could not be observed or seen to be lower than the first peak. The sudden post-peak strength loss can be attributed to the lack of confinement effectiveness that could be eliminated using more amount of FRP volumetric ratio for slightly confined specimens. Xie and Ozbakkaloglu [
19] stated that higher volume fraction of steel fibers (not less than 1.5%) can reduce the level of the strength loss right after the first peak for HSC.
The average value of the hoop rupture strain is obtained as about 0.006 for all specimens independent from FRP confinement ratio (
Table 4). Using this average value, strain efficiency factor (k
ε), which is the ratio of the hoop rupture strain to the ultimate uniaxial tensile strain of FRP, is calculated using Equation (2) as an average of 0.35. This value is lower with respect to FRP confined conventional concrete. For instance, Tamuzs et al. [
21,
22] reported the value of strain efficiency factor as 0.57 for FRP confined NSC and HSC with compression strengths up to 82 MPa. Similarly, Pellegrino and Modena [
24] recommended the value of 0.50 for NSC and HSC, whereas, in the case of steel fiber reinforced high strength concrete filled FRP tubes, Xie and Ozbakkaloglu [
19] obtained an experimental average value of 0.65.
The dilation ratio (μ) is calculated as the ratio of variations of lateral strain to axial strain (Equation (3)). As shown in
Figure 3, the dilation ratio remained almost constant up to axial strains corresponding to peak strength of the unconfined HPFRCC. Upon reaching this point, a rapid increase was observed in dilation related with the sudden increase in lateral strains due to damaging of HPFRCC. The sudden drop of strength right after the first peak can also be explained with the sudden increase of dilation upon reaching of unconfined strength for the specimens which are not sufficiently confined (i.e., the specimens jacketed with two plies of FRP sheets). It can be seen in
Figure 3 that, when stiffness of FRP jacket is higher, the increment in dilation at this point is less. Consequently, the sudden drop in axial strength is limited and the axial strength can be sustained until larger axial strains and higher deformability is achieved. For better confined specimens (i.e., specimens jacketed with 6, 8 or 10 plies of FRP sheets), dilation ratio remained almost constant until the end of tests at around the value of 0.6 after increasing suddenly to this value from the values of 0.15–0.20 at around axial strain levels of 0.003–0.005. It also worth mentioning that the characteristics of variation of dilation ratio with increasing axial strains are different with respect to that of conventional concrete reported by Ilki et al. [
7] in terms of both numerical values and general shape.