Parametric Study on Mechanical Properties of Basalt Fiber-Reinforced Pea Gravel Concrete

Abstract

Basalt fiber-reinforced pea gravel concrete (BFRPGC) has remarkable potential for use as the retrofitting covers for masonry walls. However, a quantitative understanding of the mechanical properties of the BFRPGC material is still a perceived gap in the current literature. In this study, the role of basalt fibers in pea gravel concrete was evaluated by a comprehensive experimental investigation involving compressive strength tests and splitting tensile tests. Fiber length and volume fraction were selected as the key parameters. Two fiber lengths of 6 mm and 12 mm were considered, while the volume fraction corresponding to each of the fiber lengths was increased from 0.3% to 0.8%, with a step of 0.1%. The measured strengths were not simply proportional to the fiber volume fraction. The reason behind this phenomenon, i.e., the coupling effect of the bridging role of basalt fibers on concrete microcracks and the fiber agglomeration in concrete, was analyzed. The best performance of the BFRPGC material was achieved by incorporating 12-millimeter-length fibers with a volume fraction of 0.4%. Compared to that of the reference pea gravel concrete, a significant increase of up to 44.5% in compressive strength was recorded in this case. Furthermore, the failure mechanism of basalt fibers in pea gravel concrete was disclosed via the scanning electron microscope observations. In addition, the uniaxial compressive stress–strain model of the BFRPGC material was established.

1. Introduction

Short-cut fibers are currently being considered as an approach for improving concrete, in terms of both strength and strain capacity [1,2,3,4], while unidirectional continuous fibers have been widely used for strengthening concrete structures [5,6,7]. Theoretically, the development of concrete microcracks can be prevented by the inclusion of short-cut fibers to some extent [8]. Such materials represent a new trend in the development of concrete.

Different kinds of short-cut fibers have been used to reinforce concrete, such as asbestos [3], cellulose [9], steel [10], polypropylene [11], polyvinyl alcohols [12], carbon [4], basalt [13], aramid [14], polyethylene [15], and glass fibers [16]. Among them, steel and glass fibers are the most widely used ones. The flexural toughness, flexural fatigue endurance, and impact resistance of concrete were upgraded by the use of steel fibers [17]. However, some problems still exist in the use of steel fibers, such as the reduced workability and vulnerability to corrosion [18]. Glass fiber-reinforced concrete is mainly used to produce lightweight building components. However, glass fibers suffer degradation in the alkaline environment of such concrete, which greatly limits their application [16]. In contrast, basalt fibers extruded from molten basalt represent a new type of inorganic fibers with high acid and alkali resistance [13,19]. Moreover, basalt fibers can bond well with concrete. In particular, the tensile strength of basalt fibers is several times higher than that of steel fibers. Compared to carbon fibers, basalt fibers offer a price advantage [20]. Therefore, basalt fibers can be used as a substitute for steel and glass fibers in fiber-reinforced concrete [21,22].

Dias et al. [23] evaluated the benefits of basalt fibers on the fracture toughness of concrete. Furthermore, the bridging effect of basalt fibers on concrete microcracks was revealed, and fiber length and volume fraction were confirmed as important parameters. Similar conclusions were obtained from the tests reported in Refs. [24,25,26]. In addition, the effect of basalt fibers on fly ash geopolymer concrete was investigated experimentally [13]. Basalt fibers were also used in combination with other highly elongated fibers to enhance concrete, e.g., glass [20], polypropylene [27], and steel fibers [28]. The highly elongated fibers were used to improve energy dissipation capacity during collisions.

Basalt fibers were also used to reinforce cement-based materials. During the experimental study performed by Xu et al. [22], unique strain-hardening and multiple-cracking behaviors were exhibited by basalt fiber-reinforced engineered cementitious composites. Zheng et al. [29] found that flexural strength, fracture energy, and crack resistance, although not compressive strength, were improved by the inclusion of short-cut basalt fibers to various extents. These results are not entirely consistent with those in the concrete case detailed in the previous paragraph, which implies that the effect of basalt fibers varies according to aggregate type.

Conversely, an excessive amount of basalt fibers can result in problems such as fiber agglomeration and mixing difficulties [30]. This indicates that the appropriate volume fraction should be determined [31,32].

Currently, the pea gravel concrete covers reinforced by steel wire mesh are used to improve the shear capacities of seismic resistance-deficient masonry walls [33,34]. The main difference between pea gravel concrete and normal concrete lies in aggregate size. Small-sized aggregate endows the pea gravel concrete with excellent impermeability and a relatively smooth surface, which is suitable for use as cladding material for brick masonry walls. In particular, some pea gravel aggregates are derived from solid-waste recycling [35,36]. It is important to note that the pea gravel aggregate considered in this study is not from solid waste but is instead commercial aggregate.

In view of the significant improvements gained from basalt fibers in the mechanical properties of normal concrete, as confirmed by the above-mentioned research, it is reasonable to believe that basalt fiber-reinforced pea gravel concrete (BFRPGC) has a remarkable potential for use as the retrofitting covers for brick masonry walls, as shown in Figure 1. In such cases, the BFRPGC covers themselves have sufficient resistance against brittle fracture. Hence, the cumbersome construction of the steel wire meshes for the conventional pea gravel concrete covers is replaced by the simple shear keys at the interfaces between brick masonry walls and the BFRPGC covers (i.e., the grooves in mortar joints and L-shaped steel connectors in the plot). This provides a low-cost and efficient technique for retrofitting masonry structures.

Considering the significant difference in the coarse aggregate between the BFRPGC material and normal concrete [37], the influence of basalt fibers on the mechanical properties of pea gravel concrete has its own particularity. As mentioned earlier, it is similar to the differences in performance seen between basalt fiber-reinforced cement-based material and basalt fiber-reinforced normal concrete. However, a quantitative understanding of the mechanical properties of the BFRPGC material is still a perceived gap in the current literature.

In the current study, the effects of the short-cut basalt fibers on the mechanical properties of pea gravel concrete were quantitatively evaluated by a comprehensive experimental investigation. To obtain the best mix proportion for the BFRPGC material, fiber length and volume fraction were selected as the key parameters. Two fiber lengths and six fiber volume fractions were considered. Both the compressive strength test and splitting tensile strength test were performed for each of the combinations at the different levels of the two parameters. Furthermore, scanning electron microscope observations were conducted to reveal the failure mechanism of the basalt fibers in the pea gravel concrete. Moreover, the compression stress–strain relationship of the BFRPGC material was established.

2. Materials and Methods

2.1. Materials

A commercial composite silicate cement of Grade P.C32.5 [38] was chosen for this study. Pea gravel was used as a coarse aggregate (see Figure 2a), with a maximum particle size of 16 mm. The medium sand was used as the fine aggregate, with a fineness modulus of 2.5. The short-cut basalt fibers (see Figure 2b) were produced by Zhejiang Shijin Basalt Fibers Co., Ltd., Dongyang, China. The fiber properties provided by the manufacturer are listed in

Table 1. Physical and mechanical properties of basalt fibers.

Length (mm)	Diameter (μm)	Density (g·cm−3)	Elastic Modulus (GPa)	Tensile Strength (MPa)	Breaking Elongation (%)
6	16	2.65	81	4300	3.1
12	16	2.65	81	4300	3.1

.

2.2. Mix Proportions

The mix proportion for pea gravel concrete of Grade C30 [38] was used as the basis for the different BFRPGC test scenarios. In compliance with JGJ 55-2011 [38], the water–cement ratio and the sand ratio were designated as 0.42 and 0.50, respectively. The mix proportion for this reference pea gravel concrete are shown in

Table 2. Mix proportions of the reference pea gravel concrete.

Cement (kg·m−3)	Water (kg·m−3)	Sand (kg·m−3)	Pebble (kg·m−3)
485	204	785	771

. A total of twelve BFRPGC test scenarios were considered. Each of them was associated with a certain combination of fiber length and volume fraction. Two fiber lengths were selected, i.e., 6 mm and 12 mm. The volume fraction corresponding to each of the fiber lengths was increased from 0.3% to 0.8%, with a step of 0.1%, i.e., 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, and 0.8%.

2.3. Specimen Preparation

The specimens were the cubes with a side length of 100 mm. To avoid agglomeration of the short-cut basalt fibers as much as possible, a secondary mixing method was adopted. First, cement, sand, pea gravel, water, and half the amount of the short-cut basalt fibers were mixed for 100 s. Then, the other half of the basalt fibers was added, with a mixing time of 200 s. The mixed concrete material was cast into plastic molds treated with a pre-applied release agent. Working according to the provisions prescribed in GB/T 50081-2019 [39], the specimens were cured in a non-flow calcium hydroxide-saturated solution with a temperature of 20 °C ± 2 °C (see Figure 3).

A total of 255 specimens from 3 series were made. Among them, Series R consisted of the reference pea gravel concrete blocks without basalt fibers, while Series B6 and B12 corresponded to the BFRPGC specimens with fibers of 6 mm and 12 mm in length, respectively. There are one and six specimen groups in Series R and in each of the second two series, respectively. The six groups corresponded to the aforementioned six levels of the fiber volume fraction. Each of the groups consisted of 9 specimens. The groups in series B6 and B12 were labeled as GAB. In the label, “A” denoted fiber length; “B” denoted the name for the level of fiber volume fraction. The “A” values of 6 and X corresponded to fiber lengths of 6 mm and 12 mm, respectively. The “B” values of 3, 4, 5, 6, 7, and 8 corresponded to the fiber volume fractions of 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, and 0.8%, respectively.

2.4. Experimental Program

2.4.1. Compressive Strength Test

The compressive strength tests were conducted in accordance with the provisions of GB/T 50081-2019 [39]. The WAW-1000 hydraulic universal testing machine was used, as shown in Figure 4. It was produced by Changchun Kexin Testing Instruments Co., Ltd. (Changchun, China). Its maximum loading capacity and accuracy are 1000 kN and ± 1%, respectively.

During the tests, the loading speed was 5 kN/s. To disclose the strain–strain relationship of the BFRPGC material, a pair of strain gauges were set vertically on the centers of the two opposite sides of the concrete block (see Figure 5a), while a pair of strain gauges were applied to the centers of the other opposite sides along the transverse direction (see Figure 5b). The former and latter were installed to record the compressive strain and the tensile strain caused by the Poisson effect, respectively. The strain data were recorded with a DH3820 data acquisition instrument produced by Donghua Testing Instruments Co., Ltd., Jingjiang, China. The acquisition frequency used in this study was 100 Hz.

2.4.2. Splitting Tensile Strength Test

The splitting tensile strength tests were also performed with the hydraulic universal testing machine and in compliance with GB/T 50081-2019 [39]. Except for the shapes of the loading and bearing ends, the specimens were loaded using a similar approach to that used for the compressive strength tests (as detailed in Section 2.4.1). The loading and bearing ends are arc-shaped (see Figure 4) and the radius of their cross-sections is 75 mm. Only two transverse strain gauges were set at the two opposite sides of the specimen (see Figure 5c).

2.4.3. Scanning Electron Microscope Observation

As shown in Figure 6, an EM-30 scanning electron microscope (SEM), manufactured by COXEM Co., Ltd., Daejeon, Republic of Korea, was used for clarifying the failure mechanism of the short-cut basalt fibers in the pea gravel concrete. With this instrument, the fragments from the BFRPGC blocks that remained after the tests were scanned by a narrowly focused high-energy electron beam. The various types of physical information were elucidated through interactions between the electron beam and the BFRPGC material. By collecting, amplifying, and re-imaging the information, the damage characteristics of the basalt fibers were presented. The magnification range and backscattered electron image resolution of the EM-30 SEM were from 20 to 100,000 times and 8.0 nm @30 kV SE, respectively. The morphologies of the BFRPGC material could be clearly observed at magnifications ranging from 500 times to 1000 times.

3. Results and Discussion

3.1. Compressive Strength Test

3.1.1. Failure Modes

The typical compression failure of a reference pea gravel concrete block without basalt fibers is shown in Figure 7a. Two major penetrating cracks from the upper surface to the bottom surface of the specimen can be observed. As a result of these cracks, the specimen was divided into several vertical pieces that tended to fall apart. However, intensive but relatively short cracks appeared with the failure of the BFRPGC blocks, as shown in Figure 7b. However, the major penetrating cracks that can be seen in the reference pea gravel concrete block did not occur. This difference obviously came from the bridging effect of the basalt fibers on the pea gravel concrete. When microcracks developed in the concrete, the basalt fibers across the microcracks were pulled. As a result, the further development of the microcracks was restrained to a certain extent [19]. That is, a more uniform stress profile and microcrack distribution formed in the BFRPGC material. In this case, therefore, concrete failure was delayed, and the compressive strength was increased (as detailed in the next section).

3.1.2. Compressive Strength

The compressive strength (f_cp), in MPa, was calculated in accordance with the provisions of GB/T 50081-2019 [39], as shown in Equation (1):

f_cp = F/A

(1)

where F is the maximum load in N; A is the bearing area in mm².

Each of the measured axial compressive strengths was multiplied by 0.95 to account for the size effect [39]. The average value and the coefficient of variation for the compressive strength obtained from each of the different specimen groups, the increase in the average value relative to the reference pea gravel concrete value, and the ratio between the coefficient of variation for the BFRPGC material and that of the reference pea gravel concrete are listed in

Table 3. Results of the compressive strength tests.

Series	Group	Fiber Volume Fraction (%)	Average Compressive Strength (MPa)	Increase in Average Compressive Strength (%)	Coefficient of Variation (%)	Ratio for Coefficient of Variation Ratio (%)
R	/	0.00	31.9	/	7.03	/
B6	G63	0.30	30.5	−4.39	7.23	102.8
	G64	0.40	26.4	−17.2	5.44	77.38
	G65	0.50	28.4	−11.0	5.85	83.21
	G66	0.60	28.4	−11.0	5.96	84.78
	G67	0.70	30.5	−4.39	12.9	183.5
	G68	0.80	29.8	−6.58	5.93	84.35
B12	GX3	0.30	50.0	56.7	6.22	88.48
	GX4	0.40	46.1	44.5	6.24	88.76
	GX5	0.50	49.4	54.9	4.91	69.84
	GX6	0.60	37.4	17.2	5.37	76.39
	GX7	0.70	31.8	−0.31	6.70	95.31
	GX8	0.80	33.6	5.33	15.1	214.8

. The increase in the average value is defined as (f_cp,ab − f_cp,a0)/f_cp,a0 − 1, where f_cp,ab and f_cp,a0 are the average values of the BFRPGC material and the reference pea gravel concrete, respectively. As the volume fraction of basalt fibers was increased, there were changes in the average value and coefficient of variation in Series B6 and B12, as illustrated in Figure 8a and Figure 8b, respectively.

As shown in Table 3 and Figure 8, the average compressive strength and the corresponding coefficient of variation for the reference pea gravel concrete were 31.9 MPa and 7.03%, respectively. It is noteworthy that the value of the coefficient of variation was much lower than the lower limit of the recommended range for concrete given in GB 50010-2010 [40] (i.e., from 17.2% to 18.9%).

In Series B12, when the fiber volume fraction was increased, the average compressive strength first increased and then decreased (see Figure 8a). As the fiber volume fraction varied from 0.3% to 0.5%, the average compressive strength values remained in a range from 46.1 MPa to 50.0 MPa. These average compressive strength values were 44.5% to 56.7% higher than that of the reference pea gravel concrete. This clearly demonstrates that the short-cut basalt fibers rendered significant enhancement to the pea gravel concrete. However, the average compressive strength dropped dramatically to a range from 31.8 MPa to 33.6 MPa as the fiber volume fraction was further increased. The reason for the drop in strength is detailed in the next paragraph. There was no obvious improvement in comparison with the reference pea gravel concrete in such cases.

The above phenomenon is due to the variation in the compactness of concrete when the fiber volume fraction was increased. It is noteworthy that the compactness of concrete was not sensitive to relatively small amounts of basalt fibers when the fraction ranged from 0.3% to 0.5%. Hence, the compression failure of the BFRPGC material was delayed by the bridging effect of the basalt fibers (as discussed in Section 3.1.1), and the compressive strength was improved significantly. However, the excessive amounts of basalt fibers led to the marked degradation of the fluidity of the pea gravel concrete, in terms of obstruction due to the microfiber network composed of basalt fibers. In such cases, the negative effect of the agglomeration of basalt fibers in the pea gravel concrete could not be neglected. Concurrently, there were plenty of concrete pores that formed around the agglomerated basalt fibers, which resulted in a decrease in the compactness of the concrete. During compressive loading, this tended to induce stress concentrations in the concrete, where the premature crushing of the concrete was prone to occur. Consequently, the positive role played by the bridging effect from the basalt fibers was counteracted.

With the fiber volume fraction, the coefficient of variation for the compressive strength in Series B12 also fluctuated significantly, as demonstrated in Table 3 and Figure 8b. The coefficient of variation varied in a range from 4.91% to 6.70% and was lower than the value of 7.03% for the reference pea gravel concrete when the fraction was no more than 0.7%. However, as the fraction was increased to 0.8%, the coefficient of variation increased dramatically to 15.1%. The value was 114.7% higher than that of the reference pea gravel concrete but was still lower than the lower limit of the recommended range for concrete in GB 50010-2010 [40].

As presented in Table 3 and Figure 8a, in the different specimen groups in Series B6, the compressive strength of the BFRPGC material was slightly lower than that of the reference pea gravel concrete. This implies that the bridging effect of the basalt fibers on the concrete microcracks was not significant when the fibers were relatively short. Moreover, it is reasonable to believe that the negative effect of basalt fibers on the compactness of concrete still existed, to various extents. As the fiber volume fraction increased, the trend in the variation of the coefficient of variation in this series was similar to that in Series B12 (see Table 3 and Figure 8b). However, the maximum coefficient of variation, i.e., 12.9%, was recorded at a fiber volume fraction of 0.7%, which was slightly lower than the fiber volume fraction corresponding to the maximum coefficient of variation in Series B12.

From the above analyses, it is clear that the enhancement of the compressive strength was significant when the 12 millimeter-length fibers were used and when the fiber volume fraction was in a range from 0.3% to 0.5%. Furthermore, the discreteness of the compressive strength could be controlled effectively in such cases.

3.1.3. Strain Capacity

To further explore the mechanical behavior of the BFRPGC material, the typical axial stress–axial strain and axial stress–transverse strain relationship curves recorded in the compressive strength tests of the different specimen groups in Series B12 are compared to each other at the second and first quadrants in Figure 9, respectively.

The initial linear portions of the axial stress–axial strain relationship curves associated with the different fiber volume fractions are very close to each other, as illustrated in Figure 9. It implies that the positive role of basalt fibers (i.e., the bridging effect) was not apparent before the development of microcracks in the concrete. However, an obvious discrete phenomenon exists among the nonlinear portions of the curves. It is the combined result of the bridging effect of basalt fibers on concrete microcracks and the negative effect of basalt fibers on the compactness of concrete. Furthermore, a direct relationship between the fiber volume fraction and the strain corresponding to the maximum compressive stress (hereafter referred to as “the maximum compressive strain”) was observed. The recorded maximum compressive strains reached a range from 3000 με to 5000 με, while the fiber volume fraction varied from 0.3% to 0.5%. This clearly indicated delays in the compression failure. The phenomenon could only come from the effective bridging effect provided by the short-cut basalt fibers. However, the recorded maximum compressive strains associated with other levels of fiber volume fraction were generally no more than 2000 με. The above strain measurement results were consistent with the recorded compressive strengths.

It is also noted that the axial strain value was generally much higher than the corresponding transverse tensile value at the different compression stress levels. In particular, in the nonlinear response phase, significant strain-softening phenomena were observed in the axial stress–axial strain relationship curves associated with the fiber volume fractions of 0.3%, 0.4%, and 0.5%. However, the transverse stress–transverse strain relationship curves generally terminated at the ends of their respective initial linear portions. This was apparently due to the fracturing of the strain gauges that were spanned by the concrete cracks.

3.2. Splitting Tensile Strength Test

3.2.1. Failure Modes

A typical example of the splitting tensile failure of the reference pea gravel concrete specimens is presented in Figure 10a. The concrete block was split into two halves by the main crack formed at its middle and along the vertical direction. A different failure mode was observed in the BFRPGC specimen. Relatively intensive vertical cracks occurred in the BFRPGC block (see Figure 10b), which exhibited a certain degree of plasticity. Obviously, this is due to the bridging effect of the short-cut basalt fibers in the concrete.

3.2.2. Splitting Tensile Strength

The splitting tensile strength (f_ts), in MPa, was determined in accordance with the provisions of GB/T 50081-2019 [39], as shown in Equation (2):

f_ts = 2F/πA.

(2)

The size conversion coefficient for the measured tensile splitting strengths was 0.85 [39]. The average value and the coefficient of variation for the splitting tensile strength values obtained from each of the different specimen groups, the increase in the average value relative to the reference pea gravel concrete case, and the ratio between the coefficient of variation for the BFRPGC material and that of the reference pea gravel concrete are listed in

Table 4. Results of the splitting tensile strength tests.

Series	Group	Fiber Volume Fraction (%)	Average Splitting Tensile Strength (MPa)	Increase in Average Splitting Tensile Strength (%)	Coefficient of Variation (%)	Ratio for Coefficient of Variation Ratio (%)
R	/	0.00	10.4	/	13.0	/
B6	G63	0.30	10.7	2.88	18.0	138.5
	G64	0.40	9.80	−5.77	13.7	105.4
	G65	0.50	10.1	−2.88	7.06	54.31
	G66	0.60	11.2	7.69	8.43	64.85
	G67	0.70	12.3	18.3	7.15	55.00
	G68	0.80	11.4	9.62	12.7	97.69
B12	GX3	0.30	10.6	1.92	7.95	61.15
	GX4	0.40	13.3	27.9	14.1	108.5
	GX5	0.50	10.2	−1.92	11.0	84.62
	GX6	0.60	11.5	10.6	16.1	123.9
	GX7	0.70	12.5	20.2	11.4	87.69
	GX8	0.80	12.5	20.2	14.4	110.8

. Along with the fiber volume fractions, the variations in the two values are illustrated in Figure 11a and Figure 11b, respectively.

As shown in Table 4 and Figure 11a, the average splitting tensile strength in Series B12 oscillated substantially with the fiber volume fraction. An increasing trend in the average value was observed when the fiber volume fraction was increased from 0.3% to 0.4%. Hence, the maximum value for the average splitting tensile strength was reached at a fiber volume fraction of 0.4% and at 13.3 MPa. This was 27.9% higher than the maximum value for the reference pea gravel concrete. However, a dramatic drop in the average splitting tensile strength was recorded when the fiber volume fraction was further increased to 0.5%. In this case, the average splitting tensile strength was approximately equal to that of the reference pea gravel concrete. Thereafter, it rose again in accordance with the fiber volume fraction. At fiber volume fractions of 0.7% and 0.8%, the average splitting tensile strength was stable at a value of 12.5 MPa. It is noteworthy that this value was still 20.2% higher than that of the reference pea gravel concrete. From this phenomenon, relative to the compressive strength case, it appears that the splitting tensile strength was less sensitive to fiber agglomeration at a relatively high fiber volume fraction.

The average splitting tensile strength in Series B6 also oscillated with the fiber volume fraction. At a fiber volume fraction of 0.7%, the maximum value for the average splitting tensile strength in this series was obtained. This was 12.3 MPa, 18.3% higher than that for Series R. However, at each of the fiber volume fractions of 0.4% and 0.5%, the average splitting tensile strength was slightly lower than that for Series R.

With the fiber volume fraction, the coefficients of variation for the splitting tensile strength in Series B12 oscillated in a range from 7.95% to 16.1%, as shown in Table 4 and Figure 11b. A similar trend existed in Series B6, with an oscillating range from 7.06% to 18.0%. Furthermore, it is noteworthy that the oscillations of the coefficient of variation in the case of the splitting tensile strength were more pronounced than those of compressive strength at the relatively low levels of fiber volume fraction ranging from 0.3% to 0.6%. This demonstrates the complexity of the effect of basalt fibers on the discreteness of the splitting tensile strength.

At a fiber volume fraction of 0.4% in Series B12, the maximum splitting tensile strength of this study was achieved, and the degree of dispersion was controlled at the same level as that of the reference pea gravel concrete. Hence, the mix proportion for the BFRPGC material in this test scenario is considered to be optimal for achieving the splitting tensile strength.

3.2.3. Strain Capacity

The typical axial stress–transverse strain relationship curves obtained from the splitting tensile strength tests of the different specimen groups in Series B12 are compared to each other in Figure 12. Apparently, each of the curves can be divided into two parts.

The first part is characterized by a linear response, in which the transverse tensile strain is proportional to the axial compressive stress. The ends of the first parts for most of the curves are in a range from 250 με to 500 με. Obviously, these linear parts reflect the transverse expansion of the BFRPGC material, due to the Poisson’s ratio effect.

The second part is characterized by a very slowly ascending trend. That is, the transverse tensile strain increases significantly, while the axial compressive stress changes a little. Most of the second parts terminate at approximately 1500 με, due to the fracturing of the transverse strain gauges. The characteristics of the second part reflect the development of concrete macrocracks. It is quite reasonable to suppose that the relatively high tensile strain levels reached in the second portion came from the bridging effect of the short-cut basalt fibers.

3.3. Results of Scanning Electron Microscopy

As shown in Figure 13a, the basalt fiber was a round rod and had a smooth surface. The measured fiber diameter was 14.32 μm. The typical failure characteristics of the short-cut basalt fibers during the compressive strength test are presented in Figure 13b–d. Generally, microscopic cracks existed within the pea gravel concrete itself and propagated as the compression was increased, which led to a reduction in the mechanical properties of the concrete. In the case of BFRPGC, the effective bridging role of basalt fibers in restraining the development of the microscopic cracks had the prerequisite that reliable interfacial bonds should be formed between the fibers and the concrete. This was confirmed by the phenomenon shown in Figure 13b, i.e., mortar debris adhered tightly to the fiber surface until the crushing of the BFRPGC material. The good bonding performance can be attributed to the large number of hydroxyl bonds formed on the fiber surfaces. During such a hydration process, a large amount of the hydration product continuously accumulated on the surfaces of the basalt fibers, which led to a strong adhesive force between the basalt fibers and the pea gravel concrete by forming shear keys at the interfaces [41,42].

As shown in Figure 13b, due to this good bonding performance, the short-cut basalt fibers across the macroscopic cracks suffered tensile fractures with the crushing of the BFRPGC block. The trace for the pull-out part of the fractured fiber is clearly present in the plot. As demonstrated in Figure 13c, the pull-out part of the fractured fiber shows characteristic tearing damage. This type of damage corresponds to a sheet-like fiber shape that was significantly different from the shape of the undamaged fiber (see Figure 13a). These phenomena confirmed the effectiveness of the bridging effect from the short-cut basalt fibers in the pea gravel concrete. Figure 13b,c was obtained with Series B12 and at a fiber volume fraction level of 0.4%. At the much higher fiber volume fraction of 0.8% found in Series B12, the short-cut basalt fibers agglomerated in the pea gravel concrete, as shown in Figure 13d. As a result, a larger number of voids are observed. In comparison with the 0.4% fiber volume fraction case (see Figure 13b), a significant deterioration in the compactness of the concrete can be observed. This provides intuitive evidence for explaining the oscillation in BFRPGC strength as the fiber volume fraction was increased.

3.4. Uniaxial Compressive Stress–Strain Model

As demonstrated by the test results, a significant enhancement to the mechanical behavior of the pea gravel concrete was achieved in the case of the fiber length of 12 mm, whereas the effect of the 6-millimeter-length fibers was not obvious. Therefore, the uniaxial compressive stress–strain model of the BFRPGC material was developed for the 12-millimeter-length scenario to facilitate the corresponding structural design and nonlinear analysis.

A reliable uniaxial compressive stress–strain model has been proposed for normal concrete with short-cut high-performance fibers [43,44,45,46,47,48], as presented in

Table 5. Proposed uniaxial compressive stress–strain models for fiber-reinforced concrete.

Researchers	Analytical Models
Wang et al. [43]	$y=\left\{\begin{array}{cc}{\alpha }_{1}x+(6-5{\alpha }_1)x^5+(4{\alpha }_1-5)x^6& (0\le x<1)\\ \frac{x}{\alpha {(x-1)}^2+x}& (x\ge 1)\end{array}\right.$; α1 = 1.417 + 0.697 VBF − 6.699 VPPF; here, α = 5.638 + 24.01 VBF − 468.34 VPPF; VBF and VPPF are the volume fractions of basalt fibers and polypropylene fibers, respectively.
Zhao et al. [44]	$y=\left\{\begin{array}{cc}\frac{nx}{n-1+x^n}& (0\le x<1)\\ \frac{x}{\alpha {(x-1)}^b+x}& (x\ge 1)\end{array}\right.$; n = ECεc.r/(ECεc.r + fc.r); α = 0.00022 fc.r 2.75 + 0.746; here, fc.r and ECεc.r are the peak stress and modulus of elasticity of steel fiber-reinforced expanded-shale lightweight concrete, respectively.
Chen et al. [45]	$y=\left\{\begin{array}{cc}\frac{nx}{n-1+x^n}& (0\le x<1)\\ \frac{x}{\alpha {(x-1)}^2+x}& (x\ge 1)\end{array}\right.$; n = Ec/(Ec + fcr); α = 0.157 fcr0.785 − 0.905; here, fcr and Ec are the peak stress and modulus of elasticity of basalt fiber-reinforced rubberized recycled coarse aggregate concrete, respectively.
Ding et al. [46]	$y=\left\{\begin{array}{cc}(\frac{E_{\mathrm{c}}}{E_{\mathrm{c}1}}x-x^2){\left.\left[1+(\frac{E_{\mathrm{c}}}{E_{\mathrm{c}1}}-2)x\right.\right]}^{-1}& (0\le x<1)\\ {[{(\frac{1}{1.8})}^2\xi x(x_{\lim }-1)+\frac{1}{0.9}(2x_{\lim }-1)]}^{-1}& (x\ge 1)\end{array}\right.;$$\xi =\frac{4[{(x_{\lim })}^2(\frac{E_{\mathrm{c}}}{E_{\mathrm{c}1}}-2)+2x-\frac{E_{\mathrm{c}}}{E_{\mathrm{c}1}}]}{(x_{\lim })(\frac{E_{\mathrm{c}}}{E_{\mathrm{c}1}}-2)+1}$; xlim = (0.05Ec/Ec1 + 0.9) + [(0.05Ec/Ec1 + 0.9)2 − 0.9]0.5; here, Ec and Ec1 are the tangent modulus and the secant modulus from the origin to the compressive strength, respectively; ξ and xlim are the parameters related to the elastic modulus and the limited strain at a stress of 0.9fc, respectively.
Ou et al. [47]	$y=\frac{\beta x}{\beta -1+x^{\beta }}$; β = 0.71(RI)2 − 2(RI) + 3.05; fcr = fcr.p + 2.35RI; εcr = εcr.p + 0.0007RI; RI = Vf Lf/Df; here, Vf, Lf, Df and fcr.p are the volume fraction, length, and diameter of the steel fibers and the compressive strength of plain concrete, respectively.
Abbass et al. [48]	$y=\frac{\beta x}{\beta -1+x^{\beta }}$; β = 1.401(RI)2 − 1.56(RI) + 2.42; fcr = fcr.p +5.59RI; εcr = εcr.p + 0.000261RI; RI = Vf Lf/Df.

. In general, such a uniaxial compressive stress–strain model consists of two parts, i.e., the ascending and descending elements. This indicates that the general law for the constitutive relationship of the normal concrete still applies to the fiber-reinforced concrete [45,49]. Therefore, the classical uniaxial compressive constitutive relationship suggested by Guo et al. [46] was used as the basis for developing the model for the BFRPGC material. The model is shown in Equation (3):

$$y=\left\{\begin{array}{cc}ax+(3-2a)x^2+(a-2)x^3& (0\le x<1)\\ \frac{x}{b{(x-1)}^2+1}& (x\ge 1)\end{array}\right.$$

(3)

where x is the ratio of the compressive strain to its peak value and is defined as ε/ε_cr; ε and ε_cr are the compressive strain and its value corresponding to the maximum compressive stress, respectively; y denotes the ratio of the compressive stress to its maximum value and is defined as σ/f_cr; σ and f_cr are the compressive stress and its maximum value, respectively; a and b are the critical parameters associated with the rising and falling parts of the uniaxial compressive constitutive relationship, respectively.

In this study, using the least squares method [43], the respective relationships of parameters a and b with the fiber volume fraction (V_BF) were established on the basis of the fitted uniaxial compressive stress–strain curves corresponding to the different fiber volume fractions. With the related test results, the two parts (i.e., the ascending and descending one) for each of the curves were fitted separately in the regression analyses. An example of these regression curves is shown in Figure 14 and corresponds to a fiber volume fraction of 0.4%. With regard to the fiber volume fraction range studied (i.e., from 0.3% to 0.8%), the proposed expressions for parameters a and b are shown in Equations (4) and (5), respectively.

$$a=13.0V_{\mathrm{BF}}-0.2$$

(4)

$$b=1611.5V_{\mathrm{BF}}-2.5$$

(5)

Furthermore, as discussed in Section 3.1 and Section 3.2, the best performance of the BFRPGC material was achieved by the 12-millimeter-length fibers with a volume fraction of 0.4%. This parameter combination is proposed for future engineering applications. In this case, the different stages in the mechanical behavior of this optimal BFRPGC material until compression failure is clearly reflected in the four parts of the fitted curve shown in Figure 14. Combined with the experimental observations in this study, the characteristics of the four parts can be summarized as follows.

The first stage corresponds to the initial near-linear part of the curve. At this stage, no significant changes to the initial microcracks inside the BFRPGC material commence. Hence, the mechanical behavior of the BFRPGC material is dependent on that of the matrix, i.e., the pea gravel concrete. The enhancement offered by the basalt fibers is not obvious.

In the second stage, the curve is characterized by a nonlinear ascending trend until a point of maximum compressive stress. During this process, the crushing of the concrete is delayed by the bridging effect provided by the basalt fibers across the microcracks, and higher compressive strength is reached.

This is followed by a stage dominated by the unstable expansion of cracks. After the maximum compressive stress is reached, the bridging effect degrades rapidly, while the fracture and pull-out phenomena in the basalt fibers occur across the board. Consequently, within the limited increase seen in the compressive strain, the curve dropped to an obvious inflection point that corresponds to approximately 30% of the maximum compressive stress.

The final stage reflects the residual behavior after intensive cracking. It is manifested by a relatively slowly descending trend beyond the inflection point. Correspondingly, the separation of the BFRPGC material commences. In such a severe cracking scenario, it is reasonable to consider that the bridging effect of basalt fibers is largely lost, and the residual compressive capacity comes from the damaged pea gravel concrete matrix.

4. Conclusions

In this study, the effect of short-cut basalt fibers on the mechanical properties of pea gravel concrete was quantitatively evaluated according to combinations of the different levels of the two key parameters. The main conclusions are summarized as follows:

• Fiber length is the most important parameter for the BFRPGC material. A significant bridging effect restraining the microcrack development in concrete was achieved by the 12-millimeter-length fibers, whereas no obvious role was observed in the 6-millimeter-length scenario.
• The mechanical properties of the BFRPGC material oscillated substantially with the fiber volume fraction. The phenomenon was the combined result of the bridging effect of basalt fibers on concrete microcracks and the negative effect of fiber agglomeration on the compactness of concrete.
• In both the compressive strength tests and splitting tensile tests, intensive but relatively short cracks appeared with the failure of the BFRPGC blocks. This specific failure mode demonstrates that the more uniform stress profile and microcrack distribution were formed in the BFRPGC material, in comparison with the reference concrete case.
• The parameter combination of a fiber length of 12 mm and a fiber volume fraction of 0.4% was found to be optimal. The average compressive and splitting tensile strengths in this case were 44.5% and 27.9% higher than those of the reference pea gravel concrete, respectively, and their respective dispersions were well controlled.
• The direct relationship between the fiber volume fraction and the strain corresponding to the maximum compressive stress was disclosed. The values of the strain reached a range from 3000 με to 5000 με, while the fiber volume fraction was increased from 0.3% to 0.5%. This clearly indicated a delay in the compression failure.
• By scanning electron microscopy observation, not only was the failure mechanism of the tensile fracture disclosed for the basalt fibers in pea gravel concrete but a significant deterioration in the compactness of concrete was proved in the relatively high fiber–volume fraction scenario.
• The uniaxial compressive stress–strain model of the BFRPGC material was established on the basis of the test data. Four stages of mechanical behavior are clearly indicated by the model, namely, the elastic, crack development, unstable crack expansion, and residual behavior stages.

Based on the findings of this study, the following future research efforts are recommended, including the effects of the different shear keys at the interfaces between the BFRPGC covers and masonry walls, the combined enhancement effects of basalt fibers and high-elongation fibers on the pea gravel concrete, and the impact toughness of the BFRPGC material.