1. Introduction
Advances in the areas of infrastructure and construction are closely related to the development of new technologies, especially in terms of advanced materials, durability and new construction processes [
1]. In this context, fibre-reinforced cementitious composites (FRCC) emerged at the end of the 19th century in an attempt to create a new cementitious material with two constituents, matrix and reinforcement. The matrix is a paste (cement mixed with water), mortar (paste with sand) or concrete (mortar with large aggregates). The fibres are discrete, randomly oriented and distributed through the composite volume. The matrix and reinforcement work together, producing a synergistic effect and an efficient composite. The addition of fibres allows a desired level of performance regarding a specific property (or properties) [
2]. Accordingly, high-performance fibre-reinforced cementitious composites (HPFRCC) have optimized properties and superior performance characterized by improved flexural strength, strain-hardening behaviour led by multiple cracking and relatively large energy absorption [
2,
3]. Recent studies have been conducted to broaden the knowledge on FRCCs, and many others are under development, given that the subject is broad for new discoveries.
Mechanical behavior, material characterization, and design considerations of FRCCs have been thoroughly reviewed in comprehensive works [
4,
5,
6]. Several other types of fibres such as sisal [
7] and synthetic materials (usually polymers) [
8,
9] are common alternatives to traditional steel fibres. High fibre contents (5–20%) are present in the so-called slurry-infiltrated fibre concrete (SIFCON) [
10,
11]. The fibre–matrix interaction is investigated to understand the behaviour of composite phases through pull-out tests and evaluation of associated phenomena such as adhesion, friction and anchoring [
12,
13,
14,
15]. The influence of additions is pertinent, as it changes the interfacial transition zone (ITZ), contributing to fibre bonding in the matrix [
16,
17]. The use of hybrid fibres has been promising because it combines different fibre properties, generating an optimized synergistic effect [
18,
19,
20]. The distribution and orientation of fibres in the matrix has been shown to be an influential factor in composite behaviour, so it has been the focus of recent studies [
21,
22]. Impact loads imply high loading rates and alter a material response. In this case, FRCCs are an effective alternative to combat cracking, which promotes performance increases [
23]. Fire resistance and very low temperatures can be increased by the addition of fibres. In the case of high temperatures with high heating rates, the fibres restrict spalling [
24,
25,
26]. Hydraulic shrinkage occurs during curing of cementitious materials and can be reduced by the presence of fibres [
27]. Fibrous reinforcement can also be used in self-compacting concretes, expanding the possibilities of use of the composite [
28]. Prediction models have been used to aggregate information and experimental data [
29]. In addition to the search field presented, another aspect has emerged to combine sustainability concepts with FRCCs. In this regard, attempts are made to make the composite eco-friendly, using natural aggregates, recycled fibres, cements and sustainable additions [
30,
31].
Throughout the development of HPFRCCs, SIFCON emerged. SIFCON is characterized by a high fibre content, a particular moulding process and high capacity to absorb energy [
10,
11,
32]. The production of SIFCON consists of inserting the reinforcement fibres into the moulds and subsequently pouring a fluid mortar to fill the voids between the fibres [
33]. The result is a composite with a dense fiber distribution and excellent mechanical behaviour [
34]. The use of SIFCON is widespread and can be applied to repairs and reinforcements, pavements, precast elements and structures resistant to seismic tremors [
35,
36].
Several studies have addressed the flexural strength of FRCCs, considering the different characteristics of the matrix, the reinforcement and the interfacial transition zone [
32,
37,
38,
39,
40]. According to Naaman and Reinhardt [
41], the mechanical behaviour under bending can be classified as deflection hardening when the composite withstands a greater load after the point of first fracture or deflection softening when it presents a strength lower than the load withstood by the matrix at the time of cracking. Generally, the tests are performed in a reduced model with some simplifications, using prismatic specimens and square sections. Other models are also adopted, such as panels and slabs, but they are less common [
4].
The mechanical parameters during bending (i.e., flexural strength, deflection capacity and toughness) provide the characteristics of the studied material and are influenced by several factors: curing conditions, presence of large aggregates and additions, fibre length and content, geometry and loading rate during the test [
4]. The effect of size and the fibre distribution are other sensitive parameters. Reduction in the size of the specimens accentuates the fibre alignment tendency during the moulding process, which favours mechanical performance [
42]. The distribution and orientation of the fibres are influenced by factors such as matrix fluidity, pouring method and volumetric fraction [
43]. Eventually, the distribution of fibres may be intentional, which results in specific and individual behaviours [
31].
In this scenario, it is evident that HPFRCCs have several applications and a great potential for new discoveries. Several studies have addressed the mechanical behaviour of HPFRCCs in general using conventional specimens that represent more robust structures with fibre distribution throughout the composite volume. However, experimental approaches with narrow models combined with the use of concentrated fibres are still sparse in recent studies. Thus, the present study evaluated the mechanical behaviour of thin slabs made of a high-performance steel fibre-reinforced cementitious composite. Focusing on the flexural mechanical behaviour, a new model with slender sections and fibres concentrated in the region where tensile stresses are applied was evaluated.
2. Materials and Methods
The materials used were Portland cement of high early strength (type III according to ASTM C150/C150M [
44]), active silica, washed natural sand, superplasticizer, short steel fibres and water. The methodology of preparation and analysis of the specimens included dosing of fluid mortar, moulding, flexural strength tests, statistical analysis and fibre–matrix analysis. Steel fibres with anchored ends were used for reinforcement of the cementitious matrix. The properties of the fibres are shown in
Table 1.
River sand was used as the fine aggregate, which had a maximum grain size of 4.8 mm, promoted strength gain and contributed to the increase in matrix density. A polycarboxylate-based superplasticizer modified with stabilized nanosilica was used to increase the fluidity of the mortar and allow the flow of material through the dense fibre distribution in SIFCON [
45].
Beglarigale et al. [
46] used an optimal proportion of cementitious materials, mineral additives, large aggregates and water. By following the same concept and experimentally adjusting the dosage of the constituents, with the inclusion of a superplasticizer, the constant ratio shown in
Table 2 was obtained.
The specimens had rectangular faces, slender sections and dimensions of 330 × 165 × 25 mm
3 (length, width and thickness). Moulding was performed following the slurry infiltration method [
11,
32]. The fibres were placed in the moulds at 1% vol., 3% vol. and 5% vol. The fluid mortar was then poured to fill the spaces between the fibres (
Figure 1). Two specimens were prepared for each fibre content.
After 28 days of curing in a solution saturated with calcium hydroxide (1.7 g per litre of water), the specimens were subjected to a four-point flexural strength test. The tests were performed in an electronically controlled universal testing machine (EMIC DL30000, Belo Horizonte, Brazil), with a loading rate equal to 0.1 MPa/s and a distance between supporting pins (L) equal to 240 mm. The values of applied load and deflection at the midpoint of the span were measured and adjusted in the form of loading curves. The stress values were calculated as established in the standard [
47]. The characteristic load, stress and deflection values were obtained for the point of first cracking (P
LOP, σ
LOP and δ
LOP), which is called limit of proportionality (LOP), and for the maximum loading (P
u, MOR and δ
MOR), which is identified as the modulus of rupture parameters. The toughness (T) was calculated as the area under the load-versus-deflection curve for the characteristic deflections of L/150 and L/40 and for the maximum deflection. The toughness factor was calculated according to Equation (1) [
37,
48,
49]:
where
is the width of the specimen, and
is its height. The toughness factor
in J/m³ was then obtained and is the average withstood load after matrix fissuring.
A statistical analysis was performed to evaluate the significance of the data obtained and the correlation among the response variables (load, stress, deflection, toughness and toughness factor). First, the correlation among the variables was measured by the correlation coefficient. Multivariate analysis of variances (MANOVA) was used to evaluate the influence of fibre addition, considering the response variables together. Finally, univariate analysis of variance (ANOVA) was used to verify the significance of the data in terms of the response variables individually, i.e., ANOVA does not consider the correlation among the response variables. The magnitude of the effect represents the impact of the categorical variable (fibre content) on the response variables; its value is greater than zero, and values of approximately 1 or higher indicate a broad effect. The power of the test shows the robustness of the tests in identifying differences between the evaluated groups; it ranges from 0 to 1, and values above 0.8 are considered reasonable [
50,
51]. The significance level adopted is 5% (0.05); therefore, the analyses have a confidence interval equal to 95%. A fibre–matrix analysis was performed using optical microscopy to evaluate the composite elements (matrix and reinforcement) and the interfacial transition zone.