Durability and Mechanical Performance of Sisal-Fiber-Reinforced Cementitious Composites for Permanent Formwork Applications

Abstract

Reinforced concrete structures must balance immediate structural performance with long-term durability against environmental degradation, particularly carbonation-induced corrosion. While traditional cast-in-place concrete covers serve as the primary barrier, their substitution with prefabricated permanent formworks made of fiber-reinforced cementitious composites often fails to provide the necessary protective qualities required for aggressive environments. This study evaluates the durability and mechanical behavior of sisal-fiber-reinforced cementitious composites specifically engineered for use as permanent formwork. Short sisal fibers, treated by hornification to enhance dimensional stability and fiber–matrix adhesion, were incorporated at dosages of 2%, 4%, and 6% by weight. The experimental program included tests for water absorption, ultrasonic pulse velocity, axial compression, three-point flexural strength, and accelerated carbonation. The results indicated that composites with 2% and 4% of fibers exhibited reduced water absorption, sorptivity, compressive strength, and modulus of elasticity compared to the reference cement matrix. Residual stress values further demonstrated that the composites maintain significant post-cracking strength and stress transfer capacity, confirming their viability for structural elements. Although sisal-fiber-reinforced cementitious composites exhibit higher porosity and water absorption than conventional concrete used as reinforcement cover, they show sufficient resistance to carbonation to ensure a service life exceeding 50 years for reinforced concrete elements.

1. Introduction

Reinforced concrete (RC) is one of the most widely used materials in the world and has diverse applications in civil engineering, including infrastructure projects, bridges, and skyscrapers. To make the construction process faster and more efficient, the use of prefabricated permanent formwork systems for producing reinforced concrete beams, slabs, walls, and columns has increased. These systems improve construction productivity by reducing the labor and shoring required to support conventional formwork. In addition, it provides greater geometric homogeneity and better surface finish of the structural elements.

However, permanent formwork becomes part of the structure itself, replacing the concrete cover that protects the reinforcement from the action of external aggressive agents, as shown in Figure 1. Therefore, the material used in the formwork must ensure chemical resistance to carbonation and to the penetration of aggressive ions comparable to that of conventional concrete. This requirement is essential because reinforcement corrosion is estimated to incur an annual cost equivalent to about 3.5% of the global gross domestic product [1]. This problem can be mitigated throughout the service life of reinforced concrete structures by providing concrete cover with adequate thickness and quality to protect the reinforcement, as specified in structural design standards.

Different materials have been used as permanent formwork for beams [3,4,5]. The advantages of using ferrocement permanent formwork for the production of reinforced concrete beams and columns have been reported by several researchers [6,7,8]. According to [9], beams produced with ferrocement permanent formwork exhibit higher flexural strength, first-crack strength, and energy absorption capacity compared with conventional reinforced concrete beams. However, the concrete cover provided to the steel mesh used in ferrocement formwork is often insufficient to ensure adequate durability, which limits its application due to the risk of corrosion. Corrosion of the mesh can lead to generalized cracking of the ferrocement, resulting in loss of strength and ductility [10]. The effect of using ferrocement formwork on the durability of reinforced concrete structures was investigated by [11], who showed that reduced porosity, lack of homogeneity, and the presence of microcracks in the cement matrix and in the interfacial transition zone can influence the durability performance of the system.

To ensure greater durability of permanent formwork, ferrocement reinforcement has been replaced by carbon fabrics, polymeric fibers, or glass fibers. Fabric-reinforced panels with carbon fibers used as permanent formwork for slabs have shown an increase in cracking load and ductility of structural elements [12], especially when surface treatments are applied to the inner face to improve bonding with the cast-in-place concrete [13].

Permanent formworks made of ultra-high-toughness cementitious composites (UHTCCs) reinforced with polyvinyl alcohol (PVA) fibers were produced by [14]. Due to the strain-hardening behavior of UHTCCs, these formworks can improve crack control and enhance the tensile performance of reinforced concrete beams compared with conventional reinforced concrete elements. Similarly, Leung et al. [15] developed permanent formworks using pseudo-ductile cementitious composites (PDCCs) reinforced with PVA fibers and showed that the multiple cracking behavior of the formwork under flexural loading reduces crack width. As a result, PDCC formwork can limit the penetration of aggressive agents and act as an effective protective layer against steel corrosion in reinforced concrete structures.

Glass fiber-reinforced cementitious composites were also investigated by [16], who demonstrated their feasibility as permanent formwork for slabs up to 200 mm thick. Although these studies highlight the potential of different materials for use as permanent formworks based on mechanical performance, little attention has been given to the effect of these systems on the durability of structural elements.

Biomaterials have increasingly been used in the production of cement-based construction elements to replace natural aggregates or polymeric fibers with materials that require lower energy for production, generate less CO₂, and originate from renewable sources [17,18,19]. Using experimental, analytical, and numerical approaches, it was demonstrated that wood–cement composite panels can be used in permanent formwork systems up to 1.5 m in height.

Permanent formworks reinforced with long sisal fibers (900 mm) for floor slab production were evaluated by [2]. The formworks were manufactured in arch and trapezoidal shapes with lengths of 108 mm, widths up to 80 mm, and thicknesses of 20 mm. Flexural tests conducted on the permanent formworks and on a full-scale slab produced with these systems indicated that sisal-fiber-reinforced cementitious composites are suitable for use in permanent formwork systems.

Lima et al. [2] produced permanent formworks in sisal–cement composites in prefabricated slabs, replacing conventional construction elements of expanded polystyrene (EPS) or ceramic blocks. The composite elements showed better performance in bending tests compared to the conventional elements, while reinforced concrete slabs produced with the composite material formwork exhibited equivalent performance, indicating the potential of using the material as permanent formwork.

Composites reinforced with sisal fibers have shown good performance and great potential for application in construction elements [20,21] since the mechanisms of deterioration of plant fibers in alkaline environments were identified by Li et al. [22], minimized through modification of the cementitious matrix [23] and treatment of plant fibers [24,25,26,27]. However, to be used as a permanent formwork in reinforced concrete structures, it is essential that the fiber-reinforced composite presents resistance to the degradation phenomenon (chloride penetration and carbonation), sufficient for the reinforcement to remain passive within the estimated useful life.

Some studies have indicated the possibility of producing composites with high mechanical performance using sisal fibers [2,28,29,30]. However, the presence of fibers may lead to an increase in the water absorption and porosity of the composites [31]. As a result, uncertainties remain regarding the use of permanent formworks made of plant fiber-reinforced composites, particularly due to the need to protect the internal reinforcement from the ingress of aggressive agents. Studies evaluating the effect of sisal fiber composites on the durability and performance of reinforced concrete structures have shown that their presence may facilitate the ingress of carbon dioxide into the cementitious matrix [32,33]. The interaction between sisal fibers and carbonation is primarily associated with matrix porosity and fiber durability. In certain instances, the incorporation of fibers can increase the apparent porosity and water absorption of the composite, thereby facilitating greater CO₂ diffusion and making the material more susceptible to carbonation [34]. As a durability enhancement strategy, accelerated carbonation is occasionally applied intentionally to natural fiber composites; this process lowers the alkalinity of the cement matrix, which helps reduce the degradation of sisal fibers due to alkaline hydrolysis [35]. During curing, carbonation naturally consumes portlandite (Ca(OH)₂), resulting in the formation of calcite (CaCO₃). Additionally, in activated fly ash systems, calcite formation via carbonation may eventually lead to a reduction in ettringite peaks observed during the later stages of curing [36,37].

Therefore, this study evaluates the durability and mechanical behavior of sisal-fiber-reinforced cementitious composites specifically engineered for use as permanent formwork in comparison with conventional concrete cover. We examine different sisal fiber contents (0%, 2%, 4% and 6% by mass) and investigate the influence of the fibers on durability-related physical properties, including water absorption, capillary absorption, and carbonation.

2. Materials and Methods

Permanent formwork becomes part of the structure itself, replacing the concrete cover that protects the reinforcement from the action of external aggressive agents, as shown in Figure 1. Although conventional cast-in-place concrete cover acts as the main protective barrier, replacing it with prefabricated permanent formworks composed of fiber-reinforced cementitious composites does not always deliver adequate protection for use in highly aggressive environments. Therefore, this section presents the raw materials and methods for evaluating the performance and durability of sisal–cement composites used in permanent formwork.

2.1. Sisal Fiber Treatments

The sisal fibers utilized in this investigation were sourced from Valente City, Bahia, Brazil. Initially, the raw fibers underwent immersion in water at 50 °C for one hour to eliminate residues from the decortication process [24]. The sisal fibers were rinsed with running water at room temperature and oven-dried at 40 °C for 24 h [24]. For use as reinforcement, the fibers were systematically aligned and cut into 40 mm segments before undergoing five further cycles of hornification treatment which were performed in accordance with the procedure developed by Ferreira et al. [24].

In each hornification cycle, fibers were initially saturated in water at approximately 23 °C for three hours, followed by oven-drying at a controlled heating rate of 1 °C/min until reaching 80 °C for 16 h [24]. The fibers were maintained at this temperature for sixteen hours [24]. Subsequently, all fibers underwent drying at room temperature (approximately 23 °C) [24].

2.2. Cement-Based Composite

To produce the composites, the following materials were used: (i) Portland cement of high initial strength; (ii) fly ash; (iii) silica fume; (iv) river sand; (v) sisal fiber; (vi) superplasticizer; (vii) viscosity-modifying agent; and (viii) water. The matrix design followed a binder-to-sand ratio of 1:1 and a water/binder ratio of 0.40. The binder was prepared by replacing 45% of the cement with mineral additions, specifically fly ash (40%) and silica fume (15%). The composites were produced with sisal fiber contents of 2%, 4%, and 6% by mass.

The Brazilian Portland cement of high initial strength corresponded to Type III of ASTM C150 [38], with a specific mass of 3.14 g/cm³ and a BET specific surface area of 1.36 m²/g. In addition, fly ash (specific mass of 2.15 g/cm³ and specific surface area of 0.76 m²/g) and silica fume (specific mass of 2.32 g/cm³ and specific surface area of 15.15 m²/g) compose the binder.

To preserve the proper workability of the mixture and avoid segregation, a superplasticizer (SP), a chemical additive based on polycarboxylates with a density of 1.09 g/cm³ and a solids content of 30%, as well as a viscosity-modifying agent (VMA) based on a high-molecular-weight polymer, was used.

Table 1. Chemical composition, in oxides (% by mass), and loss on ignition (LOI) of the cement, silica fume and fly ash used.

Oxide	Amount (%, by Mass)
	Cement	Silica Fume	Fly Ash
SiO2	16.18	85.51	56.61
Al2O3	3.76	1.47	27.15
Fe2O3	3.77	5.13	6.27
CaO	64.28	0.30	1.58
MgO	4.90	1.40	1.70
K2O	1.05	1.90	3.11
SO3	4.83	0.53	0.64
Na2O	0.60	2.10	1.10
Other	0.63	1.66	1.84
LOI (%)	5.07	4.40	1.64

presents the chemical compositions of the raw materials that make up the binder (cement, silica fume, and fly ash), determined by X-ray fluorescence spectroscopy (XRF) with the aid of the FRX S2 Ranger equipment of the Bruker brand.

The mixtures were produced with a binder/sand ratio, mass equal to 1, and a water/cement ratio of 0.40. The binder was composed of cement (45%), fly ash (40%), and silica fume (15%) to ensure the durability of the plant fiber and improve the workability of the composite in the fresh state, as demonstrated by Lima et al. [2]. The consumption of materials to produce a reference matrix (F0, without fibers) and cementitious composites containing 2% (F2), 4% (F4), and 6% (F6) of fibers by weight is presented in

Table 2. Raw material consumption (kg/m³) of the reference matrix (F0, without fibers) and the cementitious composites containing 2% (F2), 4% (F4) and 6% (F6) of fibers, by mass.

Raw Material Consumption (kg/m3)	Composites
	F0	F2	F4	F6
Cement	382.28	375.19	368.37	361.79
Fly ash	339.80	333.51	327.44	321.59
Silica fume	127.43	125.06	122.79	120.60
Sand	849.51	833.77	818.60	803.97
Water	329.89	329.34	328.80	316.09
Fiber	-	11.12	21.83	32.16
SP	14.16	13.90	13.64	30.82
VMA	2.12	2.08	1.23	1.21
Water/cement ratio	0.40	0.40	0.40	0.40

. The spreading verified for the self-compacting matrix without fibers (F0) was 400 ± 10 mm (without the application of blows), whereas the composites with normal consistency showed a spread flow table equal to 270 ± 10 mm after the application of blows, according to the procedure recommended in NBR 13276-2016 [39].

Brazilian Portland cement and fine aggregate were used in the production of the concrete cover (CC) and composites, respectively. The coarse aggregate of basaltic origin presented a maximum characteristic dimension and specific mass equal to 19 mm and 2.79 g/cm³, respectively.

Table 3. Mixture (in mass) and raw material consumption (kg/m³) of concrete cover (CC).

Mixing Characteristic	Raw Material
	Cement	Sand	Gravel	Water/Cement Ratio
Mixture, by mass	1.00	1.60	2.20	0.45
Consumption (kg/m3)	459.58	753.33	1011.08	206.81

presents the mixture (in mass) and consumption of materials to produce 1 m³ of concrete cover.

Cylindrical specimens 100 mm in diameter and 200 mm in height were used to evaluate the physical and mechanical properties of the concrete. To evaluate the properties of the composite under the conditions of use, impressions of the specimens were made in four layers of 50 mm, as shown in Figure 2. The permanent formwork produced with the composite material was developed with a thickness of 20 mm; thus, the short fibers (40 mm in length) were aligned in the longitudinal direction. Because the present research proposes the use of composites with plant fibers in the production of permanent formworks, it is important to verify whether the properties of this composite material are like those of the covering concrete used.

2.3. Physical and Mechanical Properties

The immersion water absorption and capillarity absorption tests were performed according to NBR 9778 [40] and NBR 9779 [41], respectively, whereas the axial compression and static modulus of elasticity tests were performed according to NBR 5739 [42] and NBR 8522 [43], respectively.

The axial compressive strength determination test was performed in a servo-controlled hydraulic press with a capacity of 1200 kN and stress increment rate of 0.45 MPa/s. The static modulus of elasticity determination test was performed using a Shimadzu servo-controlled universal testing machine with a capacity of 100 kN. All tests were performed at 28 d of age using three cylindrical specimens with a diameter of 100 mm and a height of 200 mm.

The flexure test of three notched prism points was performed at 28 days of age, with the aid of a Shimadzu servo-controlled universal testing machine at a displacement rate equal to 0.5 mm/s [44]. Three prismatic specimens (150 mm × 150 mm × 550 mm) containing a 25 mm deep notch on the underside were used (Figure 3). From the test results, the proportionality limit and residual stresses were determined using Equation (1) and Equation (2), respectively.

$$f_{R,i}={\displaystyle \frac{3\ast F_{R,i}\ast L}{2\ast b\ast h_{sp}^2}}$$

(1)

$$f_{fct,L}={\displaystyle \frac{3\ast F_L\ast L}{2\ast b\ast h_{sp}^2}}$$

(2)

where f_fct,L is the limit of proportionality; f_R,i are the residual stresses (MPa); F_L is equal to the highest value of the load up to a deflection of 0.05 mm; and F_R,i are the loads relative to the crack openings (C_MODj), as shown in Figure 4. L is the distance between supports (mm); b is the cross-sectional side (150 mm); and h_sp is the usable height of the prism (125 mm).

The dynamic modulus of elasticity was determined by determining the ultrasonic pulse propagation velocity in three cylindrical specimens at 28 d of age, with a diameter of 100 mm and a height of 200 mm, according to NBR 15630 [45]. The Pundit Lab equipment from Proceq was used. The dynamic modulus of elasticity (E_d) was calculated by determining the wave propagation velocity (V) using Equation (3) and Equation (4), respectively.

$$E_d=\rho \ast V^2\ast K$$

(3)

$$K={\displaystyle \frac{\left(1+v\right)\ast (1-2v)}{1-v}}$$

(4)

where V is the speed of propagation of the wave (km/s), Ed is the dynamic modulus of elasticity (GPa), K is a coefficient determined by the Poisson coefficient (ν) (for concrete, it is equal to 0.2, so K is 0.9), and ρ is the apparent specific mass of the material in the hardened state (kg/m³).

2.4. Carbonation

Three cylindrical specimens (100 mm diameter and 60 mm height) were used for the accelerated carbonation test. The test parameters followed ISO 1920 [46], using a QUIMIS carbonation chamber (Figure 5a) with a temperature control of (27 ± 2) °C, CO₂ concentration of (3.0 ± 0.5) %, and a relative humidity of (65 ± 5) %. The carbonation depth was measured after 56, 63, and 70 days of exposure to CO₂ by sprinkling phenolphthalein aqueous solution and analyzing the images obtained.

The images of the surfaces of the samples sprayed with phenolphthalein were analyzed with the aid of ImageJ 1.54r software (Figure 5b), estimating the carbonation depth (x), according to Equation (5).

$$x=K\ast t^{0.5}$$

(5)

where t corresponds to the time of exposure to CO₂ (days) and K refers to the accelerated diffusivity coefficient (mm*time^−0.5).

3. Results and Discussions

3.1. Physical Properties

Table 4. Physical characterization of the concrete cover (CC) and of the reference cementitious composite (F0) and those containing varying levels of fibers (F2, F4, and F6).

Physical Property	Composites *				Concrete Cover CC *
	F0	F2	F4	F6
Water absorption (%)	9.39 ± 0.12 a	7.54 ± 0.65 b	9.23 ± 0.93 a,b	11.25 ± 0.49 a,b	5.43 ± 0.30 c
Void ratio (%)	16.14 ± 0.15 a	12.66 ± 1.09 b	15.53 ± 1.59 c	19.13 ± 0.79 a	12.36 ± 0.59 b
Bulk density (%)	1.72 ± 0.01 a	1.68 ± 0.00 b	1.68 ± 0.00 b	1.70 ± 0.01 a	2.28 ± 0.02 c
Capillarity (g/cm2)	1.03 ± 0.32 a	0.77 ± 0.30 a	0.82 ± 0.29 a	1.00 ± 0.44 a	0.42 ± 0.17 b
Sorptivity (g/cm2.h0.5)	0.25 ± 0.03 a	0.18 ± 0.01 b	0.19 ± 0.01 a,b	0.20 ± 0.01 a,b	0.15 ± 0.01 c

* For each property, equal letters represent statistically equal means.

presents the average values obtained for the physical properties of the composites and concrete along with their respective standard deviations. A comparative test of means using the Tukey method indicated a significant difference between the mean values obtained. The results are presented in Table 4 using letters, where equal letters for each property indicate statistically equal averages.

According to Table 4, the composites reinforced with plant fibers exhibited superior properties compared to those observed in the cementitious matrix. This superiority can be attributed to the high water absorption capacity of the fibers and the incorporation of air during the mixing and release processes of the composite. Other researchers reached similar conclusions regarding the influence of sisal fiber additions on the physical properties of composites [2,32]. Notably, these authors found that the addition of 2% fiber content resulted in a 19% reduction in water absorption by immersion and a 21% decrease in the void index compared to the reference matrix [2]. This behavior is attributed to the use of self-compacting matrices, which enable better homogenization of the composite, good workability (despite the addition of fibers), and low air incorporation, particularly with lower fiber contents.

Another critical factor influencing the results was the layer-wise molding of the composites (see Figure 2), which led to fiber alignment in the transverse direction of the specimen. This alignment effectively reduced the interconnection between the capillary channels along the height of the specimen, resulting in lower absorptivity for all composites compared to the matrix. Notably, the fibers utilized in this study underwent a hornification process, resulting in lower water absorption compared to natural fibers without any processing [27].

From the analysis of the results of the physical properties of composites with different fiber contents (F2, F4, and F6), presented in Table 4, it can be concluded that a beam manufactured with the same covering concrete (CC) will exhibit lower water absorption than a beam with a permanent formwork made from composites reinforced with sisal fibers at contents of 2%, 4%, and 6%. All evaluated physical parameters indicated higher compactness (higher density and lower void index) and a reduced likelihood of penetration by deleterious agents (lower absorption and sorptivity) in the CC than in the composites.

The molding procedure also demonstrated the ability to produce composites with 4% fiber content that exhibited physical properties comparable to those of the matrix. This capability opens the possibility of manufacturing a material that is not only tenacious, but also durable. Nevertheless, it is noteworthy that the increase in the fiber content directly influences the physical properties (water absorption, sorptivity, bulk density, capillary absorption and void ration) of the composite, as depicted in Figure 6.

In Figure 6, a pronounced increase in water absorption (by immersion or capillarity) and void index is evident with a 2–6% increase in fiber content. The porous nature of the fibers in the composites led to a higher water storage capacity within the material [31,47]. However, the molding process has a mitigating effect on the rate of water absorption by capillarity, resulting in a less intense increase in absorptivity with increasing fiber content. Table 4 indicates that the water absorption by capillarity was statistically similar in all mixtures. The matrix’s water absorption was approximately 28% higher than that of composites containing 2% fibers. For the composites with 4% and 6% fibers, no statistically significant variation in water absorption was observed. The specific masses were similar for all the composites.

3.2. Mechanical Properties

3.2.1. Axial Compressive Strength, Static Modulus of Elasticity and Three-Point Bending Tests

Table 5. Mechanical characterization of reference cementitious composites (F0), fiber contents (F2, F4 and F6), and CC.

Mechanical Property	Composites *				Concrete Cover CC *
	F0	F2	F4	F6
Compressive strength (fc, MPa)	29.72 ± 3.07 a	28.96 ± 2.12 a	25.72 ± 2.13 a	23.98 ± 2.81 a	47.93 ± 3.60 d
Modulus of Elasticity (E, GPa)	27.10 ± 1.70 a	24.30 ± 1.55 b	20.76 ± 0.71 c	18.88 ± 0.59 c	32.95 ± 1.56 d
Limit of proportionality (ffct,L, MPa)	1.59 ± 0.28 a	2.20 ± 0.07 a,b	2.48 ± 0.09 b	2.54 ± 0.13 b	-
Residual stress 1 (fR,1, MPa)	-	1.64 ± 0.29 a	2.91 ± 0.40 b	3.57 ± 0.05 b	-
Residual stress 2 (fR,2, MPa)	-	1.55 ± 0.25 a	2.85 ± 0.45 b	3.44 ± 0.09 b	-
Residual stress 3 (fR,3, MPa)	-	1.30 ± 0.22 a	2.21 ± 0.28 b	2.68 ± 0.09 b
Residual stress 4 (fR,4, MPa)	-	1.09 ± 0.19 a	1.84 ± 0.16 b	2.15 ± 0.06 b

* For each property, equal letters represent statistically equal means.

summarizes the results for axial compressive strength, static modulus of elasticity, and three-point bending tests for the plain matrix (F0) and the cementitious composites reinforced with varying fiber contents (F2, F4, and F6). Additionally, the mechanical properties of the concrete cover (CC) were evaluated. The reported values represent the arithmetic means accompanied by their respective standard deviations. To evaluate the statistical significance of the results, Tukey’s test was performed, indicating significant differences between the mean values across the tested groups.

According to Table 5, the maximum reduction observed in the average values of the compressive strength (f_c) of the composites is 17% with an increase in fiber content. This reduction in strength was comparatively lower than that reported by other researchers [2,28,29,30]. The use of a self-compacting matrix and molding method likely contributed to this outcome, enabling the production of composites (F2 and F4) with a void index statistically lower than that of the matrix (F0). On the other hand, the modulus of elasticity (E) demonstrated reductions of approximately 10%, 23%, and 30% with the introduction of 2%, 4%, and 6% fibers, respectively.

Compared to the concrete cover, the composite that closely approximated its mechanical properties was F2, showing a 40% reduction in the axial compressive strength and a 26% reduction in the modulus of elasticity. However, it is crucial to evaluate the mechanical behavior of cementitious materials under bending stress.

Figure 7 illustrates the outcomes of the notched beam bending test, showcasing load diagrams concerning crack opening (CMOD). Linear behavior was evident until the appearance of the first crack, where, for the matrix (F0), a brittle rupture occurred at a maximum load of 0.75 kN with a crack opening (CMOD) of approximately 0.02 mm. In the composites, the linear elastic zone before the first crack is influenced by the matrix; however, the post-crack behavior differs because the fibers can transfer stress between the crack faces. This characteristic makes the post-crack behavior less brittle because the fibers contribute to stress transfer until the process of total pullout or rupture of the fibers occurs [48].

The composite with 2% fibers exhibited a significant reduction in the load after the first crack, indicating that the percentage of fibers crossing the crack was insufficient to transfer the entire applied load. Although load recovery occurs when the fibers are engaged, the residual stress values, as shown in Table 5, remain below the limit of proportionality, reflecting a behavior known as deflection softening [28,29].

In contrast, composites with 4% and 6% fibers exhibited a different behavior characterized by an increase in the resistance load after the first crack, known as deflection hardening (Figure 7). During this post-cracking period, the load resisted by the composite surpassed the load required to break (peak) the matrix [2]. Beyond the increase in flexural strength, higher fiber content contributed to better crack control in the composites. Figure 8 illustrates the cracking behavior of the composites with 2% fibers (F2) and 6% fibers (F6) during the test (CMOD = 5 mm). It is evident that a higher fiber content limits the opening of the crack propagating from the notch.

From the results of the notched prism bending test, it was possible to determine the stress versus crack opening diagram under direct tension, defined by the post-cracking behavior of fiber-reinforced concrete (FRC), which must be applied in the design of these elements in the Ultimate Limit State [49]. The Linear Model, shown in Figure 9a, uses two reference values obtained using Equations (6) and (7).

$$f_{Fts}=0.45\mathrm{\ast }{f}_{R1}$$

(6)

$$f_{Ftu}=f_{Fts}-{\displaystyle \frac{w_u}{{CMOD}_3}}\ast \left(f_{Fts}-0.5\mathrm{\ast }{f}_{R3}+0.2\mathrm{\ast }{f}_{R1}\right)$$

(7)

where f_Fts represents the post-service cracking residual direct tensile strength of the fiber-reinforced composite (MPa), f_Ftu represents the post-ultimate post-cracking residual direct tensile strength of the fiber-reinforced composite (MPa), and w_u is the maximum crack opening considered in the design, expressed in millimeters.

When considering a crack opening limit of 1.5 mm to minimize the impact of external agents on concrete and reinforcement in reinforced concrete elements protected by a composite permanent formwork, the data from Table 5 can be used to derive post-cracking constitutive laws for sisal fiber composites, as illustrated in Figure 9b. Under tension, linear behavior characterized by deflection softening was observed for all composites, with a higher fiber content resulting in higher stress. The calculated f_Fts values were 0.8 MPa, 1.3 MPa, and 1.6 MPa for reinforced composites with fiber contents of 2%, 4%, and 6%, respectively, while the respective f_Ftu values were 0.5 MPa, 0.8 MPa, and 1.0 MPa.

3.2.2. Ultrasonic Pulse Velocity and Dynamic Modulus of Elasticity

Table 6. Ultrasonic wave propagation velocity (V) and dynamic modulus of elasticity (E_d) of cementitious composites reinforced with sisal fibers.

Composites *	V (km/s)	Ed (GPa)
F0	3.81 ± 0.03 a	22.49 ± 0.02 a
F2	3.78 ± 0.02 a	21.58 ± 0.18 b
F4	3.74 ± 0.02 a,b	21.22 ± 0.21 b,c
F6	3.68 ± 0.04 b	20.76 ± 0.48 c

* For each property, equal letters represent statistically equal means.

shows the results of the ultrasonic pulse velocity and dynamic modulus of elasticity obtained from the propagation test of ultrasonic waves in the matrix (F0) and composites reinforced with sisal fibers (F2, F4, and F6).

There was a slight reduction in the ultrasonic pulse velocity (V) with an increase in the vegetable fiber content. According to Ashrafian et al. [50], the ultrasonic pulse velocity in fiber-reinforced concrete is influenced by two primary factors: the material density and void content. In homogeneous materials, ultrasonic pulse velocity is typically a function of the modulus of elasticity, Poisson’s ratio, and material density. However, in heterogeneous materials, such as concrete with fibers, the ultrasonic pulse intensity can be influenced by the presence of discontinuities, such as aggregate grains, fibers, voids, or microcracks.

Figure 10 illustrates the various internal conditions encountered during ultrasonic pulse testing according to ACI 228.2R [51]. The AA line represents the shortest path for homogeneous material. The second case (BB line) depicts the pulse passing through regions with different densities, such as an aggregate or fiber. In the CC line, the trajectory is modified by the deviation around the edge of a pore or tip of a fissure. In the latter cases, the time taken by the pulse to traverse the sample is longer than that in the first situation. In the fourth situation (DD), the passage of the pulse is interrupted by a void (fissure or large pores). Based on this concept, the British standard BS EN12504-4 [52] establishes concrete quality standards based on ultrasonic pulse speed, ranging from very weak (v < 2.0 km/s) to excellent (v > 4.5 km/s).

In the composites produced in this study, an increase in the specific mass and void index was observed (Table 4), owing to the higher fiber content. Consequently, there was a reduction in the ultrasonic pulse velocity values (Table 6), owing to the introduction of a greater number of defects in the material. This reduction in ultrasonic pulse speed due to an increase in flax plant fiber content aligns with the findings of Page et al. [53], who linked the phenomenon to the appearance of internal defects. These defects resulted from the alignment of the fibers during specimen preparation, leading to the formation of fiber agglomerates within the matrix. In concrete reinforced with manufactured fibers, there was an observed increase in porosity and a subsequent reduction in the ultrasonic pulse speed [50].

Despite the reduction in speed values, all the produced composites exhibited good quality, as per the British standard BS EN12504-4 [52]. This suggests that the introduction of fibers did not result in significant damage to the material.

The ultrasonic pulse velocity in cement-based materials depends on the modulus of elasticity, which in turn is directly correlated with compressive strength. Therefore, it is anticipated that a relationship exists between the ultrasonic pulse velocity and compressive strength of the composites, allowing the ultrasonic pulse test to be a potential estimator of this mechanical property. Numerous correlation equations have been proposed for this purpose, the most common of which is an exponential function represented by Equation (8) [54].

$$f_c=a\ast e^{b \ast V}$$

(8)

where f_c is the compressive strength and V is the ultrasonic pulse speed. The coefficients “a” and “b” were obtained from experimental results. For simple concrete, from the results of other researchers, the values of “a” are equal to 1.146 and 1.190, and the values of “b” are equal to 0.770 and 0.715 [54].

Using the experimental results obtained in this study for sisal-fiber-reinforced composites, together with data reported by [55] for coconut fiber composites and [56] for bamboo fiber composites, the relationship between compressive strength and ultrasonic pulse velocity can be fitted using Equation (8). Coefficient a shows similar values for all fiber types, while coefficient b ranges from 1.59 to 1.80. The fitted relationship is presented in Figure 11.

The results revealed a decrease in the compressive strength and dynamic modulus of elasticity, accompanied by a reduction in the ultrasonic pulse velocity. This reduction is directly proportional to the increase in the sisal fiber content of the composites, as illustrated in Table 6. Despite efforts to establish a correlation between the dynamic modulus of elasticity (Table 6) and static modulus of elasticity (Table 5) of cement-based materials, it is apparent that the absolute values differ for the studied composites. Consequently, the variation in these properties and the increase in the fiber content cannot be correlated, as shown in Figure 12, which presents the modulus of elasticity of the composite (E_composite) to the matrix (E_matrix) ratio as a function of fiber content. While the dynamic modulus of elasticity exhibited only an 8% reduction relative to the matrix with the introduction of 6% fibers, the static modulus of elasticity experienced a more substantial reduction of 30% for the same fiber content.

3.3. Carbonation Analysis

As outlined in the methodology, the assessment of carbonation depth involved aqueous–alcoholic spraying of phenolphthalein on the fractured surface of the specimens. The results, depicted in Figure 13, are further detailed in

Table 7. Average carbonated area of samples F0, F2, F4, F6 and CC after 56, 63 and 70 days of exposure to accelerated carbonation.

Mixture	Average Carbonated Area (mm2)
	56 Days	63 Days	70 Days
F0	432.19 ± 73.30	445.84 ± 28.01	552.49 ± 30.37
F2	406.60 ± 62.06	417.67 ± 35.40	426.53 ± 97.30
F4	450.37 ± 24.48	511.05 ± 21.39	600.80 ± 36.68
F6	834.19 ± 47.05	1336.77 ± 81.32	1483.32 ± 86.83
CC	316.13 ± 16.65	401.42 ± 39.19	411.96 ± 22.91

and Figure 14, illustrating the average carbonate areas and depths after 56, 63, and 70 d of accelerated carbonation.

After 70 days of exposure, it was observed that across the composites, the average carbonate area and carbonation depth increased with the increase in fiber content, excluding the F2 composite, which displayed a 22.80% reduction compared to the reference matrix (F0). Specifically, the carbonated areas increased by 8.6% and 168.48% for composites with 4% and 6% fibers, respectively. In comparison to the concrete cover (CC), composites containing fibers exhibited the following increases in carbonate area after 70 days of exposure: 3.54% (F2), 45.84% (F4), and 260.06% (F6).

The elevation in carbonate regions within the fiber-reinforced composites can be attributed to the high water absorption capacity of the sisal fibers and incorporation of air during the mixing process. These factors contribute to an overall increase in the total porosity of the composite and consequently enhance the diffusion of CO₂. A previous study [9] indicated that, owing to the hydrophilic nature and pronounced porosity of plant fibers, composites containing them exhibit a heightened ability to encapsulate carbon dioxide. This phenomenon not only preserves the integrity of the fibers, but also prevents their degradation in an alkaline environment, thereby maintaining the overall integrity of the composite.

The carbonation results obtained in this study were consistent with the observations made by Hu et al. [32] and Pavithra et al. [33], who also noted an increase in the carbonate layer due to the presence of plant fibers. These researchers concluded that the advancement of carbonation is facilitated by the resulting microstructures and physical properties of these materials. The heightened porosity at the fiber/cementitious matrix interface, as illustrated in Figure 14, promoted the passage of CO₂, contributing to the observed increase in the carbonate layer.

The progression of the carbonate front, attributed to the elevated fiber content in the composites, underscores the feasibility of utilizing these materials as permanent formworks. This phenomenon holds promise for mitigating the long-term degradation of plant fibers, primarily because of the alkalinity of the matrix, which can reduce the mechanical strength of the composite. Another mechanism of degradation involves mineralization of the lumen of the sisal fibers, driven by the migration of free calcium hydroxides. As noted by Amaral et al. [27], natural fibers establish a porous network that facilitates moisture distribution in the cementitious matrix. This, in turn, promotes delayed hydration reactions, allowing for crack control, carbonation reactions, and an overall improvement in the bearing capacity of the composites.

The carbonation depth results presented in Figure 15 are related to the time to calculate the accelerated carbonation coefficient (K_acel), obtained by means of the mathematical model presented in Equation (5). The correlation curves are shown in Figure 15.

The concrete cover (CC) exhibited a lower accelerated carbonation coefficient (9.05 mm/year^0.5) compared to the composites, where the carbonation rate increased with higher fiber content. The reference composite (F0) and those containing 2% (F2) and 4% (F4) of fibers demonstrated similar K_acel values, measuring 11.46 mm/year^0.5, 10.06 mm/year^0.5, and 12.51 mm/year^0.5, respectively. However, the composite with 6% fibers (F6) showed significantly poorer performance than the others, with a K_acel value of 29.11 mm/year^0.5.

While analyzing accelerated carbonation coefficients provides valuable insights into material quality, establishing correlation indices (x′) between carbonation coefficients determined by accelerated tests (K_acel) and those under natural conditions (K_nat) of exposure, as per Equation (9), adds another layer of meaningful information.

$$x^{\prime }={\displaystyle \frac{K_{acel}}{K_{nat}}}$$

(9)

However, studies of this nature are infrequent in the literature, primarily because of the extensive duration of the tests. Ribeiro et al. [57] conducted an extensive review and compiled data from various studies conducted worldwide, resulting in 117 correlation indices (x′). By leveraging this analysis and considering the classification of natural environments proposed by Gao et al. [58], which is based on the CO₂ concentration values, Ribeiro et al. [57] introduced the correlation indices listed in

Table 8. Correlation indices (x′) between carbonation coefficients obtained through accelerated tests and under natural exposure conditions proposed by [57].

Environment	Correlation Index (x′) Between Kacel. and Knat.
Rural or Urban Light (%CO2 < 0.04)	8.0
Medium Urban (0.04 ≤ %CO2 < 0.06)	7.5
Urban Heavy * or Light Industrial (0.06 ≤ %CO2 < 0.10)	7.0
Heavy Industrial ** (0.10 ≤ %CO2 < 0.30)	3.0

* High population density with heavy vehicle traffic. ** Specific urban sites, such as tunnels and garages (or parking lots), can have CO₂ concentrations equivalent to those in industrial areas.

.

Utilizing the correlation indices outlined in Table 8, estimates were made for the natural carbonation coefficients (K_nat) and anticipated shelf life, assuming a covering thickness of 30 mm. The results, as presented in

Table 9. Carbonation coefficients obtained through accelerated tests in addition to the useful life (UL), considering a cover thickness of 30 mm.

Environment	Mixture	Kacel. (mm/Year−0.5)	Knat. (mm/Year−0.5)	UL (Years)
Rural or Urban Light (%CO2 < 0.04)	F0	11.46	1.43	438.56
	F2	10.06	1.26	569.39
	F4	12.51	1.56	368.28
	F6	29.11	3.64	67.95
	CC	9.05	1.13	703.80
Medium Urban (0.04 ≤ %CO2 < 0.06)	F0	11.46	1.53	385.45
	F2	10.06	1.34	500.44
	F4	12.51	1.67	323.69
	F6	29.11	3.88	59.72
	CC	9.05	1.21	618.57
Urban Heavy or Light Industrial (0.06 ≤ %CO2 < 0.10)	F0	11.46	1.64	335.77
	F2	10.06	1.44	435.94
	F4	12.51	1.79	281.97
	F6	29.11	4.16	52.02
	CC	9.05	1.29	538.84
Heavy Industrial (0.10 ≤ %CO2 < 0.30)	F0	11.46	3.82	61.67
	F2	10.06	3.35	80.07
	F4	12.51	4.17	51.79
	F6	29.11	9.70	9.56
	CC	9.05	3.02	98.97

, highlight that concrete, when used as a cover, exhibits a greater capacity to protect reinforcement against carbonation than structures incorporating fiber-reinforced composites as part of the protective layer. Considering the prescribed minimum design life of 50 years [59] and a minimum thickness of 30 mm in environments with a moderate aggressiveness class [60], it is evident that all composites surpass the limits stipulated by Brazilian standards.

4. Conclusions

This study demonstrates the technical feasibility of using sisal-fiber-reinforced cement composites as permanent formworks, offering a sustainable alternative to conventional concrete cover. The tests performed are limited to standard specimens containing the material proposed as the permanent formwork. The composite with 2% fiber content (F2) exhibited the lowest porosity and mechanical properties closest to conventional concrete cover, making it suitable for structural elements subjected to axial loads, such as columns. In contrast, the 6% fiber composite (F6) showed better performance for beam formwork due to its improved stress transfer and crack control under bending.

Composites containing 2% and 4% fibers showed the best durability performance, ensuring a service life exceeding 50 years even in severe industrial environments. Although higher fiber contents increase porosity and CO₂ diffusion, the F2 composite reduced the carbonated area by 22.8% compared with the reference matrix. Carbonation may also contribute to the preservation of natural fibers by reducing matrix alkalinity, thereby limiting alkaline degradation.

Despite higher carbonation rates observed for composites with 4% and 6% fibers, formulations F2 and F4 maintained adequate durability. The F6 composite, however, showed a higher carbonation rate and did not meet durability requirements in several scenarios. Overall, sisal–cement composites, particularly F2 and F4, represent sustainable and technically viable materials for permanent formwork applications in reinforced concrete structures.

It is important to highlight that the carbonation analysis of this study is based mainly on accelerated testing, using generalized correlation coefficients that are not yet sufficiently validated for sisal fiber composites. Further studies should still be conducted to more accurately evaluate the long-term durability of structures containing this material.