Mechanical Performance, Crack Resistance and Microstructural Evolution of Engineered Cementitious Composites Reinforced with Multiscale Hybrid Fibers

1. Introduction

Engineered cementitious composites (ECC) are a class of high-ductility cement-based composites developed based on micromechanical design theory [1]. They typically exhibit a tensile strain capacity of 3-8%, exceeding that of conventional concrete by more than 300-fold, while exhibiting stable strain-hardening behavior and multiple fine cracking under uniaxial tension [2]. Unlike conventional concrete, which typically fails through the formation of a single dominant crack, ECC maintains crack widths below 100 μm under loading, thereby significantly reducing the ingress of chloride ions, moisture, and other aggressive agents, and consequently enhancing structural durability and service life [3]. Owing to these advantages, ECC has been widely applied in bridge connections, seismic-resistant structures, rehabilitation and strengthening, and marine engineering, demonstrating significant potential for improving structural toughness, enhancing seismic energy dissipation, and extending infrastructure service life [4,5] Despite its outstanding mechanical performance and durability, the widespread application of ECC remains limited by several challenges. Conventional ECC typically employs polyethylene (PE) or polyvinyl alcohol (PVA) fibers as the primary reinforcement. However, the high cost of these fibers, coupled with their relatively high dosage, substantially increases material costs, thereby limiting large-scale engineering applications. Moreover, the macroscopic mechanical performance of ECC is highly dependent on the fiber-matrix interfacial bond and the crack-bridging capacity of the reinforcing fibers [6]. Although single-fiber systems can achieve excellent strain-hardening behavior, they remain limited in terms of compressive strength, flexural performance, early-age crack resistance, and crack contro [7]. Furthermore, fibers differ considerably in elastic modulus, tensile strength, aspect ratio, and interfacial characteristics, making it difficult for a single fiber type to simultaneously satisfy the multiple requirements of matrix strengthening, crack control, and toughness enhancement. Therefore, developing novel ECC systems that combine high ductility, high strength, and superior crack resistance has become a major focus of current research [8].

In recent years, multiscale hybrid fiber-reinforced ECC has attracted increasing research interest worldwide [9,10]. By integrating fibers with different length scales and mechanical properties, complementary reinforcing mechanisms can be achieved, enabling synergistic crack-bridging during crack initiation, propagation, and ultimate failure, thereby enhancing the toughness of ECC across multiple scales. Previous studies have shown that hybridization of steel fibers, polypropylene (PP) fibers, basalt fibers (BF), polyethylene (PE) fibers, and polyvinyl alcohol (PVA) fibers can improve the compressive strength, tensile strength, flexural toughness, and crack control capability of ECC to varying degrees [11,12]. Among these fibers, PP fibers are characterized by low density, low cost, and excellent chemical resistance, making them effective in suppressing plastic shrinkage cracking and improving early-age crack resistance [13]. BF possesses a high elastic modulus and tensile strength, together with excellent interfacial bonding with cementitious matrices, thereby enhancing the load-carrying capacity and crack-bridging performance [14]. PE fibers, owing to their high strength, high ductility, and superior crack-bridging capability, are indispensable for achieving the strain-hardening behavior of ECC [15]. Lu et al. [16] fabricated concrete beams reinforced with hybrid glass fibers (GF), steel fibers, and PVA fibers. Four-point bending tests combined with digital image correlation (DIC) revealed that steel fibers significantly enhanced the load-carrying capacity and ductility. However, among the hybrid systems investigated, only the steel fiber-GF combination exhibited a positive hierarchical synergistic effect, whereas the remaining hybrid configurations failed to generate effective synergy and instead compromised the integrity of the cementitious matrix. Lian et al. [17] investigated the mechanical behavior of BF-PP fiber-reinforced concrete through experimental testing and numerical simulation. The results showed that fiber incorporation enhanced toughness, while the tensile and compressive strengths first increased and then decreased with increasing fiber content, indicating the existence of an optimum dosage. Acoustic emission analysis further revealed the crack evolution process, confirming that BF and PP fibers effectively suppressed crack initiation and propagation, with higher fiber contents resulting in greater improvements in residual strength. Zhou et al. [18] conducted axial pull-out fatigue tests on BF-steel fiber composite specimens and reported that incorporating 0.5% and 1.0% steel fibers increased the fatigue life by 215.44% and 334.72%, respectively. Piotr S. et al. [19] reported that hybrid BF-PP fibers significantly improved the workability, compressive strength, fracture energy, and splitting energy absorption capacity of cementitious composites.

With the development of nanomaterials, carbon nanotubes (MWCNTs) have been widely utilized for fiber surface modification due to their ultrahigh elastic modulus, outstanding mechanical properties, and excellent interfacial regulation capability [20,21]. He et al. [22] coated carbon nanofibers (CNFs) onto PE fibers via hydrophobic interactions, increasing the tensile strength and strain capacity of ECC by 15.0% and 20.0%, respectively. Lu et al. [23] developed a GO-coated PE fiber modification strategy, in which the wrinkled structure and oxygen-containing functional groups of GO enhanced both mechanical interlocking and chemical bonding at the fiber-matrix interface. Consequently, the tensile strength and strain capacity of ECC increased by 46.3% and 70.4%, respectively, while the average crack width decreased from 138μm to 58μm. Previous studies have shown that MWCNTs increase fiber surface roughness, enhance mechanical interlocking with the cementitious matrix, and facilitate the deposition of hydration products at the fiber interface, thereby improving interfacial bonding and fiber bridging capacity. These effects provide a promising approach for enhancing ECC performance [24]. However, the development of multiscale hybrid fiber systems combining MWCNT-modified PE fibers with PP and basalt fibers remains relatively unexplored. In particular, the influence of fiber hybridization on strain-hardening behavior, crack evolution, and microscale reinforcement mechanisms remains insufficiently understood [25]. Therefore, this study employs carbon nanotube-modified polyethylene (M-PE) fibers as the primary reinforcement and incorporates PP and basalt fibers to develop a multiscale hybrid fiber-reinforced ECC system. The effects of different fiber combinations on the compressive strength, uniaxial tensile behavior, flexural performance, and early-age crack resistance of ECC were systematically investigated. Moreover, scanning electron microscopy (SEM), X-ray diffraction (XRD), and mercury intrusion porosimetry (MIP) were conducted to reveal microstructural evolution and clarify the synergistic toughening mechanisms of the multiscale fiber-reinforced ECC. This study provides theoretical insights for the design and engineering application of high-performance multifunctional ECC materials.

2. Experimental Materials and Methods

2.1. Materials

Portland cement (P·O 42.5) was used as the primary binder in this study. Natural quartz sand supplied by Fengping Mineral Products Co., Ltd. was selected as the fine aggregate, with an apparent density of 2580kg/m³, a fineness modulus of 1.5, and a maximum particle size of 0.60mm. Class F fly ash (Grade II) from Beilun Power Plant was incorporated as a mineral admixture to improve fresh-state workability and refine the matrix microstructure. A polycarboxylate-based superplasticizer from Ningbo Zhongshuike Chemical Technology Co., Ltd., with a water-reducing rate of 25%, was used to maintain mixture flowability. Nano-silica (Nano-SiO₂) from Shanghai Macklin Biochemical Technology Co., Ltd. was added to accelerate cement hydration and enhance matrix densification. The main physical properties and chemical compositions of the raw materials are summarized in Table 1.

The reinforcing fibers used in this study included polyethylene (PE), basalt (BF), and polypropylene (PP) fibers, with their surface morphologies and microstructures presented in Figure 1. Among them, PE fibers modified with multi-walled carbon nanotubes (MWCNTs) were used as the primary reinforcement to enhance fiber-matrix interfacial bonding and load-transfer efficiency [26].

PE fiber modification was performed using an MWCNT suspension impregnation method. First, PE fibers were ultrasonically cleaned for 10 min to remove surface contaminants and improve surface cleanliness. The cleaned fibers were then immersed in an MWCNT suspension with a concentration of 4kg/m³ and subjected to intermittent mechanical stirring at 100r/min (15 min stirring followed by 5min resting cycles). This process facilitated uniform MWCNT adsorption on the fiber surface while minimizing fiber entanglement and agglomeration caused by continuous stirring. After 8 h of impregnation, the fibers were removed, drained, and dried in a forced-air oven at 80 °C for 24h. The resulting MWCNT-modified polyethylene fibers were denoted as M-PE [27].

2.2. Specimen Preparation

During sample preparation, cement, fly ash, standard sand, and other dry constituents were first added to a mortar mixer according to the designed mix proportions and dry-mixed to ensure uniform dispersion. Water and superplasticizer were then added, followed by further mixing to obtain a homogeneous mixture with appropriate flowability. Subsequently, fibers were gradually introduced in batches and mixed at high speed to minimize fiber agglomeration and improve fiber dispersion within the matrix.

The prepared ECC specimens were cured at room temperature (20±2 °C) for 24h before demolding and then transferred to a standard curing environment (20±2 °C, relative humidity ≥95%) for further curing. The mix proportions of ECC mixtures are listed in Table 2.

In this study, nano-silica (NS) was incorporated at 0.1 wt.% of the binder. The mixtures containing M-PE/PP and M-PE/BF hybrid fibers were designated as the PP and BF series, respectively. The detailed mix proportions are listed in Table 3.

2.3. Test Methods

2.3.1. Uniaxial Compression Test

Cube specimens (40mm×40mm×40mm) were used for compressive strength testing. After curing under standard conditions to the designated ages, the specimens were tested using a STYE-300E automatic cement flexural and compressive testing machine under continuous loading at a rate of 2.4kN/s. Three replicate specimens were tested for each mixture, and the average value was reported.

2.3.2. Uniaxial Tensile Test

Uniaxial tensile tests were performed on dumbbell-shaped specimens (330mm×60mm×30 mm) using a WA-50 electronic universal testing machine under displacement-controlled loading at a rate of 0.5mm/min. A YYU-80-25-SH extensometer was attached to the central gauge section to continuously monitor tensile deformation, and the test setup is illustrated in Figure 2. Tensile stress and strain were calculated from the measured load and displacement data, and the corresponding stress-strain curves were obtained to evaluate the strain-hardening behavior and tensile ductility of ECC.

2.3.3. Three-Point Bending Test

Three-point bending tests were conducted on prismatic specimens (40mm×40mm×160mm) after curing under standard conditions to the designated ages. The specimens were tested using a WA-50 electronic universal testing machine under displacement-controlled loading at a rate of 0.5mm/min, with a support span of 100mm. Load-deflection curves were recorded to evaluate the flexural performance and deformation capacity of ECC. Three replicate specimens were tested for each mixture, and the average value was reported.

2.3.4. Plate Crack Resistance Test

The plate cracking test was conducted using a steel restraint mold with internal dimensions of 600mm×600mm×63mm. A fan was used to accelerate surface water evaporation and induce plastic shrinkage cracking. Crack widths were measured using an MG10085-1 150× illuminated reading microscope, while crack lengths and numbers were recorded to evaluate the early-age cracking resistance of ECC.

2.3.5. Pore Structure Testing (MIP)

Hardened paste samples were cut into approximately 1cm³ specimens, vacuum-dried at 60 °C to constant weight, and subjected to mercury intrusion porosimetry analysis. The test pressure ranged from 0.1 to 60,000psi, and pore size distribution, total pore volume, and porosity were determined using the Washburn equation to characterize the pore structure.

2.3.6. Scanning Electron Microscope Analysis (SEM)

Fresh fracture surfaces of hardened paste samples were obtained, vacuum-dried at 60 °C to constant weight, and sputter-coated with gold for scanning electron microscopy observation. The matrix morphology and fiber-matrix interfacial features were observed under high-vacuum conditions at an accelerating voltage of 15kV. SEM images were used for qualitative morphological comparison only; phase identity, hydration-product content, and interfacial bond strength were not determined solely from image morphology.

2.3.7. X-Ray Diffraction Analysis (XRD)

Specimens cured for 28 days were selected, and hydration was stopped using anhydrous ethanol. The samples were subsequently ground and sieved through a 75μm mesh to obtain XRD powders. X-ray diffraction analysis was conducted over a 2θ range of 5-90° with a scanning rate of 5°/min. Crystalline phases were identified qualitatively from diffraction peak positions and characteristic patterns. No quantitative phase analysis was performed, and poorly crystalline C-S-H was not quantified from XRD peak intensity.

3. Experiment Results and Analysis

3.1. Mechanical Property Analysis

The effects of hybrid fiber content and proportion on the mechanical properties of fiber-reinforced cementitious composites were investigated through uniaxial compressive, uniaxial tensile, and three-point bending tests on M-PE fiber-reinforced and hybrid fiber-reinforced specimens. The test results are presented in Table 4.

3.1.1. Uniaxial Compressive Strength

Figure 3 presents the uniaxial compressive strength of the specimens. The 28-day compressive strength showed a slight increase compared with the 7-day values. For the M-PE/PP hybrid fiber system, introducing an appropriate amount of PP fibers effectively enhanced the compressive strength while maintaining M-PE fibers as the dominant reinforcement. Specifically, compared with the M-PE group, the 7-day compressive strengths of the PP-0, PP-1, and PP-2 groups increased by 4.5%, 10.7%, and 2.3%, respectively, while the corresponding 28-day increases were 7.2%, 13.4%, and 4.9%. The PP-1 group exhibited the highest compressive strength, reaching 92.0MPa at 28 days, which was 13.4% higher than that of the M-PE group. This improvement can be attributed to the multiscale synergistic reinforcement provided by hybrid fibers. The M-PE fibers modified with carbon nanotubes exhibited increased surface roughness, facilitating the deposition of hydration products, enhancing interfacial mechanical interlocking, and improving stress transfer efficiency. Meanwhile, the smaller size and better dispersion of PP fibers enabled them to fill localized defects and restrain microcrack propagation, thereby improving matrix integrity and compressive performance. However, excessive PP incorporation resulted in a gradual reduction in compressive strength. When the M-PE fiber content decreased below 60% of the total fiber volume fraction, the 28-day compressive strengths of the PP-3, PP-4, and PP-5 groups decreased by 3.0%, 6.5%, and 18.5%, respectively, compared with the M-PE group. This reduction is mainly attributed to the lower elastic modulus and weaker interfacial bonding capability of PP fibers compared with M-PE fibers. Excessive replacement of M-PE fibers by PP fibers weakened the load-transfer capacity of the fiber network, thereby reducing the reinforcing efficiency [6].

The M-PE/BF hybrid fiber system exhibited a similar trend to the PP-based system but achieved a more pronounced improvement in compressive strength. With a high proportion of M-PE fibers, BF incorporation further enhanced the compressive performance. Specifically, the 28-day compressive strengths of the BF-0, BF-1, and BF-2 groups increased by 5.2%, 15.7%, and 7.5%, respectively, compared with the M-PE group. The BF-1 group exhibited the highest compressive strength (93.8MPa), representing a 15.7% increase over the M-PE group. This enhancement indicates that an appropriate BF content can establish an effective multiscale reinforcing network with M-PE fibers. Due to their high elastic modulus and hydrophilic characteristics, BF fibers promote the deposition of hydration products on the fiber surface and enhance interfacial bonding with the cementitious matrix, thereby improving the load-bearing capacity of the interfacial transition zone. In addition, BF fibers effectively restrain crack initiation and propagation, reducing damage to matrix integrity and contributing to the enhanced compressive strength [28]. However, further increasing the BF content and replacing more M-PE fibers led to a gradual reduction in compressive strength. Although BF fibers exhibit high strength and elastic modulus, their lower ductility and inferior crack-bridging capacity compared with M-PE fibers limit their ability to provide sustained load transfer during crack propagation.

3.1.2. Uniaxial Tensile Properties

Figure 4 presents the uniaxial tensile strength of all specimens, while Figure 5a,b illustrate the 28-day tensile stress-strain curves of the M-PE/PP and M-PE/BF hybrid fiber systems, respectively. Overall, the 28-day tensile strengths exceeded those at 7 days, with the 7-day values reaching approximately 84% of the corresponding 28-day strengths. This improvement is attributed to prolonged hydration, which generates additional C-S-H gels, densifies the fiber-matrix interface, and enhances interfacial bonding and stress-transfer efficiency.

For the M-PE/PP hybrid system, when the M-PE fiber content was maintained, the incorporation of a small amount of PP fibers enhanced tensile strength. Specifically, the 28-day tensile strength of the PP-0 group increased by 7.6% compared with the M-PE group. This enhancement is mainly attributed to the synergistic interaction between PP and M-PE fibers, where PP fibers alleviate stress concentration at crack tips, delay microcrack propagation, and improve post-cracking load-carrying capacity. However, further increasing the PP fiber proportion resulted in continuous decreases in both tensile strength and ultimate tensile strain, particularly when the M-PE fiber content fell below 60%. Compared with the M-PE group, the PP-5 group exhibited reductions of 51.5% and 86.6% in 28-day tensile strength and ultimate tensile strain, respectively. This deterioration can be explained by the lower elastic modulus and weaker interfacial bonding of PP fibers, which facilitate interfacial debonding and fiber pull-out during crack propagation, thereby reducing the sustained bridging effect. Furthermore, excessive replacement of M-PE fibers decreases the fraction of high-ductility bridging fibers, promoting premature crack localization and resulting in concurrent losses in strength and ductility.

The M-PE/BF hybrid system exhibited a similar trend but achieved a more pronounced strengthening effect. The 28-day tensile strength of the BF-0 group increased by 23.7% compared with the M-PE group, reaching 7.35MPa, the highest value among all specimens. This improvement is attributed to the high elastic modulus and excellent hydrophilicity of BF fibers, which facilitate stronger fiber-matrix interfacial bonding. During microcrack initiation, BF fibers effectively bear tensile stress and dissipate energy through fiber fracture and pull-out, thereby delaying M-PE fiber damage and enhancing the overall tensile load-bearing capacity. However, further increasing the BF content resulted in gradual reductions in both tensile strength and strain. Although the BF-5 group exhibited only a 13.1% decrease in tensile strength, its ultimate tensile strain decreased by 85.8%, indicating that BF fibers can enhance strength but are insufficient to maintain stable crack bridging due to their limited elongation capacity, leading to a more brittle failure mode.

Overall, M-PE fibers act as the dominant reinforcing phase controlling the tensile performance of ECC. When the M-PE fiber volume fraction was ≥60%, PP and BF fibers served as secondary reinforcements by optimizing stress redistribution during crack initiation. Specifically, PP fibers mainly contributed to energy dissipation through frictional pull-out, whereas BF fibers improved microcrack-stage load-bearing capacity through their higher stiffness and stronger interfacial bonding. Therefore, the BF hybrid system provided a more effective strengthening effect than the PP system. With further reduction in M-PE content, the stable crack-bridging network became difficult to sustain, resulting in rapid degradation of tensile performance. As shown in Figure 5, when the M-PE fiber content was ≥60%, all stress-strain curves exhibited the typical three stages of linear elasticity, strain hardening, and strain softening. In the linear elastic stage, the load was primarily carried by the cementitious matrix. After matrix cracking, M-PE fibers bridged the cracks and continuously transferred stress, inducing stable multiple cracking and initiating the strain-hardening stage. After reaching the peak stress, crack localization occurred gradually, accompanied by progressive fiber pull-out or fracture, resulting in stress decay and transition to the strain-softening stage. Conversely, when the M-PE fiber content was <60%, the insufficient amount of high-ductility bridging fibers led to rapid crack localization after cracking, eliminating the strain-hardening response. These results indicate that M-PE fibers govern the multiple-cracking capability and steady-state crack propagation of ECC, whereas PP and BF fibers mainly provide microcrack control and supplementary energy dissipation. The synergistic interaction among the three fibers establishes an effective multiscale hybrid reinforcement mechanism [29].

3.1.3. Three-Point Bending Performance

Figure 6 presents the three-point flexural strength of all specimens, while Figure 7a,b illustrate the 28-day load-deflection curves of the M-PE/PP and M-PE/BF hybrid fiber-reinforced specimens, respectively. Overall, the 28-day flexural strengths were higher than those at 7 days, with the 7-day values reaching approximately 79% of the corresponding 28-day strengths. This enhancement is attributed to prolonged hydration, which generates additional C-S-H gels, densifies the fiber-matrix interface, and improves interfacial stress-transfer efficiency, thereby enhancing the flexural load-bearing capacity of the composites.

The 7-day flexural strength was approximately 79% of the corresponding 28-day value, indicating that prolonged curing enhanced the flexural performance. For the M-PE/PP hybrid fiber system, compared with the M-PE group, the 7-day flexural strengths of the PP-0, PP-1, PP-2, PP-3, and PP-4 groups increased by 14.1% and decreased by 0.9%, 19.4%, 30.1%, and 41.8%, respectively, accompanied by reductions in deflection of 10.0%, 22.7%, 40.8%, 56.2%, and 58.5%, respectively. At 28 days, the corresponding flexural strengths increased by 4.2% and decreased by 9.4%, 27.0%, 37.1%, and 47.5%, while the deflections decreased by 10.7%, 25.1%, 44.9%, 59.9%, and 65.8%, respectively. The PP-5 group, reinforced solely with PP fibers, exhibited reductions of 51.8% and 56.4% in 7-day and 28-day flexural strength, respectively, compared with the M-PE group, with corresponding deflection reductions of 69.6% and 74.9%.

For the M-PE/BF hybrid fiber system, compared with the M-PE group, the 7-day flexural strengths of the BF-0, BF-1, BF-2, BF-3, and BF-4 groups increased by 16.4% and decreased by 2.2%, 11.4%, 34.0%, and 44.6%, respectively. Meanwhile, the corresponding deflections decreased by 7.0%, 24.4%, 44.1%, 62.9%, and 82.9%, respectively. At 28 days, the flexural strengths of these groups increased by 5.5% and decreased by 6.3%, 18.9%, 39.5%, and 49.4%, respectively, accompanied by reductions in deflection of 8.2%, 26.6%, 41.2%, 65.3%, and 83.9%, respectively. The BF-5 group, reinforced solely with BF fibers, exhibited reductions of 46.6% and 51.5% in 7-day and 28-day flexural strengths, respectively, compared with the M-PE group, while the corresponding deflections decreased by 90.6% and 91.2%. When the M-PE fiber content was maintained, the incorporation of PP or BF fibers in the PP-0 and BF-0 groups enhanced the three-point flexural strength but reduced the deformation capacity. With decreasing M-PE fiber content, the flexural strength of the hybrid fiber-reinforced specimens gradually declined. When the M-PE fiber content was ≥60%, M-PE fibers remained the dominant reinforcing phase, and the specimens exhibited pronounced strain-hardening behavior. However, strain hardening disappeared when the M-PE fiber content decreased below 60%. These results indicate that M-PE fibers play a critical role in determining the three-point flexural performance of fiber-reinforced cementitious composites.

3.2. Crack Resistance Analysis

3.2.1. Early Crack Resistance

The effects of fiber type, dosage, and hybridization ratio on the early-age cracking resistance of cementitious composites were investigated through plate cracking tests on M-PE fiber-reinforced and hybrid fiber-reinforced specimens. The results are presented in Table 5.

The evolution of crack number, maximum crack width, crack length, and nominal total crack area during the plate cracking test is summarized in Table 5 and Figure 8. For all mixtures, the crack number, maximum crack width, and total crack length continuously increased with increasing curing age, consistent with the typical early-age shrinkage cracking behavior of cementitious materials. During early hydration, the combined effects of chemical shrinkage, drying shrinkage, and thermal shrinkage generate tensile stresses within the restrained specimens. When these stresses exceed the tensile strength of the matrix, cracks initiate and progressively propagate and coalesce with ongoing hydration, eventually forming a stable crack network. The M-PE reference group exhibited the most severe cracking behavior, with crack initiation occurring as early as 1 day. At 7 days, this group developed 18 cracks, with a maximum crack width of 0.43mm and a total crack length of 372.1mm. Its crack propagation rate was also significantly higher than those of the hybrid fiber-reinforced groups. In contrast, the hybrid fiber systems generally showed delayed crack initiation, slower crack propagation, and reduced final cracking severity, confirming the effectiveness of hybrid fiber reinforcement in mitigating early-age shrinkage cracking of cementitious composites [30].

The cracking resistance of the PP series gradually improved with increasing PP fiber content. The PP-0 group showed only slight reductions in crack width and total crack length compared with the M-PE reference group, while the crack number remained relatively high, indicating limited improvement in cracking resistance. With further increases in PP fiber content, all crack-related parameters exhibited a marked decreasing trend at different curing ages. The PP-2 group achieved effective control of maximum crack width, with a value of only 0.32mm at 7 days, representing a 25.58% reduction compared with the reference group and indicating effective suppression of wide cracks. The PP-5 group exhibited the best overall cracking resistance among the PP-based systems. At 7 days, its crack number and total crack length decreased by 38.89% and 35.15%, respectively, compared with the reference group. In addition, no cracks were observed at 1 day, and only two cracks formed at 3 days, demonstrating that PP fibers effectively delayed crack initiation and reduced the risk of early-age cracking.

The BF system exhibited a more pronounced improvement in cracking resistance, following a trend of initial enhancement followed by a slight decline with increasing BF content. Although the BF-0 group showed limited improvement, further BF incorporation significantly reduced the crack-related parameters. At 7 days, the crack numbers of the BF-4 and BF-5 groups decreased to 11 and 10, respectively, corresponding to reductions of 38.89% and 44.44% compared with the reference group. Among all mixtures, the BF-4 group exhibited the most effective crack-length control, with a total crack length of only 188.9mm at 7 days, representing a 49.23% reduction compared with the reference group. Furthermore, no cracks were detected at 1 day, and only one crack developed at 3 days, indicating excellent early-age cracking resistance. Overall, both PP and BF fibers enhanced cracking resistance by bridging cracks, redistributing shrinkage stresses, and restricting crack propagation. However, BF fibers exhibited superior crack-control performance compared with PP fibers, mainly due to their stronger interfacial bonding with the cementitious matrix, which enables more efficient stress transfer and crack restraint.

Table 6 presents the cracking resistance grades of the specimens at 7 days calculated using the corresponding equation. The M-PE reference group exhibited the largest nominal total crack area (86.2mm²) among all mixtures, indicating the most severe cracking behavior. Based on this group as the reference, the crack reduction coefficient of each mixture was calculated to quantitatively assess the improvement in cracking resistance provided by fiber modification. For the PP fiber series, the nominal total crack area continuously decreased with increasing PP fiber content, demonstrating a progressive improvement in cracking resistance. The PP-0 group exhibited a 24.25% reduction compared with the reference group, indicating only limited enhancement. With further increases in PP fiber content, the nominal total crack area of the PP-1 to PP-4 groups decreased from 61.5mm² to 41.1mm², corresponding to reductions of 28.65%-52.32% relative to the M-PE group, suggesting an enhanced crack-bridging effect of PP fibers. The PP-5 group showed the best cracking resistance among the PP-based systems, with the nominal total crack area reduced to 36.2mm², representing a 58.00% decrease compared with the reference group. This substantial reduction in cracking severity agrees well with the improved control of crack number, crack length, and crack width observed in the plate cracking test.

The nominal total crack area of the BF fiber series exhibited a trend of slight fluctuation followed by a pronounced decrease with increasing BF content, indicating superior overall crack-control performance compared with the PP fiber system. The BF-0 group showed a nominal total crack area of 63.1mm², representing a 26.80% reduction relative to the reference group and indicating limited improvement in cracking resistance. In the BF-1 group, the nominal total crack area slightly increased to 66.3mm², while the crack reduction coefficient decreased to 23.09%. This fluctuation was attributed to inadequate fiber dispersion and localized fiber agglomeration, which introduced initial defects within the matrix. With further increases in BF content, a continuous three-dimensional fiber-bridging network was gradually formed, leading to a significant enhancement in crack-bridging efficiency. The nominal total crack areas of the BF-2 to BF-5 groups decreased continuously from 54.2mm² to 27.6mm², corresponding to reductions of 37.12%-67.98% compared with the reference group, demonstrating a substantial improvement in cracking resistance. Among all specimens, the BF-5 group exhibited the best crack-control performance, with a nominal total crack area of only 27.6mm², representing a 67.98% reduction compared with the reference group.

Due to the relatively high cracking resistance of the M-PE reference group, the PP-0, PP-1, BF-0, BF-1, and BF-2 groups exhibited relatively low crack reduction coefficients and failed to achieve Grade III cracking limitation. The PP-2, PP-3, PP-4, and BF-3 groups reached Grade III cracking limitation, indicating a moderate improvement in the cracking resistance of cementitious composites but insufficient enhancement for achieving a higher grade. In contrast, the PP-5, BF-4, and BF-5 groups achieved Grade II cracking limitation, demonstrating a significant improvement in crack-control performance. Notably, the PP-5 group was the only mixture in the PP fiber system to reach Grade II classification. The crack reduction coefficients of the BF-4 and BF-5 groups were 65.78% and 67.98%, respectively, both substantially exceeding the critical threshold for Grade II cracking limitation. This suggests that BF fibers can achieve Grade II cracking resistance at a lower dosage and provide more efficient crack control. The difference in cracking limitation performance between the two fiber systems is mainly attributed to their distinct capabilities in regulating crack initiation and propagation. During the crack initiation stage, both fibers effectively redistribute shrinkage stresses, increase the cracking threshold of the matrix, and delay crack formation, thereby reducing crack occurrence. However, during crack propagation, BF fibers, as inorganic mineral fibers, exhibit better chemical compatibility with cement hydration products, stronger interfacial mechanical interlocking and bonding strength, and a significantly higher elastic modulus than PP fibers. Consequently, BF fibers bridging across cracks can more effectively transfer tensile stresses, mitigate stress concentration at crack tips, and restrict crack width growth and length propagation [31].

3.2.2. Crack Limit Level

The cracking strengths of the specimens in the uniaxial tensile and three-point bending tests are presented in Table 7 and Figure 9.

The cracking strength at 28 days was higher than that at 7 days. With prolonged curing, continued hydration of the cementitious matrix resulted in a denser microstructure and enhanced cracking strength. For the M-PE/PP hybrid fiber system, all mixtures exhibited increased cracking strengths in the uniaxial tensile tests. Compared with the M-PE group, the 7-day and 28-day cracking strengths of the PP-0 to PP-5 groups increased by 54.0% and 52.1%, 47.2% and 56.3%, 49.7% and 39.1%, 46.0% and 43.8%, 34.8% and 25.5%, and 60.2% and 50.0%, respectively. In the three-point bending tests, the cracking strengths of the PP hybrid fiber groups generally decreased, except for the PP-0 group. The 7-day and 28-day cracking strengths of the PP-0 group increased by 3.5% and 5.9%, respectively, whereas those of the PP-1 to PP-5 groups decreased by 10.3% and 11.1%, 1.7% and 3.7%, 8.4% and 8.4%, 12.5% and 12.4%, and 12.6% and 11.8%, respectively, at 7 and 28 days. For the M-PE/BF hybrid fiber system, the cracking strengths obtained from the uniaxial tensile tests showed a more pronounced enhancement. Compared with the M-PE group, the 7-day and 28-day cracking strengths of the BF-0 to BF-5 groups increased by 90.1% and 84.4%, 84.5% and 68.2%, 70.8% and 58.9%, 100.6% and 102.1%, 118.0% and 125.0%, and 162.1% and 168.8%, respectively. The cracking strength increased progressively with increasing BF fiber content, demonstrating the significant reinforcing effect of BF fibers.

In the three-point bending tests, the cracking strengths of the BF-3 and BF-5 groups at 7 days decreased by 9.2% and 10.6%, respectively, while that of the BF-5 group at 28 days decreased by 3.2%. The other groups exhibited increasing trends in cracking strength. Notably, the BF-0, BF-1, and BF-2 groups showed pronounced enhancements, with 7-day and 28-day cracking strengths increasing by 38.3% and 37.7%, 40.4% and 39.2%, and 37.4% and 34.7%, respectively. The results indicate that hybrid fiber incorporation can enhance the cracking strength of cementitious materials under tensile loading. However, the M-PE/PP hybrid fiber system showed limited effectiveness in improving flexural cracking strength, whereas the M-PE/BF hybrid fiber system exhibited a significant enhancement. The superior strengthening effect of BF fibers compared with PP fibers can be attributed to their hydrophilic characteristics, which promote stronger bonding with cement hydration products. In addition, BF fibers possess a higher elastic modulus than PP fibers, resulting in better resistance to deformation under external loading. Consequently, BF fibers can more effectively inhibit microcrack initiation and propagation, thereby improving the cracking strength of cementitious composites.

3.3. Microstructure and Hydration Product Analysis

3.3.1. SEM Analysis

Figure 10 presents representative SEM images of the fracture surfaces of the fiber-reinforced cementitious composites. Because SEM primarily provides morphological information, the images are interpreted qualitatively and in conjunction with the MIP and XRD results. In Figure 10(a), fine particulate features and hydration-related products are visible within the matrix. These features cannot be unambiguously assigned to nano-silica (NS) from morphology alone; therefore, the image is used to describe matrix morphology rather than to identify or quantify specific hydration products.

Figure 10(b) shows the PP fiber-reinforced specimen. The PP fiber surface appears comparatively smooth and contains less visible attached matrix material in the representative field of view. Several local voids and pull-out-related cavities are also visible. These observations are consistent with comparatively weak local PP-matrix contact, but they are not used to quantify the overall porosity; pore-structure differences are evaluated using MIP in Section 3.3.2.

Figure 10(c) presents the BF fiber-reinforced specimen. Compared with the PP fiber shown in Figure 10(b), more matrix material is visibly retained on the BF surface in the representative image, suggesting closer local contact between the BF and cementitious matrix. However, SEM morphology alone cannot determine the chemical identity or amount of the attached products, nor can it quantify bond strength or bulk porosity. The corresponding pore-structure trends are therefore discussed together with the MIP results.

Figure 10(d) illustrates the PP/M-PE hybrid fiber-reinforced specimen. Matrix residues are visibly retained on portions of the M-PE fiber surface, whereas some PP fibers are associated with pull-out cavities. These features indicate differences in local interface morphology and failure mode between the two fibers. The SEM images do not directly quantify interfacial bond strength; rather, they provide qualitative morphological support for the distinct load-transfer and pull-out behaviors discussed from the mechanical response [32].

Figure 10(e) presents the M-PE/BF hybrid fiber-reinforced specimen. Matrix residues are visible on both M-PE and BF surfaces, and fractured BF segments can be observed in the representative field. This morphology is consistent with the coexistence of fiber pull-out and fracture during crack development. When considered together with the tensile and flexural responses, the images support a complementary role in which BF contributes to early-stage crack restraint while M-PE provides sustained crack bridging at larger deformation. Nevertheless, these SEM observations remain qualitative, and no quantitative interfacial bonding parameters or hydration-product contents are inferred from the micrographs.

3.3.2. Pore Structure Analysis

Based on the improvements in mechanical performance and cracking resistance of the hybrid fiber-reinforced cementitious composites, selected specimens were subjected to mercury intrusion porosimetry (MIP) testing to characterize their pore structures. In the experimental design, the total fiber contents of the PP-0 and BF-0 groups were 1.8%, while those of the PP-1 and BF-1 groups were 1.5%. The pore size distribution curves and cumulative pore volume curves were obtained, and the results are presented in Figure 11.

Figure 11(a) presents the pore size distribution curves. The peak position represents the most probable pore diameter, while the peak intensity reflects the relative intrusion volume within the corresponding size range. All specimens exhibited a unimodal distribution, with the most probable pore diameters concentrated within 20-50 nm. Relative to the M-PE group, the PP-1 and BF-1 groups showed higher peak intensities, and within each hybrid system the 1.8 vol.% fiber groups (PP-0 and BF-0) showed higher peaks than the corresponding 1.5 vol.% groups (PP-1 and BF-1). At matched total fiber content, the PP-based specimens also exhibited higher peak intensities than the corresponding BF-based specimens. These are direct MIP observations and demonstrate that fiber type and total fiber content are associated with changes in the pore-size distribution. Possible contributions from fiber dispersion, interfacial defects, and differences in fiber-matrix affinity [33]. may explain these trends; however, MIP does not spatially distinguish bulk-matrix pores from fiber-matrix interfacial pores, and the specific interfacial mechanisms should therefore be regarded as interpretations rather than direct measurements.

Figure 11(b) presents the cumulative pore volume distribution curves. The measured total intrusion volumes followed the order PP-0>BF-0>PP-1>BF-1>M-PE. Thus, all selected hybrid-fiber groups exhibited higher total intrusion volumes than the M-PE reference, while the 1.8 vol.% fiber groups showed higher values than the corresponding 1.5 vol.% groups. At the same total fiber content, the BF-based specimens exhibited lower total intrusion volumes than the PP-based specimens. These quantitative MIP results indicate that both fiber type and total fiber content affected the accessible pore structure. The lower intrusion volume of the BF-based mixtures is consistent with the comparatively denser local interface morphology observed by SEM; however, the present measurements do not independently quantify hydration degree, fiber dispersion, or the extent of any secondary chemical reaction. Accordingly, these factors are discussed only as possible contributors to the measured pore-structure differences.

3.3.3. Analysis of Hydration Products

As shown in Figure 12, the XRD patterns of the PP and BF series specimens were compared with that of the M-PE reference group. All specimens exhibited similar principal diffraction-peak positions, and no additional crystalline peaks were detected after fiber hybridization. The main crystalline phases identified included SiO₂, Ca(OH)₂, and AFt. Importantly, SiO₂ is not a cement hydration product; the SiO₂ reflections are attributed to crystalline silica-bearing constituents remaining in the composite, such as quartz/residual silica. In addition, C-S-H is poorly crystalline and cannot be reliably quantified from sharp conventional XRD peaks. Therefore, the XRD results are interpreted as showing no new detectable crystalline phases caused by fiber hybridization, rather than as evidence that all hydration-product contents are unchanged.

Compared with the M-PE group, the PP-0 and PP-1 patterns showed lower intensities for several crystalline reflections, whereas the BF-0 and BF-1 patterns were generally closer to the M-PE reference. Because no internal standard or quantitative phase analysis was used, these peak-intensity differences are interpreted qualitatively and are not used to calculate hydration degree or phase contents. The weaker PP-matrix interfacial affinity reported for hydrophobic PP fibers [34] may contribute to the observed differences, while the hydrophilic BF surface may favor closer local contact with the matrix. Although BF contains SiO₂- and Al₂O₃-bearing components and may exhibit limited surface reactivity, the present XRD data do not directly demonstrate a secondary pozzolanic reaction or quantify additional C-S-H formation. The similar peak positions within each hybrid system indicate only that changing the fiber ratio did not produce new detectable crystalline phases.

3.4. Economic Benefit Analysis

Compared with unmodified PE fiber-reinforced cementitious composites, hybrid fiber-reinforced composites incorporating 20% and 40% PP or BF fibers as partial replacements for M-PE fibers maintained over 80% of the performance of the unmodified PE fiber composites. Based on current market prices, the costs of PE, PP, and BF fibers are 200,000, 8,500, and 30,000CNY/t, respectively, while the modification cost of PE fibers is approximately 25,000CNY/t. Cost analysis showed that the fiber cost of PE fiber-reinforced composites was 2910CNY/m³. Replacing 20% and 40% of M-PE fibers with PP fibers reduced the fiber cost by 267.8CNY/m³ (9.2%) and 899.3CNY/m³ (30.9%), respectively. Similarly, replacing 20% and 40% of M-PE fibers with BF fibers decreased the fiber cost by 52.5CNY/m³ (1.8%) and 468.7CNY/m³ (16.1%), respectively. These results demonstrate that partially replacing high-cost PE fibers with PP or BF fibers provides considerable economic benefits. PP fibers offer the greatest cost-reduction potential, whereas BF fibers provide a better balance between cost efficiency and mechanical performance enhancement.

4. Hybrid Fiber Reinforcement Mechanism

M-PE, PP, and BF fibers exhibit significant differences in tensile strength, elastic modulus, and surface characteristics, and their hybridization can induce complementary synergistic effects. Due to their relatively low modulus, PP fibers can bear partial stress during the initial loading stage, whereas BF fibers with high modulus and excellent hydrophilicity can improve the interfacial transition zone and refine the pore structure [35]. Meanwhile, M-PE fibers provide superior ductility and toughness [36]. As illustrated in Figure 13, fibers with different moduli enable hierarchical stress transfer under external loading, which not only enhances the overall stiffness of the composite but also delays premature PE fiber fracture, thereby achieving an effective hybrid reinforcement effect.

In fiber-reinforced cementitious composites, PE fibers serve as the primary reinforcing phase responsible for bridging tensile stresses across cracks and maintaining high ductility. In contrast, PP and BF fibers mainly act as auxiliary reinforcements by enhancing cracking strength, dissipating energy, and redistributing stress. Hybrid fibers provide multi-point restraint during the early loading stage, significantly improving first-cracking strength and effectively suppressing the initiation and propagation of microcracks [11]. Before crack formation, PP or BF fibers preferentially undertake applied stresses and dissipate energy through frictional pull-out or fiber fracture, thereby alleviating stress concentration on M-PE fibers and preventing their premature rupture or pull-out during crack initiation. This mechanism allows M-PE fibers to maintain a more stable bridging effect during critical loading stages, thereby improving the overall toughness and deformation capacity of the composite. After crack penetration through the matrix, stress transfer across cracks is primarily controlled by the bridging effect of M-PE fibers. The high-M-PE-content groups exhibit more effective fiber-bridging networks due to the greater proportion of PE fibers, resulting in superior tensile and flexural properties. Conversely, reducing the M-PE fiber content decreases the number of effective bridging fibers, leading to reduced overall toughness. Owing to the higher elastic modulus and tensile strength of BF fibers compared with PP fibers [12], the M-PE/BF hybrid fiber system demonstrates superior performance to the M-PE/PP hybrid fiber system.

5. Conclusions

This study systematically investigated the effects of M-PE/PP and M-PE/BF hybrid fiber systems on the mechanical properties, cracking behavior, and microstructure of fiber-reinforced cementitious composites, elucidating the synergistic reinforcement mechanisms and cost-reduction potential of hybrid fibers. The main conclusions are summarized as follows:

(1) M-PE fibers are essential for maintaining strain-hardening behavior and high ductility in ECC, with their content governing the performance evolution of hybrid fiber systems. When the M-PE fiber content was ≥60%, the specimens exhibited distinct strain-hardening behavior; below this threshold, strain hardening was lost. The pure PP and BF fiber groups showed reductions of 51.5% and 13.1% in tensile strength and 86.6% and 85.8% in tensile strain, respectively. Partial replacement of M-PE fibers improved composite performance, with 20% PP and BF substitution increasing compressive strength by 13.4% and 15.7%, respectively. Furthermore, the M-PE/PP and M-PE/BF systems enhanced tensile strength by 7.6% and 23.7%, respectively, indicating a stronger synergistic reinforcement effect of BF fibers.

(2) Hybrid fibers significantly improved the cracking resistance of cementitious composites through crack bridging and stress redistribution. BF fibers exhibited superior crack-control efficiency due to their higher elastic modulus and stronger interfacial bonding capability. Under tensile loading, the cracking strengths of the PP and BF systems increased by 25.5%-60.0% and 84.4%-168.8%, respectively. Under flexural loading, the BF-0, BF-1, and BF-2 groups exhibited increases of 37.7%, 39.2%, and 34.7% in cracking strength, respectively. Plate cracking tests further confirmed that hybrid fibers effectively reduced crack number, length, and width, achieving a maximum crack reduction coefficient of 67.98%.

(3) Fiber type and hybridization ratio were associated with changes in interfacial morphology and pore structure but did not produce new detectable crystalline phases. SEM provided qualitative evidence of different local interface and failure morphologies for M-PE, PP, and BF fibers; these images were not used to quantify hydration-product content or bond strength. MIP quantitatively showed that the most probable pore diameters were concentrated within 20-50 nm and that total intrusion volume followed the order PP-0>BF-0>PP-1>BF-1>M-PE. At equivalent total fiber contents, the BF-based specimens exhibited lower total intrusion volumes than the PP-based specimens. XRD identified crystalline SiO₂ as residual/quartz-containing silica rather than a hydration product and showed no additional crystalline phases due to fiber hybridization; poorly crystalline C-S-H was not quantitatively assessed by XRD.

(4) The hybrid fiber strategy effectively reduced material costs while maintaining satisfactory mechanical performance. The fiber cost of PE fiber-reinforced cementitious composites was approximately 2910 CNY/m³. Replacing M-PE fibers with 20% and 40% PP fibers reduced the fiber cost by 9.2% and 30.9%, respectively, whereas 20% and 40% BF replacement reduced the cost by 1.8% and 16.1%, respectively. PP fibers provided the greatest cost-saving potential, while BF fibers achieved a better balance between performance preservation and cost reduction, providing an effective strategy for developing low-cost, high-performance ECC materials.