Enhancing self-healing in ultra-high-performance fiber-reinforced concrete: The role of superplasticizer anchor groups

2.1. Materials

The Portland CEM I 52.5 cement used in this study was sourced from Shahre-Kord Cement Factory, Iran. Cement supplementary materials included GGBFS collected from the Esfahan Steel Company, Iran, and silica fume (SF) powder obtained from OCI Company Ltd., South Korea (Table 1). Two different sizes of quartz sand (0.2–0 mm and 0.6–0 mm) were used in the mix. The superplasticizers, PCE (dark brown, neutral pH = 7) and PCE-P (light brown, neutral pH = 7) in liquid form were acquired from Peivand Shimi, Iran (Figure S3). Additionally, two types of synthetic fibers, polypropylene (PP) and polyvinyl alcohol (PVA), were used at a 1% volume ratio of each (Table 2). These fibers were sourced from Faratav Co., Iran.

2.2. Preparation and mixing process of UHPFRC

To prepare UHPFRC, all dry components (powder and sand) were first mixed for 5 mins in a Hobart-type mixer. Following this, half of the water was gradually added to the dry mixture. The hydrophilic nature of PVA fibers aids in the concrete self-healing process by creating nucleation sites, which enhance repair efficiency in samples with PVA fibers compared to non-polar fibers. These nucleation sites help facilitate the formation of healing materials within cracks. Additionally, longer fibers are more likely to align over crack surfaces, creating a larger bonding area between fibers and the cement matrix, thus improving bridging capacity and helping to prevent crack propagation. This ultimately enhances post-crack flexural performance. To leverage these benefits, this study used 6 mm PVA fibers and longer 12 mm PP fibers, each at a 1% volumetric ratio. PP fibers are also used in reinforced concrete due to their unique properties. These fibers resist water and chemicals and do not easily absorb water, which means they have minimal impact on the water content in the concrete mix. Typically, PP fibers improve the mechanical properties of concrete, such as toughness and resistance to shrinkage and cracking. The longer length of PP fibers increases the bonding surface with the cement matrix, enhancing crack-bridging capacity. These characteristics lead to improved flexural strength and durability, and, in combination with hydrophilic PVA fibers, create a synergistic effect that contributes to more effective crack healing. The remaining water, combined with the superplasticizer, was then added to the mixer and blended at a low speed for an additional 5 mins. Next, fibers were introduced gradually, with mixing continuing at a high speed for 5 mins to prevent fiber clumping and ensure optimal concrete performance. In accordance with the requirements of standard C1856/C1856M 17, the concrete molds were filled in a single layer, and the specimens were compacted by striking the mold sides with a specialized hammer 30 times. Immediately after finishing, within 1 min, the molds were covered to prevent drying. Specimens were coded using a three-part system where ‘U’ denotes Ultra-high-performance concrete, ‘S’ or ‘N’ indicates the presence or absence of GGBFS respectively, and ‘S’ or ‘P’ represents the superplasticizer type (standard PCE or phosphate-modified PCE-P), resulting in four designations: USP (GGBFS + PCE-P), USS (GGBFS + PCE), UNP (No GGBFS + PCE-P), and UNS (No GGBFS + PCE). Table 3 presents the composition of concrete components in the tested samples.

2.3. Preloading and cracking

Controlled crack width tests were conducted three days after casting the specimen. To monitor the crack width, a linear variable differential transformer (LVDT) was mounted on the top front of the cube specimen (as shown in Figure S4). The LVDT, with a range of 500 μm and an accuracy of 1 μm, recorded the crack width as it increased at a rate of 0.3 $\mu {\text{m s}}^{-1}$. This controlled crack growth was achieved using a servo-hydraulic testing machine, which applied a specific loading rate to ensure the crack widened at the predetermined rate. Once the crack width reached 200 μm, the specimen was unloaded, leading to partial crack closure. The crack width was then measured by a digital microscope, yielding values of 150 ± 15 μm.

2.4. Method and characteristics

In this study, the impact of PCE and PCE-P on the self-healing properties of UHPFRC was compared. Microcracks were induced in concrete specimens, which were then immersed in water at room temperature to evaluate their self-healing capacity. The evaluation involved examining the specimens using digital microscopy, measuring compressive strength recovery, and using SEM combined with EDS analysis. The findings are detailed below.

The self-healing performance of cracks was evaluated at 3 days (the age of crack formation) and 90 days using digital microscopy. In this analysis, the crack area before and after the healing process was isolated from the rest of the image and analyzed using the histogram thresholding method in ImageJ software. Crack images were converted into binary format, with black pixels (value of 1) representing the crack area and white pixels (value of 0) representing the rest of the specimen. The overall healing ratio for each sample was calculated using the following Eq. (1):

Where A_b and A_h represent the crack area (sum of black pixels) before and after healing, respectively. To ensure high-quality images, a panoramic technique was employed to combine images from different sections of the sample, enabling detailed comparison at 3 and 90 days.

The recovery of compressive strength for the specimens is evaluated using 5×5×5 cm³ samples according to ASTM C109/C109M-16 test methods [21]. The compressive strength recovery percentage, denoted as “${\text{C}}_{\text{P}}$“ is calculated using Eq. (2) for the specimens:

Where “${\text{C}}_{\text{h,90}}$“ and “${\text{C}}_{\text{u,28}}$“ represent the compressive strength values of the healed specimen at 90 days and the uncracked specimen at 28 days, respectively [22,23].

3.1. Evaluation by a digital microscopy

A comparison of USP (GGBFS + PCE-P), USS (GGBFS + PCE), UNP (no GGBFS + PCE-P), and UNS (no GGBFS + PCE) specimens at these intervals has been presented in Figure 1. The average self-healing percentages for the UNS, USS, UNP and USP specimens, were found to be 50.39%, 93.68%, 64.43%, and 98.35%, respectively. These findings suggest a significant improvement in the self-healing capabilities of specimens containing GGBFS and PCE superplasticizer when compared to those without GGBFS but with the same superplasticizer, with the self-healing percentage increasing from 50.39% to 93.68%. This notable enhancement underscores the positive role of GGBFS in promoting the self-healing properties of UHPC. GGBFS, as a pozzolanic material, reacts with calcium hydroxide produced during cement hydration to form additional calcium silicate hydrate (C-S-H), which helps in healing microcracks by filling voids and improving the concrete’s durability [24]. Moreover, specimens incorporating GGBFS and PCE-P (phosphate-modified PCE superplasticizer) demonstrated an even greater increase in self-healing percentage, from 64.43% to 98.35%, compared to specimens without GGBFS but with the same superplasticizer. This result indicates that PCE-P is more effective than PCE in enhancing the self-healing properties of UHPFRC. The modification of PCE with phosphate groups likely improves the dispersion of the superplasticizer in the mix, leading to better workability and the formation of a denser microstructure. This denser matrix facilitates the self-healing process by improving the sealing of microcracks.

Close

Figure 1.

Digital microscopy images of crack surfaces before (3 days) and after (90 days) of healing for specimens: (a) UNP, (b) UNS, (c) USP, and (d) USS.

Export to PPT

The specimens containing GGBFS exhibited the highest self-healing performance, with an average healing rate of 96%. This can be attributed to the low hydration level of GGBFS particles, which allows unreacted particles to activate upon cracking, thereby enhancing the crack healing process [25]. When compared to other specimens with GGBFS (USS and USP), the USP specimens, which incorporated PCE-P as a superplasticizer, showed superior performance in terms of self-healing. A similar trend was observed for specimens without GGBFS, where the self-healing performance of the UNP specimen was approximately 14% better than that of the UNS specimen. However, it is worth noting that the effect of the superplasticizer on self-healing was substantially less significant than the effect of GGBFS.

PCEs are known to be highly effective dispersants in cementitious systems containing silica fume [24]. The mechanism by which PCE interacts with silica fume differs from its interaction with cement particles due to the negative charge on silica fume particles. PCEs are adsorbed onto the surface of silica fume particles through hydrogen bonding between the terminal hydroxyl groups of polyethylene glycol (PEG) chains and the silanol groups on the silica fume surface [26,27]. When cracks form in the material and the specimens are immersed in water, the unreacted superplasticizer molecules undergo hydrolysis. At the same time, silica fume particles at the crack interface come into contact with water. As hydration continues, calcium hydroxide (Ca(OH)₂) is formed, leading to the generation of Ca²⁺ and OH^- ions. This is due to two factors: (a) the formation of an electric double layer and a negative zeta potential on the silica fume surface, and (b) the increased anionic charge density of PCE caused by the separation of hydrogen from the carboxylate group in the polymer backbone. The complexation of Ca²⁺ ions with counterions results in the rapid detachment of PCE from the silica fume surface [28].

The detached PCE groups exhibit a strong affinity for absorbing Ca²⁺ cations, promoting the formation of new materials within the crack and facilitating crack closure. The carboxylate (COOH^-) group in standard PCE acts as the anchor group, while in PCE-P, the anchor group is phosphate ($-{\text{PO}}_4^{3-}$). Based on the experimental findings, PCE-P was found to enhance the self-healing of UHPFRC more effectively than conventional PCE. This improvement can be explained through Pearson’s theory of hard and soft acids and bases. According to this theory, soft acids interact more with soft bases. On the other hand, hard acids tend to form stronger bonds with hard bases, resulting in more stable compounds. The bond between hard acids and hard bases is primarily based on electrostatic forces. However, the bond between soft acids and soft bases is more covalent due to the high polarizability of soft bases. In the case of the coordination bond between a calcium ion (Ca²⁺), a hard acid, and the carboxyl group (-COOH), a soft base, the bond is weak. This weak bond leads to the instability of the CaCOOH compound [29]. In contrast, the phosphate group ($-{\text{PO}}_4^{3-}$) is a hard base that forms a much stronger bond with calcium, resulting in the formation of stable calcium phosphate compounds.

3.2 Recovery of compressive strength and microstructural (SEM/EDS) analysis

Figure 2 illustrates the average compressive strength values for uncracked specimens at 28 days and for cracked specimens at 28 and 90 days across all specimens. The average recovery percentages of compressive strength are 71.94, 83.08, 70.95, and 85.16 for UNS, USS, UNP and USP specimens, respectively.

Close

Figure 2.

Average compressive strengths at 28 and 90 days for healed specimens, and at 28 days for uncracked specimens, for (a) UNP, (b) UNS, (c) USP, and (d) USS.

Export to PPT

The highest recovery of compressive strength is observed in specimens containing GGBFS and using PCE-P at 85.16%. However, specimens with GGBFS and PCE also exhibit a high recovery rate of 83.08%, indicating that the pozzolanic effect of GGBFS in restoring compressive strength in healed specimens is greater than the effect of the superplasticizer. In all specimens, the compressive strength recovery rate at 90 days surpasses that at 28 days, which can be attributed to the delayed action of both GGBFS and PCEs.

The recovery of compressive strength in the specimens indicates that those with GGBFS and PCE-P demonstrated superior performance. The significance of GGBFS was highlighted, showing that when cement is replaced by GGBFS in UHPC mixtures, hydration is initially delayed, which leads to reduced compressive strength. However, the long-term compressive strength remains consistent [30]. In this study, replacing 15% of the cement’s weight with GGBFS resulted in approximately 7% lower 28-day strength in uncracked specimens compared to those without GGBFS. The pozzolanic activity of GGBFS improves the microstructure and properties of UHPFRC by utilizing calcium hydroxide and generating additional C-S-H gel [31]. The generation of additional C-S-H gel through pozzolanic activity not only increases long-term compressive strength but also facilitates better crack healing. The enhanced recovery of compressive strength in specimens with PCE-P can be attributed to a stronger bond between the phosphate anchor group ($-{\text{PO}}_4^{3-}$) and calcium ions, which is induced by the hydration product (Ca(OH)₂) within the crack. The high capacity of phosphate groups in superplasticizers to complex with calcium enhances their performance in concrete [32]

A cubic section of the specimen, including the fixed crack, measuring 1×1×1 cm³, is sliced using a diamond saw blade (Figure 3). SEM images and EDS results for the samples have been presented in Figure 4(a-d). Both UNP and USP samples contain phosphorus (P), while USS and UNS samples do not, which is related to their anchor group types. Furthermore, the Ca/Si ratio in USS and UNP samples is greater than 1. The K-α line of phosphorus (P) and the M-α line of gold (Au) have similar energies (2.013 keV and 2.123 keV, respectively). The SEM images showing the morphology of the healing materials within the crack, and the EDS results indicate that the Ca/Si ratio in the GGBFS samples is lower. The compressive strength of C-S-H pastes increases as the Ca/Si ratio decreases [33]. An increase in the pH of the mixture leads to enhanced desorption of superplasticizers from the surface of silica fume particles [34]. The alkalinity of GGBFS-containing samples increases desorption, resulting in higher phosphate ion (PO₄³⁻) concentration and greater phosphorus (P) content in samples with PCE-P mixed with GGBFS. When the Ca/P ratio reaches 1.67, phosphate ions combined with calcium ions trigger hydroxyapatite formation [35]. Studies on hydroxyapatite describe its SEM appearance as needle-like, and upon closer inspection, the SEM images reveal the hydroxyapatite structure (Figure 5).

Close

Figure 3.

Specimen preparation for SEM.

Export to PPT

Close

Figure 4.

SEM images of (a) UNP, (c) UNS, (e) USS, and (g) USP specimens with (b,d,f,h) corresponding EDS analyses from marked zones.

Export to PPT

Close

Figure 5.

The hydroxyapatite structure in previous electron microscopy studies.

Export to PPT

3.3. Comparing the self-healing mechanism with other methods

Bacterial concrete utilizes carbonate-producing bacteria (such as Bacillus spp.) to generate calcium carbonate (CaCO₃) and seal cracks [36,37]. This method generally enhances self-healing in thin cracks (less than 0.8 mm), but its performance depends on the metabolic activity of bacteria and environmental conditions (such as humidity and pH) [38,39]. In contrast, the method examined in this study, which is based on the use of GGBFS and modified PCE-P, relies on a chemical mechanism and is independent of biological environmental conditions [40]. The self-healing rate observed in this study (98.35%) is higher than most reported values for bacterial concrete, which typically range between 50% and 80%.

In some studies, healing agents such as silicates, epoxies, or cementitious liquids are encapsulated within microcapsules or polymeric fibers and released when the concrete cracks [41]. The advantage of this method lies in its precise control over the amount of healing material and its applicability in dry conditions [42]. However, one of the challenges of this approach is the stability of the microcapsules and the potential for premature degradation during the concrete mixing process [43]. In comparison, the method using GGBFS and PCE-P in this study offers a more natural and durable mechanism, as it leverages internal chemical reactions within the concrete. This method has also demonstrated the ability to restore compressive strength up to 85.16%, whereas encapsulation-based approaches typically have a more limited impact on compressive strength recovery due to their dependence on the release efficiency of the healing agent.

In this study, compressive strength recovery in specimens containing GGBFS and PCE-P exceeded 85%. In contrast, bacterial concrete typically exhibits compressive strength recovery between 50% and 75%, while microcapsule-based methods, depending on the type of encapsulated material, achieve recovery rates ranging from 60% to 90%. This indicates that the method utilizing GGBFS and PCE-P not only exhibits high efficiency in crack healing but also contributes to the structural reinforcement of concrete [44,45]. While bacterial concrete and microcapsule-based methods are effective in enhancing concrete self-healing, the approach in this study, based on GGBFS and PCE-P, demonstrates superior performance in crack repair (up to 98.35%) and compressive strength recovery (up to 85.16%). Furthermore, this method is more stable and does not require specific environmental conditions (such as high humidity for bacterial activation), which is a key advantage in practical applications.

3.4. The importance of self-healing concrete

The findings of this study have significant implications for the development and practical application of self-healing concrete in real-world infrastructure, particularly in structures subjected to harsh environmental conditions such as bridges, tunnels, and maritime structures [46]. Cracking in concrete, if not properly addressed, can lead to severe durability issues, increased maintenance costs, and structural failures due to the ingress of water, chlorides, and other aggressive agents [47]. This study contributes to the growing body of research on self-healing UHPFRC and its ability to enhance durability and extend the lifespan of critical infrastructure.

One of the major challenges in infrastructure engineering is the high cost associated with repairing cracks in concrete structures. The incorporation of PCEs with tailored anchor groups and pozzolanic additives like GGBFS can significantly improve the autogenous self-healing capacity of UHPFRC. Studies have shown that pozzolanic materials enhance calcium carbonate precipitation and secondary hydration reactions, effectively sealing microcracks over time [28,48]. This self-healing mechanism can reduce maintenance cycles and prolong service life, making it highly beneficial for large-scale infrastructure.

Bridges, tunnels, and maritime structures are constantly exposed to moisture, chloride attack, freeze-thaw cycles, and sulfate exposure, which accelerate deterioration. Research indicates that phosphate-functionalized PCEs exhibit superior resistance to sulfate attack, which is particularly useful in marine environments [49,50]. Additionally, the steric hindrance effect of phosphate groups enhances adsorption on cement particles, leading to better dispersion, lower permeability, and reduced risk of chloride-induced corrosion [51]. These properties are crucial for structures like offshore platforms, ports, and underwater tunnels, where traditional concrete often experiences rapid degradation.

The use of GGBFS as a supplementary cementitious material (SCM) not only enhances self-healing but also aligns with sustainable construction practices by reducing the carbon footprint of cement production. According to research by Raghuvanshi and Singh [52], the incorporation of GGBFS in concrete mixtures can significantly lower CO₂ emissions while improving mechanical properties and long-term durability. Given the growing emphasis on green construction and climate-resilient infrastructure, self-healing UHPFRC offers a viable solution for future projects.

Advancements in digital monitoring and self-sensing concrete have made it possible to integrate self-healing mechanisms into smart infrastructure systems. The ability of self-healing concrete to autonomously repair microcracks can be combined with embedded sensors to provide real-time data on crack propagation and healing efficiency [53,54]. This integration can be particularly useful in high-traffic bridges, underground metro systems, and high-speed rail networks, where early detection and self-repair mechanisms can prevent catastrophic failures.

The insights gained from this study contribute to the next generation of self-healing concrete materials that can reduce maintenance costs, enhance durability, and promote sustainable construction. By understanding the role of PCEs with optimized anchor groups and pozzolanic additives, engineers can develop more resilient and long-lasting concrete structures suited for aggressive environments such as coastal regions, seismic zones, and high-load transportation infrastructure. Future research should explore the scalability and cost-effectiveness of these materials in large-scale construction projects, as well as the potential to integrate biomimetic self-healing mechanisms for further improvements.

3.5. Future research directions

Despite the promising advancements in self-healing concrete using PCEs with PCE-P and GGBFS, further research is essential to assess its long-term performance, economic feasibility, and large-scale applicability. Future studies should explore several key areas. Firstly, long-term durability under various environmental conditions must be addressed. While short-term experiments have demonstrated the self-healing potential of UHPFRC, its durability under real-world conditions over extended periods remains uncertain. Future research should focus on long-term chloride penetration resistance, particularly given the widespread use of concrete in marine structures and bridges. Studies should assess whether self-healing properties remain effective over decades in chloride-rich environments [55,56]. Additionally, resistance to freeze-thaw cycles is crucial; in cold climates, self-healed concrete should be tested for microcrack expansion due to repeated freeze-thaw cycles, which could compromise structural integrity [57]. Another important aspect is behavior under dynamic loading, as bridges and high-rise buildings experience constant fatigue loading, necessitating investigations into whether self-healing mechanisms can repeatedly activate under such conditions [58]. Secondly, the economic viability and large-scale implementation of self-healing concrete need thorough evaluation. While laboratory studies highlight the potential of PCE-P and GGBFS in enhancing self-healing concrete, their cost-effectiveness in large-scale applications remains unclear. Key aspects for future research include conducting a cost-benefit analysis by comparing the initial cost of self-healing concrete with traditional repair costs over a structure’s lifespan to determine financial feasibility [59]. Additionally, optimizing the mix design to identify the optimal ratio of PCE-P and GGBFS for maximum self-healing while minimizing additional costs is essential. Investigating the scalability for industrial applications is also important, specifically how batch production and field implementation challenges affect the practical use of self-healing concrete in large infrastructure projects such as tunnels, highways, and offshore platforms.

Thirdly, integrating self-healing concrete with smart monitoring systems presents an innovative direction for future research. Embedding self-sensing materials like carbon-based nanomaterials or piezoelectric sensors could provide continuous data on crack closure and healing efficiency [60,61]. Additionally, AI-driven predictive models using machine learning could be developed to predict self-healing performance based on environmental conditions and historical data, enhancing proactive maintenance strategies.

Finally, exploring alternative eco-friendly additives for sustainability is imperative, given the increasing push for low-carbon construction. Future studies should investigate the use of biopolymer-based superplasticizers as biodegradable alternatives to synthetic PCE-P. Moreover, nano-engineered self-healing agents such as nano-silica or graphene oxide could be explored to further accelerate crack healing and enhance mechanical properties [62].

Addressing these research gaps will facilitate the widespread adoption of self-healing concrete in large-scale infrastructure. Long-term durability studies, cost-effectiveness assessments, and the integration of digital monitoring technologies will be crucial in ensuring that PCE-P and GGBFS-based self-healing concrete can be effectively deployed in real-world applications such as bridges, tunnels, marine structures, and high-rise buildings.