Study on the effect of rubber content on the frost resistance of steel fiber reinforced rubber concrete

Change in mass loss

Figure 7 illustrates the influence of varying rubber content levels on the mass reduction of SFRRC during F-T cycles. It is shown that the rate of mass loss increases slowly during the initial stages of the experiment and accelerates significantly after 100 cycles. The respective mass losses for SFR0C, SFR5C, SFR10C, and SFR15C are recorded as 3.36%, 3.24%, 2.92%, and 4.01% after 300 cycles. It is clear that there is a gradual decrease in the rate of mass loss as the rubber content increases from 0 to 10%. Notably, SFR10C exhibits the lowest mass loss and demonstrates superior resistance against F-T peeling. However, the mass loss rate significantly increases when the rubber content reaches 15%, surpassing even that of SFR0C without rubber, which exhibits the lowest F-T peeling resistance.

As an organic elastomer, rubber has a dual effect on the frost resistance of SFRRC. On the positive side, due to its high toughness, rubber possesses significant tensile and compression deformation abilities. This characteristic helps reduce hydrostatic pressure and osmotic pressure during F-T processes, effectively mitigating concrete crack expansion and reducing connectivity. Simultaneously, the inclusion of rubber particles acts as an air-entraining agent by increasing the quantity of enclosed small pores in the concrete²⁰. Consequently, this reduces frost heave stress within the concrete and improves its resistance against freezing conditions. On the negative side, due to its organic nature, the chemical inertness of the rubber surface leads to insufficient adhesion between rubber and cement²². This can result in a weak interface layer that is prone to cracking and failure under frost heave stress. Therefore, adding an appropriate amount of rubber into concrete has been found effective in enhancing its frost resistance. However, excessive amounts of rubber tend to increase the occurrence of weak interface layers, thus reducing its frost resistance.

Change in RDME

Figure 8 illustrates the influence of varying rubber content levels on the RDME loss of SFRRC during F-T cycles. It can be observed that during the initial phase of the test, there is a slow decrease in RDME, followed by a significant decrease in the subsequent phase.After undergoing 300 cycles, the RDMEs of SFR0C, SFR5C, SFR10C and SFR15C are reduced to 68.18%, 71.69%, 75.81% and 62.57% respectively. As the rubber content is increased from 0 to 10%, a gradual decrease in the RDME is observed. Notably, SFR10C exhibits minimal loss in RDME and demonstrates superior frost resistance performance compared to other types of SFRRC. However, increasing the rubber content to 15% results in a significant increase in RDME loss. This can be attributed to the higher rubber content, which leads to an increase in internal defects within SFR15C and consequently decreases its frost resistance performance. During the first 100 F-T cycles, the bond between steel fibers and the concrete matrix remains strong, effectively maintaining crack resistance and causing a slow decline in RDME. However, with repeated application of temperature stress and frost heave stress, the bond strength at the interface between the steel fibers and the matrix weakens, leading to a reduction in crack resistance¹⁷. Subsequently, as more cracks emerge and their width increases, SFRRC experiences an accelerated decrease in RDME.

Relative dynamic modulus of elastically of SFRRC.

The decrease in RDME primarily results from an increase in cracks and a decrease in compactness within SFRRC. Cracks serve as the main channels for water and ice crystals to expand and damage the concrete. During crack propagation, randomly distributed rubber particles exhibit a blocking effect, while staggered steel fibers help alleviate stress at crack tips, thereby inhibiting further growth of the crack. Additionally, elastic rubber and high-strength steel fibers can absorb and disperse part of the stress during F-T cycles³⁵. The combined effect significantly enhances the frost resistance of SFRRC.

Change in compressive strength

Figure 9 illustrates the influence of varying rubber content levels on the compressive strength of SFRRC during F-T cycles. After 300 cycles, the compressive strengths of SFR0C, SFR5C, SFR10C, and SFR15C decrease to 73.22%, 78.36%, 81.36%, and 67.98% respectively. As the proportion of rubber in SFRRC increases from 0 to 10%, a gradual decrease in compressive strength is observed. Among different types of specimens, SFR10C exhibits the least decrease in relative compressive strength. The incorporation of rubber particles at this level significantly enhances the internal pore structure, leading to a predominantly positive influence on SFR10C and mitigating the reduction in compressive strength.

Relative compressive strength of SFRRC.

However, incorporating 15% rubber into SFR15C results in a significant decrease in the relative compressive strength compared to SFR0C without the addition of rubber. This phenomenon can be attributed to two main reasons. Firstly, rubber particles, as elastomers, have low strength. The inclusion of rubber particles at this higher content increases weak defects within the SFR15C, thereby reducing its effective bearing area and resulting in a reduction of its mechanical properties¹². Secondly, rubber, due to its high-molecular organic nature, exhibits a relatively weak bonding interface between its particles and cement mortar. When subjected to external forces, concrete tends to fail along this interfacial zone⁸. Therefore, a higher rubber content adversely affects the concrete, resulting in a faster decline in compressive strength.

Change in damage layer thickness

Figure 10 illustrates the process of damage evolution in SFRRC. During repeated F-T cycles, the outer layer of SFRRC is initially affected by freezing pressure, which leads to the formation of microcracks once the pressure exceeds the tensile strength of the concrete. As the number of F-T cycles increases, these microcracks gradually expand, causing a progressive deterioration in the internal microstructure of the concrete and ultimately resulting in the formation of a damaged layer. The progression and accumulation of this damage advance the boundary of the damage layer into the interior of the concrete, which is reflected by an increase in H_f. This degradation can be observed macroscopically as a reduction in strength and V_f.

The process of damage evolution in SFRRC.

The variations of ultrasonic velocity and H_f in SFRRC with different rubber contents subjected to F-T cycles are illustrated in Fig. 11. As the experiment progresses, there is a gradual increase in the H_f, while the V_f gradually decreases. After undergoing 300 F-T cycles, measurements reveal that the respective H_f values for SFR0C, SFR5C, SFR10C and SFR15C are recorded as 17.39 mm, 16.97 mm, 16.39 mm and 19.07 mm. It can be observed that SFR10C demonstrates superior frost resistance performance due to its smallest H_f and minimal reduction in the V_f. However, once the rubber content reaches 15%, there is a noticeable increase in H_f, which exceeds that observed in SFR0C without rubber. The change in the H_f resulting from F-T cycles corresponds to the experimental results illustrated in Figs. 7, 8, 9, as evident from the data. This indicates that a higher rubber content negatively impact the frost resistance capability of SFRRC. Due to the presence of a large number of microcracks and pores in the damaged layer, as these defects accumulate and expand, the originally solid concrete in the affected areas transforms into gas phase spaces³². This results in a decrease in V_f, as ultrasound propagates more slowly in air compared to concrete. Similar findings have been reported in reference³⁶, which observed that under sulfate attack and F-T conditions, the H_f of recycled concrete increases while V_f gradually decreases with an increasing number of F-T cycles.

Change in damage layer thickness of SFRRC: (a) The V_f; (b) The H_f.

The experiment findings indicate that the fluctuation pattern of H_f aligns with the accelerated decrease stage of RDME, as shown in Fig. 8. This suggests a significant increase in the damage level of SFRRC after 100 F-T cycles. Both H_f and RDME are measured using ultrasonic techniques, and Fig. 12 illustrates the correlation between RDME and H_f in SFRRC subjected to F-T cycles. The graph clearly demonstrates a gradual increase in H_f as RDME decreases, indicating a strong association between these two factors. During F-T cycles, progressive deterioration occurs in SFRRC from its surface towards its interior. Therefore, utilizing H_f as an evaluation parameter can effectively assess the extent of degradation in SFRRC. Previous studies^31,32 have also indicated that recycled aggregate concrete or coal gangue concrete exposed to F-T conditions exhibit increased severity of damage and reduced compactness, which corresponds to an increase in H_f.

Relationship between H_f and RDME.

Calculation of the compressive strength in damage layer

As the experiment progresses, an increase in the H_f of SFRRC leads to the emergence of defects in its meso-structure, resulting in a gradual decrease in the strength of the damage layer. This directly affects the bearing performance of the SFRRC structure. The calculation results for the f_d of SFRRC are presented in Table 4. It is evident that the f_d gradually decreases with an increase in F-T cycles. With an increase in F-T cycles, there is a gradual reduction in the f_d. After undergoing 300 cycles, the damage layer of SFR0C (without rubber content) exhibits a reduction in compressive strength to approximately 47.33%, while for SFR5C, SFR10C, and SFR15C, the decrease amounts to around 40.09%, 36.03%, and 53.56% respectively. In comparison to the test results depicted in Fig. 9, the reduction in f_d is more pronounced. SFR15C exhibits the highest rate of strength reduction in its damage layer, surpassing even that of SFR0C without rubber. The observation suggests that SFR15C undergoes significant deterioration, indicating that the damage layer is the weak area of SFRRC when exposed to F-T conditions. Therefore, effectively addressing deterioration in the damage layer is essential for improving the mechanical properties and long-term durability of SFRRC.

Microscopic morphology of SFRRC

SEM was conducted on the SFRRC samples both before and after undergoing freezing–thawing cycles to observe their microstructure, as depicted in Fig. 13. It is evident that the hydration products in SFRRC have cemented together to form a continuous phase prior to F-T damage. The overall structure appears uniform and compact without any microcracks, as depicted in Fig. 13a. During the F-T test process, the walls of SFRRC holes are subjected to both hydrostatic and osmotic pressures. In situations where the tensile stress exceeds the concrete’s tensile strength, microcracks form within the SFRRC. Figure 13b and c respectively illustrate the presence of cracks in SFR0C and SFR10C after 200 cycles. The crack width of SFR10C is found to be smaller compared to that of SFR0C. As an elastic material, rubber particles exhibit excellent resistance to compressive deformation, thereby mitigating tensile stress caused by F-T cycles. Consequently, the progression of cracks in concrete is effectively hindered.

Microcosmic morphology of SFRRC specimens during F-T test: (a) Undamaged specimen; (b) Microcracks were observed in SFR0C after 200 cycles; (c) Microcracks were observed in SFR10C after 200 cycles; (d) ITZ between rubber and cement matrix; (e) ITZ between rubber and cement matrix; (f) Microcracks expand and the structure is obviously loose in SFR15C after 300 cycles.

The frost resistance properties of concrete are further improved by the air-entrainment effect caused by rubber particles, which is evident in the reduced mass loss, RDME loss, and compressive strength loss exhibited by SFRRC. Additionally, a decrease in H_f is also achieved. However, due to its hydrophobic nature, rubber does not participate in the hydration reaction of cement. The primary bonding mechanism between rubber particles and the cement matrix is physical adhesion rather than strong chemical bonding³⁷. Consequently, the interfacial transition zone (ITZ) between them becomes a weak area, as depicted in Fig. 13d. The large amount of rubber increases the number of micropores and the ITZs within SFR15C, resulting in a noticeable decline in its bonding strength at the interface and frost resistance. The limited adhesion between rubber and cementitious materials in SFRRC makes it prone to crack formation within the ITZs, with a tendency for the cracks to widen further (Fig. 13e). After undergoing 300 cycles, the SFR15C experiences an expansion and increase in internal cracks. Moreover, the cracks within the pores interconnect, resulting in a loosening of the structure of hydration products as depicted in Fig. 13f. Consistent with the macroscopic performance test results mentioned above, the SFRRC with a 15% rubber content exhibits significant deterioration. The macroscopic properties of SFRRC are inferred to be influenced by its microscopic structure.

Analysis of the pore structure

Table 5 presents the parameters of SFRRC’s pore structure obtained through mercury intrusion testing, while Fig. 14 illustrates the distribution of pore sizes. The incorporation of rubber particles is observed to increase the number of enclosed and fine bubbles in SFRRC, thereby enhancing its porosity and effectively functioning as an air-entraining agent. However, the macro performance of SFRRC is demonstrated by a reduction in compressive strength. The SFR0C without rubber exhibits minimal porosity, but it has a relatively large average pore size. Based on the data presented in Fig. 14, it is evident that SFR0C exhibits a reduced number of harmless (d < 20 nm) and less harmful (20 nm ≤ d < 50 nm) pores compared to SFR5C and SFR10C. Conversely, there is a higher proportion of harmful (50 nm ≤ d < 200 nm) and more harmful (d ≥ 200 nm) pores in SFR0C. This is reflected in the decreased frost resistance of SFR0C in terms of its macro performance.

The influence of varying rubber content on the distribution of pore size.

Compared to SFR5C, the number of harmless pores in SFR10C increases by 35.23%, while the numbers of harmful and more harmful pores decrease by 27.20% and 30.71% respectively. Due to the hydrophobicity of rubber particles, an air–water film forms on the contact surface with mortar during the mixing process. The large pores on the surface of rubber particles are uniformly dispersed into stable and closed micro-pores under water adsorption³⁸. Meanwhile, the air–water film attached to the rubber particles also participates in the hydration reaction of cementitious materials, further refining the pore structure³⁹. Therefore, the data indicates that SFR10C demonstrates a decrease in total porosity, total pore volume, and area compared to SFR5C. The literature²⁴ also notes that as the rubber powder content increases from 5 to 10%, the total porosity of SFRRC decreases to a certain extent, and the rate of increase in total porosity diminishes with an increasing number of F-T cycles. Furthermore, the increase in the ratio of harmless and less harmful pores also contributes to improving the frost resistance capability in SFR10C. The most favorable pore structure and the optimal frost resistance performance of SFRRC are achieved with a rubber content of 10%.

However, when the rubber content is increased to 15%, the number of harmless pores with d < 20 nm in SFR15C decreases by 53.20% compared to SFR10C, while the proportions of pores with sizes ranging from 20 nm ≤ d < 50 nm, 50 nm ≤ d < 200 nm, and d ≥ 200 nm increase notably. Therefore, a significant increase in pore structure parameters of SFR15C is observed. Specifically, compared to SFR10C, there is a 37.64% increase in total porosity, a 35.70% increase in total pore volume, and a 53.36% increase in total pore area. Moreover, both the average and most probable pore diameters in SFR15C demonstrate a significant increase, accompanied by an observed rise in the proportion of harmful and more harmful pores. In terms of macroscopic performance, this results in a significant decrease in both RDME and compressive strength for SFR15C, as well as an increase in H_f. It is evident that the addition of rubber in an appropriate volume ratio improves the pore structure of SFRRC. However, excessive rubber content adversely affects both the pore structure and frost resistance of SFRRC.