Improving the Creep Resistance of Hardened Cement Paste through the Addition of Wollastonite Microfibers: Evaluation Using the Micro-Indentation Technique

Abstract

This paper evaluates the creep behavior of wollastonite-microfiber-reinforced hardened cement paste. A parametric study was performed for cementitious mixtures with partial replacement of cement using wollastonite. The samples were subjected to uniaxial compression and microindentation experiments. The compressive strength significantly improved by up to 30% for the mixture containing 10% wollastonite fiber. Microindentation experiments were performed to evaluate the creep response and time-dependent performance of both the control and the wollastonite-fiber-reinforced samples. Approximately 36% increase in creep modulus was observed with 10% wollastonite fiber content. This indicates a significant improvement in creep behavior, which can be attributed to the micro-reinforcing effect of the fibers. No significant impact was observed for time characteristic of creep with the incorporation of wollastonite fibers. Overall, the study establishes a cost-effective, sustainable, and efficient route to enhance the creep behavior of hardened cement paste for a wide range of infrastructural applications.

1. Introduction

Creep is a significant mechanical behavior of cementitious materials that causes time-dependent deformations throughout a structure’s service life [1,2,3,4,5,6]. This phenomenon occurs when deformation grows excessively under constant stress [7]. Understanding the creep mechanism in cementitious composites is a challenging task due to the environmental influences and complexity between creep and shrinkage [8] as in the absence of an applied load, shrinkage also causes the material to deform [9]. However, the uncertainty between these processes offers the opportunity to investigate the mechanical behavior of cement-based materials at various time and length scales. Research over decades has been dedicated to understanding short- and long-term creep effects, in concrete [10]. Short-term creep is identified by an instantaneous strain and elastic response of the material [11,12]. Alternatively, long-term creep occurs when the calcium-silicate-hydrate (C-S-H) of cement paste begins to slide and rearrange leading to water diffusion that can form microcracks [13]. These small microcracks transform into larger cracks over time which allow water and sodium to seep in and corrode the reinforcement inside [14]. Additionally, creep-induced cracks in prestressed concrete structures have notably larger deflections and weaker bearing capacity [15,16]. These types of failures have been responsible for over billions of dollars needed to maintain reinforced concrete structures [2]. Fiber additives in the form of micro- and nano-fillers in cement paste are being implemented to limit creep and enhance the overall mechanical performance of the binder [17]. Mallick et al. [18] studied the effects of fly ash addition into cement paste on the creep response under several loading conditions. Microindentation tests revealed that capillary water has significant influences on creep compliance [18,19]. Meanwhile, Mangat et al. determined that steel fiber-reinforced cementitious composites have better creep performance than normal cement paste [20,21]. Several other studies introduce silica fume and other reinforcements to improve the creep response of the system [22,23,24,25].

Recently, researchers have been interested in naturally occurring materials, such as wollastonite [26,27,28]. This raw mineral is found in many regions of the world and has shown an increase in production over the last two decades [29,30]. Wollastonite is a non-harmful material used in ceramics, paints, and polymers [31,32]. It is recognized for exhibiting high strength, water, and heat resistance [30,33]. The chemical structure of wollastonite contains natural chains of calcium and silicate-based inorganic compounds which is chemically very similar to calcium-silicate-hydrate (C-S-H) in cement paste. Thus, the interface between these two materials presents better chemical compatibility and bonding between fiber and matrix. Previous studies have shown significant improvement in mechanical performance with the incorporation of wollastonite fibers in cement-based composites [34,35,36]. Dey et al. found that wollastonite fibers prevent shrinkage in hardened cement paste [37,38]. Doner et al. [39], performed microindentation experiments and finite element simulations that showed about a 33% increase in fracture toughness when 10% mass-based cement replacement by wollastonite is considered. Lyngdoh et al. [40] performed multiscale numerical simulation to elucidate the influence of wollastonite microfibers on the dynamic compressive response of mortars and the results suggest about a 45% increase in dynamic energy absorption capacity with 10% mass-based replacement of cement by wollastonite fibers. While these previous studies focus primarily on the mechanical response of wollastonite-fiber-reinforced cementitious composites in terms of compressive/flexural strength and fracture under static and dynamic loading conditions, no study has been conducted thus far to evaluate the creep response. Thus, the significant success of the micro-reinforcing effect of wollastonite microfibers in cementitious composites, as established in the above-mentioned previous studies, provides the necessary impetus to elucidate the influence of wollastonite microfibers on the creep response of cementitious composites which is the ultimate objective of this paper. In particular, this paper evaluates the influence of mass-based cement replacement with wollastonite fibers on the creep response of hardened cement paste using the microindentation technique. Indentation and hardness tests have been performed at the micro and nanoscale to understand the creep behavior of cementitious materials within the interfacial transition zone (ITZ) [41,42,43,44,45,46,47,48,49]. Moreover, the microindentation technique has been leveraged to elucidate the creep response of hardened cement paste [50,51,52,53,54,55,56].

The purpose of this study is to examine the impact of reinforcing hardened cement paste with wollastonite fibers, up to 15% mass-based replacement of cement, on its creep response through microindentation experiments. This investigation is motivated by the successful use of wollastonite fibers in enhancing the static and dynamic mechanical performance of cementitious composites. The goal is to provide a cost-effective and sustainable solution to improve the creep tolerance of cement-based materials. Thus, evaluating the creep behavior of hardened cement paste with varying wollastonite fiber content is a crucial step towards developing a cost-effective and sustainable solution for microinforced cementitious composites.

2. Materials and Experimental Methods

2.1. Materials and Mixture Proportions

For the experiments, 4 µm (diameter) × 12 µm (length) wollastonite fibers manufactured by NYCO Minerals Inc. (Willsboro, NY, USA) were used. A parametric study was performed by replacing cement with a percentage of wollastonite fibers by mass. Type I/II Ordinary Portland Cement (OPC) was substituted with 5%, 10%, and 15% wollastonite fibers by mass. No more than 15% of wollastonite was added due to the improper workability of the cement paste and associated deterioration of mechanical performance as evidenced by previous studies. A control sample with 0% fibers was made to compare the effects of the wollastonite fiber reinforcement. A constant mass-based water-to-cement (w/c) ratio of 0.3 was chosen for all samples. Metflux© 4930F superplasticizer manufactured by BASF Construction Chemicals was used in the range of 0.5–2.5% of cementitious solids by mass to compensate for the loss in workability due to the addition of wollastonite fibers. The density of the used wollastonite fiber is 2.8 g/cm³, which is similar to that of hardened cement paste (2.4 g/cm³) used in this study. As a result, the addition of wollastonite did not significantly alter the density of the resulting samples. The density of the wollastonite fiber reinforced samples ranged from 2.42 to 2.49 g/cm³, which was within a relatively narrow range. For each of the mixtures, 50 mm cube samples were prepared and the samples were kept in a moist environment containing 95% humidity with a temperature of 23 °C ± 2 °C for 28 days. While some of these samples were used for uniaxial compression tests, small cubes of 4 mm sides were cut from the larger cubes for microindentation experiments. These cubes were polished to a 0.04 μm colloidal silica finish.

2.2. Compressive Strength Experiment

Compression testing of cement mortars was performed in accordance with ASTM C109 guidelines [57] using standard 50 × 50 × 50 mm cube specimens after 28 days of curing. For all the cases six replicate samples were tested. A Shimadzu AGX-225 universal testing machine was set up for compression loading (Figure 1). A loading rate of 1 mm/min was set. Prior to testing, the dimensions of the cubes were accurately measured. The peak load of the samples was divided by the cross-sectional area to obtain the compressive strength.

2.3. Microindentation for Creep Response Evaluation

A Shimadzu DUH-211S dynamic micro-indentation machine (Figure 2) was used to evaluate the creep response of cementitious composites. The machine was equipped with a Berkovich indenter that has a conical angle of 115°. Each sample was held under a constant loading rate of 20 mN/s until a maximum load of 1500 mN was achieved. Next, this force was kept constant for 600 s to replicate creep conditions. For each case, ten indentations were performed by varying the locations. Four replicate samples were tested for each case.

For indentation analysis, Galin [42] has provided the load penetration depth equation for an axisymmetric probe. The contact stiffness can be determined by differentiating the Galin penetration equation as seen in Equation (1) [42,58].

$$S=\frac{dP}{dh}=\beta \left(\theta ,\vartheta \right)\frac{2}{\sqrt{\pi }}{E}_{eff}\sqrt{A_c}$$

(1)

where $E_{eff}$ is the effective modulus (indentation modulus) and A_c is the area of contact. The correction factor, $\beta \left(\theta ,\vartheta \right)$, is 1.034 for a Berkovich indenter. This method originates from the Bulychev–Alekhin–Shorshoro (BASh) equation to determine Young’s Modulus using nanoindentation [59]. The relationship between effective modulus ($E_{eff}$) and contact stiffness ($S)$ as per Oliver and Pharr’s Equation [41] is given as:

$$E_{eff}=M=\frac{\sqrt{\pi }S}{\beta \left(\theta ,\vartheta \right)\sqrt{A_c}}$$

(2)

The slope of the unloading curve of the force-displacement graph represents the contact stiffness from the indentation test. The contact area of the microindenter is determined by:

$$A_c=\pi {\left(h_{c}tan\theta \right)}^2$$

(3)

where h_c is the contact depth along the vertical axis and θ is the effective conical angle of the indenter. Equation (4) relates the effective modulus to the moduli of the material and indenter tip [41].

$$\frac{1}{E_{eff}}=\frac{1-{\vartheta }_{in}^2}{E_{in}}+\frac{1-{\vartheta }^2}{E_{it}}$$

(4)

where $E_{in}$ and ${\upsilon }_{in}$ are the elastic modulus and Poisson’s Ratio of the rigid indenter tip, respectively. For the Shimadzu DUH-211S indenter $E_{in}$ and ${\upsilon }_{in}$ are given as 1140 GP and 0.07, respectively. $E_{it}$ is Young’s modulus of the material of interest, and υ is Poisson’s Ratio of the cement paste. During dwell time the penetration depth of the indenter increases in relation to the creep response. Plasticity is induced at every stage of loading due to the sharp corners of the indenter. The contact creep compliance is given as a function L(t) seen in Equation (5) [60].

$$L\left(t\right)-L\left(0\right)=\frac{2a_c\mathsf{\Delta }h\left(t\right)}{P_{max}} L\left(0\right)=\frac{1}{M_0}$$

(5)

where $a_c$ is contact radius, $\mathsf{\Delta }h\left(t\right)$ is time-dependent indentation depth, $P_{max}$ is the maximum load, and $M_0$ is indention modulus. Equation (6) describes the logarithmic function of $\mathsf{\Delta }h\left(t\right)$ that was introduced by Vandamme during the holding phase [2,61].

$$\mathsf{\Delta }h\left(t\right)=x_1\ln \left(x_{2}t+1\right)$$

(6)

Substituting $\mathsf{\Delta }h\left(t\right)$ into Equation (5), the following expression is obtained:

$$L\left(t\right)-\frac{1}{M_0}=\frac{2a_c{x}_1\ln \left(x_{2}t+1\right)}{P_{max}}= \frac{\ln \left(t/\tau +1\right)}{C}$$

(7)

where $C$ is the contact creep modulus and $\tau$ is the time characteristic given by:

$$C=\frac{P_{max}}{2a_c{x}_1} and \tau =\frac{1}{x_1}$$

(8)

The time characteristic is a long-term creep dominator in the logarithmic model that was created by Vandamme [2,61].

3. Results and Discussion

3.1. Compressive Strength

The 28-day compressive strengths of various wollastonite-reinforced hardened cement pastes are plotted in Figure 3 along with the control sample. In line with ASTM C109, this mechanical property is achieved by locating the maximum stress before failure [57]. As can be seen in Figure 3, the compressive strength increases significantly with increasing wollastonite fiber content for up to 10% fiber content beyond which the strength starts dropping. About a 35% increase in compressive strength is observed for 10% wollastonite dosage. Such strength enhancement can be explained by the fact that stiffer wollastonite fibers can carry significantly higher stress without breaking than the hardened cement paste matrix which ultimately reduces stress on the matrix and prevents it from cracking at smaller loads thereby increasing the strength capacity of the composite. Beyond, 10% wollastonite content, any further reduction in cement content translates into strength loss when compared to the best-performing wollastonite-fiber-reinforced cementitious composite. At 15% wollastonite dosage, the compressive strength is still higher than the control sample. The error bars in Figure 3 (and all the figures in this paper) range across all replicate samples. Tukey’s HSD (honestly significant difference) test on the compressive strength data reveals a p value of less than 0.00001 which suggests that the compressive strength results for mixtures with varying wollastonite fiber dosage are statistically different.

3.2. Creep Response from Micro Indentation Experiments

The values for elastic modulus, hardness, creep modulus, and time characteristics from microindentation experiments are reported in this section. The indentation modulus of the wollastonite-reinforced hardened cement paste, based on Equation (2), is plotted as a function of fiber content in Figure 4a. Considering average values, a comparison between 0% and 10% wollastonite fibers reveals a slight linear increase in the modulus of elasticity. Beyond 10% the detrimental impact of binder loss in the cement becomes significant. Thus, a reduction in the modulus is observed. A similar trend is observed in the study by Low and Beaudoin where an optimal dosage of 11.5% cement replacement was reported [34,35,36]. This trend is also reflected in the hardness values shown in Figure 4b. Overall, to elucidate the statistical significance of the modulus and hardness values, Tukey’s HSD test was conducted on the data. While the differences in Young’s modulus values were not found to be significant (p = 0.096), the difference in hardness values with change in wollastonite fiber content was significant (p = 0.0414) at p < 0.05.

The creep behavior of a material is represented by two main parameters: creep modulus and the time characteristic. Figure 5a shows the result of the creep response using Equation (8) for the wollastonite-reinforced cementitious composites. A similar trend to that of the elastic modulus and hardness values of the data is seen. Compared to the control sample, the 10% wollastonite replacement of cement has a 30% increase in creep response. This can be attributed to the crack-tolerance enhancing effect of the wollastonite fibers through micro-reinforcement as established elsewhere [39]. Figure 5b presents the creep response in terms of time characteristic (τ(s)) which suggests how creep progresses throughout a structure’s service life. This function, τ(s), has been derived from the logarithmic behavior of Δh(t) in Equation (6). During dwell time, the creep modulus determines how much load the samples can withstand before this mechanism starts to cause high strain and deformation. Later, time characteristics dictate the creep curve response. As exemplified in Figure 5b, the material time characteristic does not increase with wollastonite fibers in the system. After a certain amount of replacement, this feature falls due to the binder replacement. Please note that the wider spread of the data is due to the highly heterogeneous nature of hardened cement paste at a small length scale, which is investigated using microindentation technique. This results in more variable data compared to macroscale experiments, which generally produce more consistent results. To shed more light on the statistical significance of the trends, Tukey’s HSD test was conducted on the creep modulus as well as time characteristic data. The creep modulus data showed very clear significant difference in results between various wollastonite fiber dosage cases with a p value of 0.000026. On the other hand, the results for time characteristic data showed insignificant difference with a much higher p value of 0.44.

Figure 6 presents the inter-relationships between indentation modulus, hardness, creep modulus, and time characteristic. As can be seen in Figure 6a, the relationship between indentation modulus and creep modulus follows an increasing trend where an increase in indentation modulus in wollastonite-fiber-reinforced cementitious composites translates into an improvement in creep behavior. Here, a linear relationship is observed between Indentation Young’s modulus and creep modulus with an R² value of 0.8. In terms of indentation hardness, however, such trend is not observed as can be seen in Figure 6b. Figure 6c,d present the relationship of time characteristic with respect to indentation modulus and hardness, respectively, where no trend was found with respect to varying wollastonite fiber dosage. These results further reinforce the fact that wollastonite microfiber does not influence the time characteristic significantly and it primarily resists creep-induced microcracks by virtue of significant enhancement in strength and fracture toughness when compared to the control-hardened cement paste.

4. Conclusions

This study systematically implements microindentation experiments to elucidate the influence of wollastonite fiber incorporation on the creep response of hardened cement paste. The fiber content in the specimens ranged from 0% to 15%. Each sample was cured for 28 days in a humid environment. The results establish the effectiveness of the wollastonite fibers in improving the creep behavior of hardened cement paste. The following conclusions are drawn from this experimental study:

• The compressive strength of the hardened cement paste increases with the introduction of wollastonite fibers into the matrix. The mixture containing 10% wollastonite fibers as mass-based cement replacement was found to be optimal in terms of compressive strength.
• Enhancement in the indentation modulus and hardness was observed with up to 10% wollastonite fiber content. While the average elastic modulus increased from 24 GPa to 26 GPa with 10% wollastonite fiber incorporation with respect to control-hardened cement paste, the average indentation hardness increased from 0.6 GPa to 0.68 GPa for the same dosage of wollastonite fibers.
• The creep modulus showed significant improvement with increasing wollastonite content up to 10%. This can be explained by two facts. First, there are chemical similarities between the wollastonite fiber and the C-S-H in cement paste. Because of this chemical similarity, wollastonite fibers naturally have better compatibility and bonding with the cement paste as explored and established elsewhere [62,63,64,65]. A better bond between the fiber and the matrix likely translates into better creep performance. Second, the micro-reinforcing effect [39] of the wollastonite fibers arrests creep-induced microcracking in the hardened cement paste thereby improving the creep-tolerance of the composite. Overall, the significantly higher creep modulus observed in the wollastonite fiber reinforced samples compared to the control samples suggests that the incorporation of fibers has effectively reinforced the material, enabling it to better withstand sustained loading and resist excessive deformation. Besides, the practical applications of wollastonite in terms of significant cement replacement also reduce cement consumption, which causes 5% of global dioxide emissions.
• The fact that no significant difference in the time characteristic (i.e., the shape of the creep curve) is observed between the control and fiber reinforced samples suggests that the fibers have not significantly affected the time-dependent behavior of the material. This means that the material still likely undergoes the same three stages of deformation (primary, secondary, and tertiary creep) over time under sustained loading, but with a higher resistance to deformation.

Overall, the improved creep modulus in the wollastonite fiber reinforced samples indicates that the addition of wollastonite fibers has improved the material’s ability to withstand sustained loading with significantly higher resistance to deformation over time, which can be beneficial for applications where long-term structural stability is important.