Performance of green high-strength concrete incorporating palm oil fuel ash in harsh environments

The concretes in this study were produced using OPC river sand fine aggregate, granite coarse aggregate, U-POFA, and superplasticizer.

2.1.2.

Aggregates

The fine aggregates were natural river sand with a specific gravity of 2.75, water absorption of 0.61%, and a fineness modulus of 3.10, complying with the requirements of ASTM-C128 [28] and ASTM-C136 [29]. The coarse aggregate used was granite crushed with a maximum size of 12.5 mm, and the bulk density was 1550 kg/m³, according to ASTM-C29 standard [29]. The specific gravity was 2.7, and water absorption was 0.49%, based on ASTM-C127 [30]. A sieve analysis was conducted on aggregates in order to generate the gradation curves according to ASTM-C33/C33M [31], as depicted in Figure 1.

Fig. 1.

2.1.3.

Steel bars

This investigation involved the utilization of a typical steel bar (rod) of 10.0 mm in diameter for the purpose of conducting corrosion tests, simulating the conditions experienced by embedded steel within a HSC specimen. In this particular scenario, in order to achieve precise outcomes, it is imperative to pre-clean the steel rods. Consequently, the cleaning procedure was implemented in the following manner: Initially, the steel rods underwent a brushing process using an electrical brush equipped with a broad wired head. Subsequently, in order to eliminate any oil or other contaminants, the bars underwent a washing process with acetone, followed by a thorough drying using a pristine towel. Figure 2 depicts the steel bars utilized in this study, showcasing a visual contrast between their condition prior to the cleaning phase and their condition following the cleaning phase.

Fig. 2.

2.1.5.

Palm oil fuel ash

This study utilized the procedure previously introduced by Johari et al. [13] for the production of the U-POFA. The original POFA (O-POFA) was obtained from a local palm oil mill in Penang, Malaysia. It was then dried in an electric oven at 100 ± 5°C for 24 hours to eliminate any moisture. The dried O-POFA was then passed through a 300-μm sieve to separate out the larger particles. Afterwards, the O-POFA underwent grinding, using a ball milling to create the ground POFA (G-POFA). The G-POFA was then heated in a furnace at 500°C for 90 min to reduce the unburnt carbon content and produce the treated POFA (T-POFA). Finally, the T-POFA was further ground in the same ball milling machine to obtain the U-POFA. The uniformity and consistency of the U-POFA produced were controlled by taking several measures, such as the amount of POFA put in the ball milling machine, the milling speed, the grinding time, the quantity of the POFA in the furnace, the heating time, and the heating temperature. The production process of the U-POFA is depicted in Figure 3.

Fig. 3.

In this study, the initial target strength for the concrete specimens across all groups was set to be a minimum of 80 MPa at 28 days with a 0.22 waterbinder ratio (W/B ratio). Four concrete mixes were prepared with the proportions presented in Table 1.

The first mix was the control with 100% OPC, while the OPC was replaced with various ratios of U-POFA in other mixes. Three replacement ratios of U-POFA were used: 20% (20UP), 40% (40UP), and 60% (60UP), as shown in Table 1. The mixing procedure commenced by introducing the predetermined quantities of fine and coarse aggregates into the pan-type mixer. Subsequently, the OPC or binder was included in order to facilitate thorough mixing. The duration of this step typically ranges from three to four minutes, during which the aggregates and binder are thoroughly mixed to achieve a homogeneous dry mix. Subsequently, the water and the superplasticizer (SP) were included in the mixture, and the mixing procedure was sustained for a duration of approximately four to six minutes.

The concrete mixture was placed in the molds in three layers, and then subjected to shaking using a vibrating table to eliminate air bubbles and obtain a homogeneous mixture. The molds were allowed to sit for a period, after which they were then covered with a moist, dense fabric and permitted to harden for a duration of 24 hours. On the subsequent day, the specimens were extracted from the molds and subsequently immersed in a water tank for curing purposes. The samples were then left to cure for specified periods of 7, 28, 60, 67, and 90 days.

Table 2 displays the changes observed in the physical characteristics of the POFAs during all stages. The U-POFA exhibited the smallest median particle size of 2.21 μm. When compared with the median particle sizes of the O-POFA and OPC, it is obvious that the U-POFA achieved a size reduction of 86.73% and 78.29%, respectively. Moreover, the SEM micrographs, which show the morphology of the particle size of the different types of POFA, are depicted in Figure 5 in which a notable reduction in the grain dimensions of the POFA particles can be seen at various phases of the treatment procedure. This aligns with the numerical outcomes presented in Table 2. The chemical compositions of OPC, G-POFA, and U-POFA determined via XRF analysis are presented in Table 3. The findings indicate that the treatment procedure resulted in an improvement in the chemical composition of the G-POFA and significantly reduced the carbon content by 85.4% (from 30.60% to 4.74%). In addition, a reduction of 87.12% was noticed in the content of the LOI between the G-POFA and the U-POFA. Thus, the treatment of POFA contributed to the enhancement of its chemical compositions and gave it significant potential to have superior resistance against corrosion and aggressive environmental conditions.

Fig. 5.

The slump results for all HSGC mixes are depicted in Figure 6. The data presented in the table indicate that when compared to the control concrete (i.e., OPC), the slump test value increases by 25 mm, 40 mm, and 45 mm for HSGC mixes 20UP, 40UP, and 60UP, respectively. The slump increased by 13.5%, 18.9%, and 21.6% for 20UP, 40UP, and 60UP, respectively, compared with OPC. However, the rate of increase has experienced a minor decrease compared to a lower rate of replacement percentage of U-POFA, which is similar to the finding reported in previous research [14].

Fig. 6.

According to Johari et al. [14], the occurrence can be attributed to the U-POFA having a higher surface area and lower specific gravity in contrast to OPC. The inclusion of U-POFA in HSC leads to a larger paste volume of binder compared to HSC consisting only of OPC. As a result, the reduced grain size of U-POFA, as compared to OPC, may produce enhanced workability and slump rates. Hence, it is advisable to employ an elevated amount of U-POFA in order to reduce the dose of superplasticizer or the water/binder ratio, contingent upon the target slump value.

As Figure 7 shows, the CS of HSGC specimens deteriorated significantly with the inclusion of the U-POFA, especially at an early age. For instance, at 7 days, the CSs of the 20UP, 40UP, and 60UP were 76.03 MPa, 70.63 MPa, and 60.1 MPa, respectively. Similar results were reported in a few related research reports [25]. The reason behind this deterioration might be due to the dilution effect caused by the partial replacement of OPC with U-POFA. However, with prolonged curing, the percentage of relative strength of the U-POFA specimens increased and surpassed that of the OPC sample as a result of the pozzolanic reaction between the U-POFA and calcium hydroxide (CH). After a period of 90 days, the findings indicate that the strength of the HSGC exhibits improvements when U-POFA is used as a replacement for OPC by 20%, 40%, and 60%. This suggests that incorporating U-POFA in high volume has a positive impact on the CS of HSGC. Meanwhile, it is evident that a substitution of up to 60% U-POFA in terms of weight of OPC has the potential to improve the CS of the HSGC to a comparable extent as concrete consisting only of 100% OPC after a period of 90 days, as shown in Figure 7a. With U-POFA replacing 20% of OPC, Figure 10b shows that the CS of the concrete specimens is higher after 7 and 28 days, though only slightly higher than the concrete made with 100% OPC. Nevertheless, after 90 days, the strength exhibits a notable increase. Despite observing a decrease in CS at 7 and 28 days when the replacement rate of OPC was increased to 40%, a slightly higher CS was recorded at 90 days compared to the control concrete. The HSGC’s strength decreased more when 60% of U-POFA was added, and this decrease is more noticeable at 7 and 28 days of curing compared to all other samples. It is noteworthy, however, that the strength continues to increase from 7 to 28 days. Nevertheless, after a curing period of 90 days, the 60UP mixture exhibited a CS that was comparable to that of the OPC concrete and approached nearly the same level of CS, as shown in Figure 7. However, it is important to mention that according to this study, the utilization of 20% substitution of U-POFA did not result in a reduction of CS over the 7, 28, or 90 days. In fact, it significantly improved the strength. Hence, in situations where there is a requirement for enhanced CS during the early ages, it is advisable to ascertain and implement the optimal volume of U-POFA replacement. In this investigation, it has been shown that a replacement dose of 20% is indicated. Similar results were reported by several researchers in related studies [17]. Jaturapitakkul et al. [15] observed the effect of fineness and filler effects on the enhancement of CS in concrete, including concrete incorporating G-POFA. Therefore, in order to achieve this objective, it is crucial to examine at least two key aspects. Firstly, the utilization of SCM possessing enhanced pozzolanic characteristics has the potential to increase the formation of calcium silicate hydrate (C-S-H) in the cement pastes. This phenomenon holds significant advantages in terms of enhancing concrete strength. This particular possibility arises within the context of HSGC that incorporates SCMs. Additionally, the second component to consider is the particle size of the pozzolanic materials, whereby smaller particles contribute to the formation of a denser microstructure via the micro-filler effect [25]. Based on previous studies, increasing the replacement content to 60% by weight of OPC led to a minor decrease in the CS, which was 105.2.5, 109.9, and 108.6 MPa, to replacement rates of 20%, 40%, and 60%, respectively [14]. This is consistent with the results obtained in this study, which indicate that a replacement rate of 60% by weight may be sufficient to maintain high CS in addition to satisfactory transport and durability properties. CS, in addition to durability, is an important characteristic and criterion for producing HSGC [25].

Fig. 7.

Table 4 presents the water absorption and porosity results of all HSGC mixes. Based on the porosity test findings presented in Table 4, the OPC concrete sample exhibited a porosity of 12.3% after 7 days of curing. In contrast, the 20UP, 40UP, and 60UP samples displayed porosities of 11.6%, 9.6%, and 8.8%, respectively, with reduction percentages compared to the OPC sample of 5%, 22%, and 28%, respectively. Looking at the results after 28 days of curing, it is evident that the samples of all HSGC mixes that contain U-POFA become less porous as the curing time increased. The reduction values reached 13%, 15%, and 32% for 20UP, 40UP, and 60UP, respectively, as compared to the OPC specimens. In the case of the porosity after 90 days, the reduction values reached 23%, 36%, and 47% for 20UP, 40UP, and 60UP, respectively, as compared to the OPC sample. The significant decrease in porosity clearly illustrates that the use of U-POFA has the capacity to effectively diminish the porosity of HSGC. Moreover, the findings of the present study align well with those of related prior research works [14]. These studies suggest that the fineness of POFA has a direct impact on pore size, indicating that increased fineness can effectively mitigate the aforementioned problems. Therefore, it can be observed that the HSGC incorporating U-POFA demonstrates a reduced porosity rate when subjected to a longer curing period as compared to OPC concrete. Achieving a larger decrease rate can be accomplished by employing higher replacement levels of U-POFA. Regarding the water absorption after 7 days of curing, the reduction values reached 11%, 24%, and 33% for 20UP, 40UP, and 60UP, respectively, as compared to the OPC sample. The same trend was noticed after 28 days of curing, where the reduction values reached 15%, 16%, and 34% for 20UP, 40UP, and 60UP, respectively, as compared to the OPC sample. Moreover, after 90 days of curing, the 20UP, 40UP, and 60UP exhibited reduction values of 22%, 34%, and 45%, respectively, as compared to the OPC sample, as shown in Table 4. Therefore, it can be shown that the water absorption of HSGC samples follows a similar trend to the porosity. Specifically, as the samples age and the U-POFA replacement level increases, the water absorption decreases in contrast to the samples containing OPC. Similar results were reported by several researchers, such as Kroehong et al. [37].

Figure 8a depicts the average initial currents of the samples obtained from the IVT. It is evident that the initial currents of the OPC samples were the highest at all ages. For instance, at 7 days, the initial current of the OPC specimen was the highest at 23.3 mA, over the experiment, the 20UP, 40UP, and 60UP had initial currents of 13.5 mA, 10.3 mA, and 8.9 mA, respectively. The same trend was also noticed at the other three ages. Therefore, it can be concluded that the initial current was inversely proportional to the amount of U-POFA in the mixtures, indicating that the use of U-POFA enhanced the permeability properties of the HSGC., that is, making the concrete more impermeable. Similar results were reported by Topçu and Boğa [35], who reported that including 10% and 20% fly ash as a replacement for OPC decreases the primary currents and corrosion rate in comparison to concrete zero-FA. The reduction in the initial current of HSGCs with the inclusion of U-POFA could be linked to the refinement in the pore structure of the HSGC containing U-POFA as a result of its effective pozzolanic reaction with CH to produce C-S-H, particularly at higher contents of U-POFA. Earlier, Johari et al. [14] reported significant improvement of transport properties of HSGC containing U-POFA. Hence, the results of the current study corroborate the findings of Johari et al. [14] on the improvement of the HSGC transport properties after the inclusion of U-POFA. The steady state of the current time was also measured on the basis of the start and end times of the IVT. Okba et al. [38] stated that the longer the steady state of the current time during testing, the better the corrosion resistance of the concrete is. Table 5 and Figure 8b present the start-end times of the steady state and the average currents at the steady state, respectively. The results show that the OPC sample exhibited the shortest durations with the highest currents at both initial and steady state stages compared with the samples containing U-POFA at all ages. This proves that the corrosion-resistant performance of the HSGCs containing U-POFA is superior to that of the OPC specimen, with much superior corrosion-resistant performance at higher contents of U-POFA. Okba et al. [38] noticed a sharp rise in the current during the IVT measurements, accompanied by multiple simultaneous cracks in the samples. In this study, the sharp rising phases with crack time were investigated. Figure 9 shows the cracks that occurred during the IVT at 7, 28, 60, and 90 days. The occurrence time of the initial cracks is indicated by arrows (Fig. 10) and illustrated in Table 5. Figure 9 and Table 5 show that all the OPC samples at 7, 28, 60, and 90 days. Conc20133 experienced cracks after 176 h, 188 h, 229 h, and 236 h, respectively. It is noticed that the time to the initial crack increased with age (curing time) due to the reduction in porosity and the increase in strength of the samples with continued hydration over a longer curing period. This improvement can be clearly noticed in the specimens with U-POFA, which could be attributed to the refinement in the structure of the pores as a result of the pozzolanic reaction of the U-POFA, especially when the replacement level of U-POFA is high. At 7 and 28 days, the 20UP samples cracked after 234 h and 353 h, respectively, but no sign of cracks was detected at 60 and 90 days. Chindaprasirt et al. [39] employed the G-POFA to enhance the corrosion resistance of HSC with replacement levels of 10, 20, and 30% by mass of OPC at 7, 28, and 90 days. At a 20% replacement level, the cracks were initiated after 180 h, 220 h, and 275 h at 7, 28, and 90 days, respectively. By comparing the results, the inclusion of U-POFA in this study delayed the crack time by 23%, 38%, and 100% at 7, 28, and 90 days, respectively. This indicates that U-POFA is more efficient than the G-POFA in terms of enhancing the corrosion resistance of HSC. For the 40UP samples, a crack only occurred after 245 h, i.e., at 7 days. In the case of the 60UP samples, however, no crack was observed at any age, serving as more evidence that increasing the volume of the U-POFA contributed greatly to the prevention of corrosion and the formation of cracks.

Fig. 8.

Fig. 9.

Fig. 10.

Table 6 presents the rates of corrosion determined for the OPC, 20UP, 40UP, and 60UP specimens at 7, 28, 60, and 90 days. The results show that the rates of corrosion decreased with the increase in U-POFA content at all ages of the HSGCs. Compared with the corrosion rate of the OPC at 7 days, those of the 20UP, 40UP, and 60UP decreased by 46%, 67%, and 91%, respectively. The corrosion rates observed for OPC, 20UP, 40UP, and 60UP specimens at the 28-day mark were 6.945, 1.488, 0.717, and 0.331 mm/year, respectively. The observed reductions in corrosion rate for the 20UP, 40UP, and 60UP samples were approximately 79%, 90%, and 95% of the corrosion rate exhibited by the OPC sample. The same behavior was displayed at 28, 60, and 90 days. This could be attributed to the fact that the U-POFA decreased the pore size and permeability of the concrete via the pozzolanic reaction, resulting in lower currents and corrosion rates. Figure 10 depicts all the samples after the mass loss test. The presence of corrosion products is evident for the OPC sample, and the reinforcement bar exhibits a significant degree of corrosion, as shown in Figure 10a. Based on the visual analysis of the 20UP sample, it is readily apparent that the reinforcement exhibits a relatively confined degree of corrosion, as shown in Figure 10b. It can be noticed that the amount of corrosion on the reinforcement bars of the 40UP and 60UP specimens is almost negligible to none, as shown in Figure 10c, d.

SEM micrographs in Figure 10 illustrate the effect of incorporating U-POFA as a substitute for OPC, with varying proportions of 0%, 20%, 40%, and 60% by weight, on the microstructural characteristics. From the morphology shown in Figure 11a–d, it is observed that the microstructures of cement paste have different shapes, and the changes in the microstructures depend on the rate of replacement of U-POFA with OPC. It is evident that the rise in the rate of OPC replacement by U-POFA led to an improvement in the microstructure with higher density (low permeability) through a decrease in the pore size and bonding of the cement paste compared to the cement paste zero-U-POFA or low content (20%) U-POFA. Two main compounds are made when OPC hydrates: C-S-H gel and CH. Utilizing supplementary cementitious materials like U-POFA typically has a substantial influence on the microstructures of the cement paste matrix [14]. The utilization of the ultrafine pozzolanic material enhances and purifies minute structures at a microscopic level. The large amount of silicon dioxide (SiO2) and noncrystalline aluminum oxide (Al₂O₃) in the POFA causes a chemical reaction with CH, leading to the creation of calcium silicate hydrate (C-S-H), calcium alumino-silicate hydrate (C2ASH8), and calcium aluminate hydrate (C4AH13). The pozzolanic reaction has a dual effect: it reduces the quantity of CH, while increasing the quantities of C-S-H and C₂ASH₈, and thus lowering the amount of ettringite [16, 24]. The interaction between CH and amorphous SiO2 and Al2O3 leads to the creation of supplementary C-S-H and calcium-aluminate-silicate-hydrate (C-S-A-H) gels. This technique further improves the microstructure and plugs the pores.

Fig. 11.