Enhancing performance and sustainability of GGBFS-based self-compacting geopolymer concrete blended with coal bottom ash and metakaolin by using RSM modelling

Abstract

This research study is performed on the self-compacting geopolymer concrete (SCGC) combining coal bottom ash (CBA) and metakaolin (MK) as a substitution for GGBFS alone and combined for analysing the fresh properties (slump flow, V-Funnel, and T50 flow), mechanical characteristics (compressive, splitting tensile and flexural strengths) and durability tests (permeability and sulfate attack test). Though, total 195 SCGC samples were made and tested for 28 days. It has been revealed that the consumption of CBA and MK as a substitution for GGBFS alone and combine in the production of SCGC is decreased the workability of SCGC while mechanical characteristics of SCGC are enhanced by utilizing CBA and MK as a substitution for GGBFS alone and combine up to 10%. In addition, the compressive, splitting tensile and flexural strengths were calculated by 59.40 MPa, 5.68 MPa, and 6.12 MPa while using the 5CBA5MK as a substitution for GGBFS in the production of SCGC after 28 days correspondingly. Furthermore, the permeability is decreased by growing the quantity of CBA and MK by the weight of GGBFS alone and jointly in the production of SCGC after 28 days. Besides, the minimum change in length of the SCGC specimen is recorded by 0.062 mm at 7.5MK7.5CBA while the maximum change in length is calculated by 0.11 mm at 10CBA10MK as a substitution for GGBFS at 180 days correspondingly. In addition, the embodied carbon is recorded reduce as the addition of CBA while it is getting higher when the accumulation of MK alone or combined with CBA in SCGC. Besides, response models for prediction were constructed and confirmed using ANOVA at an accuracy rate of 95%. The models' R² fluctuated from 88 to 99%. It has been observed that the utilization of CBA and MK alone and together up to 10% as substitution for GGBFS in geopolymer concrete provides the best results therefore it is suggested for structural applications.

Introduction

Concrete is frequently employed in construction applications throughout the globe¹. Portland cement (PC) is an imperative component of concrete, and it is not universally sustainable. Sustainability issues associated with Portland cement production are well-known. Portland cement production discharges large quantities of carbon dioxide (CO₂) and additional dangerous gases into the environment due to the mechanism of decarbonation of limestone and the burning of energy sources such as fossil fuels. 1 tonne of PC needs approximately 1.6 tonnes of raw materials and 6.5 million British thermal units of energy, resulting in the release of 1 tonne of CO₂ into the natural environment^2–4.

Globally, the PC industry releases about 1.65 billion tonnes of global warming annually, contributing to about 7% of global CO₂ emissions^5–8. The CO₂ production from the manufacture of PC is projected to increase by approximately 50% from existing levels by 2050^9,10. To protect the environmental issues from the impacts of PC production, new options for producing a more economical and environmentally friendly concrete material to substitute standard PC concrete must be identified and evaluated¹¹. To resolve global warming concerns, enormous initiatives have been made to minimize the usage of PC on a global scale. Among these are the recycling of waste by-products and the production of PC replacements¹².

Geopolymer concrete (GC) is a significant advancement in unique product development, resulting in an economical, environmentally acceptable constituent that can be used in place of PC^12–15. It is an alumino-silicate-based synthetic polymer concrete that can be produced in a strong alkaline solvent using silicon and aluminum-rich geological elements or by-product constituents¹⁶. The GC is invented by substituting the PC entirely^17–21. Apart from PC, the manufacturing of natural components for geopolymers are not necessarily involve a lot of energy consumption. The binder used in geopolymer has been proven to discharge between a 5th to 6th of the Carbon dioxide generated by Portland cement²². As a result, geopolymer concrete not only minimizes CO₂ production associated with the PC sector, and waste materials residues of alumino—silicate concentration are also used to make high-performance constituents for construction^23,24.

Moreover, GGBFS is obtained through the collection of waste components from a particular industry, namely metallurgical. The slag is produced in the form of molten iron, which is formed when a mixture of components such as coke, iron ore, and limestone is heated to between 15,000 and 16,000 degrees Celsius. The resulting slag has a chemical structure alike to PC, with approximately 45% CaO and 35–45% silicon dioxide. Additionally, glassy particles are formed when silicon-aluminum slag is rapidly immersed in liquid form after molten iron is removed^25–28. Dewatering and converting such glassy particles to the desired size^29,30. This crushed slag is referred to as GGBFS, and because it is an environmentally friendly building material, which can be utilized in place of PC in concrete, thereby reducing carbon dioxide production and PC use in the construction industry³¹. GGBFS can enhance durability assets, resistance to corrosion, and sulfate resistance. By considering these properties, it is feasible to prolong the infrastructure life and reduces maintenance budgets. Due to the high concentration of GGBFS in the eco-friendly circumstances utilized to substitute cement, concrete not only recycles waste but also conserves natural assets and reduces energy utilization^32,33. As a consequence, GGBFS was preferred as the starting material for geopolymer synthesis in this investigation to maximize the utilization of this industrial waste and to produce geopolymer concrete based on GGBFS.

The enhancement of concrete construction throughout the last 2 decades has increased focus on improving efficiency and striving to improve the effective atmosphere. Self-compacting concrete (SCC) is a comparatively novel improvement in the field of concrete production which is gaining traction. Distinct from conventional concrete, SCC has the capability to flow through and around obstacles by itself, completely covering the formwork and self-compacting without segregation or blocking³⁴. It has a variety of technical, economic, and ecologic benefits over existing concrete, like enhanced concrete quality, decreased duration of construction, relatively easy installation in clogged reinforcement, and consolidation completion, enhanced bond strength, reduced overall costs, and a quieter working environment due to the absence of vibration^35–37. This results in a reduction in the presence of guidance from a professional, along with a decreased tolerance for pictorial errors, excessive noise, and workplace injury. The essential components in the production of SCC are the same to plane concrete, with the exception that SCC is composed of a variety of different percentages and supplementary mineral admixtures blends. SCC mixtures typically contain a larger quantity of ultra-fine constituents. SCC has superior flowability owing to its lesser concentration of coarse aggregates, lesser water/binder ratio, and utilization of superplasticizers and viscosity-modifying agents^38,39. The cement substitution ingredients are extensively utilized to decrease costs and increase the performance of fresh and hardened concrete^40,41. The usage of SCMs like groundnut shell ash (GSA), GGBFS, fly ash (FA), coconut shell ash (CSA), millet husk ash (MHA), and silica fume (SF) is well recognized for improving concrete properties and for environmental and economic reasons³⁴.

Metakaolin (MK) is created by burning kaolin clay at temperatures fluctuating from 650 to 800 °C. It has possessed the pozzolanic property, therefore, MK can be used as a SCM in concrete, and when it is mixed in concrete as cementitious material that improves the ITZ of concrete compared to other ingredients which results in improving the strength⁴². For decades, MK has been utilized commercially in construction sector⁴³. The majority of studies used concrete containing MK as supplementary cementitious material (SCM), showing a notable enhancement in the mechanical characteristics. Poon et al.⁴⁴ performed research work on MK as SCM in concrete which enhanced the mixture’s mechanical properties while decreasing its porosity. It has been realized that by using MK dosages enhanced in concrete, pore spaces are minimized. Jin and Li⁴⁵ explained an associated tendency. Ahmed et al.,⁴⁶ stated that the consumption of MK up to 15% in matrix increases crushing and flexural strength. The majority of prior investigation has focused on using MK in geopolymers mixture^47,48. MK is a crucial strategy for reducing the carbon footprint of concrete^47,48.

Moreover, the CBA is produced when coal is burned in thermal power plants which is the primary waste product of fly ash (FA). Owing to the existence of SiO₂, Fe₂O₃, and Al₂O₃ in the CBA, it possesses pozzolanic properties. The calcium hydroxide reacts with calcium aluminate hydrate (CAH) throughout cement hydration to form supplementary CAH and calcium silicate hydrate (CSH)⁴⁹. According to prior research⁴⁹, suitable crushing can enhance the CBA pozzolanic activity, and six hrs of crushing CBA can result in a 27% growth in strength. As a result, it is partially substituted in concrete by FA and PC⁵⁰. CBA usage in building sector is one of the most effective ways to mitigate environmental issues due to a lack of discharge sources and augmented CBA production. India generates almost 105 million metric tons of CBA every year for many thermal power plants, which provide 68% of the nation’s energy production. The inclusion of CBA in the current and future building industries is being advocated for owing to its comparable particle size distribution and supplementary pozzolanic characteristics to NFA⁴⁹.

Furthermore, based on the literature study, there are no experimental investigations performed on SCGC mix inclusion with MK and CBA as a substitution for GGBFS individually and combine. Therefore, this study examined the fresh properties, mechanical properties, durability properties and embodied carbon of GGBFS-based SCGC inclusion with MK and CBA as a substitution for GGBFS. The outcomes demonstrate a noteworthy enhancement in the strength properties of GGBFS-based SCGC blended with MK and CBA separate and combined and these materials can be advantageously utilized in the production of SCGC mixture by applying response surface methodology (RSM) tool for modelling and optimization.

Concept of the geopolymerization mechanism

The concept of the geopolymerization mechanism is fundamental to understanding the formation and performance of SCGC with MK and CBA as substitutions for GGBFS. Geopolymerization is a complicated chemical process that starts with aluminosilicate materials dissolving in an alkaline solution⁵¹. The ions that are dissolved then condense into a stable network of three-dimensional tetrahedral structures^52,53. The reaction of the alumina and silica available in the constituents (like MK and CBA) with the alkaline activator (in this case, sodium carbonate) drives this process. The resulting geopolymer matrix is composed of interconnected Si–O–Al bonds, forming a rigid, stable structure that contributes to the material’s mechanical strength and durability^54–57.

The inclusion of MK and CBA in the SCGC affects the geopolymerization process and the final microstructure. Metakaolin, because of its high reactivity and small particle size, easily dissolves in alkaline conditions, delivering a plentiful supply of reactive alumina and silica. This promotes the creation of aluminosilicate gels, which are essential for the construction of the tetrahedral network⁵⁸. While CBA has lower reactivity than MK, it plays a role in the overall matrix by occupying empty spaces and decreasing porosity, potentially enhancing the strength and longevity of the geopolymer concrete. The synergistic impact of MK and CBA results in a more condensed and tightly packed microstructure, characterised by a well-established tetrahedral network that exhibits high resistance to environmental factors such as water and sulfates⁵⁸.

The geopolymerization mechanism clarifies the several steps involved in the geopolymerization process, starting with the initial dissolution of aluminosilicate minerals and culminating in the production and expansion of the tetrahedral network^59,60. Additionally, it might elucidate how the distinctive attributes of MK and CBA contribute to this procedure and how the fine-tuning of their proportions impacts the overall efficacy of the SCGC. This feature would enhance readers' comprehension of the development of the material’s microstructure and its correlation with the observed mechanical characteristics and durability by offering a clear and comprehensive explanation of the underlying processes. Gaining a more profound comprehension of this subject matter will not only enhance the present investigation but also provide direction for future investigations and actual implementations in the realm of sustainable construction.

Experimental program

Materials

The GGBFS was obtained with permission from the steel mill and thereafter dried for 24 h in the open environment. After sieving the dried GGBFS from a 75 µm mesh to disregard any undesired particles, it is applied as a binding constituent in SCGC. However, in this investigational study, metakaolin (MK) served as a substitution for GGBFS in the production of SCGC. It was obtained and then dried for 24 h under the environment before being sieved from a 75 µm mesh to disregard unwanted constituents and then it was utilized in SCGC as a substitute for GGBFS material. Additionally, coal bottom ash (CBA) was stored with permission from the Coal Power Plant. After collecting the CBA, it was dried for 24 h at 105 °C and then allowed to pass from a 75 µm mesh to disregard bulky particles. Sieved ash was then utilized in place of GGBFS in the production of SCGC. Table 1 summarises the chemical configurations of GGBFS, MK, and CBA. Furthermore, both the manufacturing of geopolymers is needed to absorb alumina, and silica, and the polymerization catalyst procedure, which is required for an alkaline solution⁶¹. This research was using an alkaline liquid composed of sodium silicate (SS) and sodium hydroxide (SH). For this experimental investigation, SH and SS solutions were used. To prepare an SH solution, dissolve SH particles in enough drinkable water to reach the desired concentrations and left the solution for 24 h to cool down. Besides, to develop an alkaline solution by combining the two aqueous mixtures. Additionally, the superplasticizer (SP) is added for increasing the fresh property which is to the level needed for SCGC mixture compliance with the EFNARC³⁶ code procedure. Additionally, in this investigational study, drinking water was provided to create the alkaline solution. In addition, the coarse aggregates (CA) with a 12 mm particle size and fine aggregates (FA) with possess 4.75 mm in size were applied for this study.

Table 1.

Oxides’ composition of GGBFS, MK, and CBA^62–66.

Binding materials	Oxides (%)						Physical Property
	SiO2	Al2O3	Fe2O3	CaO	Na2O	SO3	Specific Gravity
GGBFS	37.22	10.37	1.23	35.66	0.23	0.34	2.25
CBA	36.67	26.13	14.02	5.11	2.54	0.90	2.15
MK	62.18	21.67	3.01	3.22	1.04	0.78	2.60

Mix proportions

This experimental procedure was adopted to thirteen SCGC mixes, one of which was composed entirely of GGBFS as a binding material, four of which contained 5–20% GGBFS replaced with MK, four of which contained 5–20% CBA as a substitution for GGBFS, and four of which contained a combination of MK and CBA as a replacement for GGBFS. All SCGC mixes are examined in this investigational study, as seen in Table 2. Though, to generate a homogeneous mixture, dry materials such as GGBFS, MK, CBA, and aggregates are mixed in a concrete mixer for approximately 2.50 min in the dry state, monitored by the accumulation of a well-shaken premixed solution combination comprising an alkaline solution, SP, and additional water for three minutes. The SCGC mixture was characterized after the homogeneous SCGC mixture was examined for key tests on workability such as slump flow, V-Funnel, and T₅₀ flow. After checking the SCGC mixture’s workability, it was placed without compaction into steel moulds (cubes, cylinders, and prisms), entirely occupying all gaps in the moulds with self-weight of mixtures. In addition, cubical samples were cast to analyse the compressive strength, cylinders to determine the indirect tensile strength, and beam samples were prepared to perform the flexural strength. Furthermore, concrete samples were cast to determine the permeability and sulfate attack test of SCGC.

Table 2.

Mix Proportions of GGBFS based SCGC.

Mix ID	Binders (%)		Quantity of ingredients utilized in SCGC (kg/m3)										SH
	GGBFS	MK	CBA	GGBFS	MK	CBA	FA	CA	SS	Extra Water	SP	Kg/m3	Mol/L
CM	100	0	0	400	0	0	840	980	143	54	8	57	10
MK5	95	5	0	380	20	0	840	980	143	54	12	57	10
MK10	90	10	0	360	40	0	840	980	143	54	16	57	10
MK15	85	15	0	340	60	0	840	980	143	54	20	57	10
MK20	80	20	0	320	80	0	840	980	143	54	24	57	10
CBA5	95	0	5	380	0	20	840	980	143	54	12	57	10
CBA10	90	0	10	360	0	40	840	980	143	54	16	57	10
CBA15	85	0	15	340	0	60	840	980	143	54	20	57	10
CBA20	80	0	20	320	0	80	840	980	143	54	24	57	10
2.5CBA2.5MK	95	2.5	2.5	380	10	10	840	980	143	54	12	57	10
5CBA5MK	90	5	5	360	20	20	840	980	143	54	16	57	10
7.5CBA7.5MK	85	7.5	7.5	340	30	30	840	980	143	54	20	57	10
10CBA10MK	80	10	10	320	40	40	840	980	143	54	24	57	10

Testing methods

Workability of SCGC

It was accomplished in terms of slump flow, V-Funnel, and T₅₀ flow blended with several extents of MK and CBA as a substitution for GGBFS separately and in combination were determined utilizing EFNARC’s European self-compacting concrete guidelines³⁶.

Mechanical properties of SCGC

The cubical samples made of SCGC inclusion of MK and CBA as a substitution for GGBFS were cast for compressive strength while the cylindrical samples of SCGC addition with MK and CBA as a substitution for GGBFS were tested for splitting tensile strength by obeying the BS EN 12390-3⁶⁷ and BS EN 12390-6⁶⁸ code procedure at 28 days consistently. In the same way, the prism specimens made of SCGC mixture introduction of MK and CBA as a substitution for GGBFS were prepared for analysing the flexural strength by obeying the BS EN 12390-5⁶⁹ code standard.

Durability of SCGC mixture

The permeability was performed on the concrete samples made of SCGC inclusion with MK and CBA as a substitution for GGBFS by using BS EN-12390-8⁷⁰ code procedure while the sulfate attack test was executed on the prism samples made of SCGC incorporating CBA and MK as a substitution for GGBFS and length change of samples was evaluated by using the ASTM C 1012-04⁷¹ code procedure at 180 days.

Results and discussions

Fresh properties

Slump flow

It is employed for assessing the workability of SCGC production accumulation with CBA and MK as a substitution for GGBFS alone and combined as shown in Fig. 1. According to the EFNARC³⁶, the slump flow is lying between 650 and 850 mm for all fresh mixes of SCGC. However, the highest slump flow is measured by 752 mm in the SCGC mixture which is made of GGBFS only while the lowest value is measured by 582 mm at 20% of GGBFS replaced with MK in the production of SCGC. It has been noted that as the amount of MK as a substitution for GGBFS grows in SCGC is decreased the slump flow. This finding is related to Arun et al.,⁷² where the workability of mix decreases but MK increases. Bheel et al.,² reported that the usage of FA replaced with MK rises in the production of SCGC reducing the slump flow. A similar kind of study was done by Guneyisi and Gesoglu⁷³. In the same way, the maximum slump flow is predicted to be 695 mm at 5% CBA, while the lowest value is projected to be 550 mm with 20% of CBA substituted with GGBFS in the production of SCGC. The addition of CBA to SCGC decreases the slump flow. This trend has been executed by Bheel et al.,³ that the slump flow is declined as content of WSA as a substitution for MK rises in SCGC. Furthermore, the CM has the highest slump flow of 752 mm, while the smallest value is noted by 578 mm at 10MK10CBA as a replacement for GGBFS in the production of SCGC. The combined amount of MK and CBA used to replace GGBFS rises in SCGC declined the slump flow. The greater specific surface area of MK and CBA which absorbs more water, leading to a drop in slump flow, and resulting in a stiffer mix. In real-world construction scenarios, a significant amount of MK may require the incorporation of superplasticizers to ensure the material remains flowable, as it tends to have a reduced slump flow. On the other hand, a higher amount of CBA can result in decreased workability, making it more difficult to place the material in complicated formwork or heavily reinforced portions. Although MK improves the material’s mechanical characteristics and durability by virtue of its pozzolanic reactivity, CBA may also make a significant contribution to the material’s strength and durability at an early stage. Local availability, affordability, and sustainability factors may also impact the selection between MK, CBA, and GGBFS. CBA, being an industrial by-product, is often more affordable and easily accessible, hence aiding waste management efforts as well as decreasing the carbon footprint associated with concrete manufacturing. Conversely, because MK is a manufactured material, it has a higher level of embodied energy. Thus, in order to meet the demands of structural purposes that require exceptional strength and durability, it may be more advantageous to use a mixture with a higher percentage of GGBFS and MK, even if it results in lower workability. On the other hand, for non-structural purposes where workability is of utmost importance, it is essential to carefully control the quantities of CBA. Ensuring a proper balance of these elements is crucial for optimising the mix design of SCGC to meet particular building requirements. This assertion is connected to Guneyisi and Gesoglu⁷³, who discovered that slump flow reduces while MK rises in the SCC. Bheel et al.,³ conveyed that the accumulation of WSA and Millet husk ash (MHA) as a substitution for MK increases in the production of SCGC is reduced the slump flow. A similar kind of trend was perceived by Memon et al.,⁷⁴, and Bheel et al.,².

Fig. 1.

Slump flow of GGBFS-Based SCGC.

V-funnel test

It was executed to measure the workability of SCGC addition with MK and CBA as a substitute for GGBFS alone and combined in the production of SCGC as revealed in Fig. 2. The flow time of the fresh mix is documented during the experiments. Based on the experimental outcome, the V-funnel flow time is lying in ranges from 8.40 to 13.10 s when various amounts of MK and CBA are used to substitute GGBFS separately and combined in the SCGC combination. Except for MK15, MK20, CBA20, 7.5CBA7.5MK, and 10CBA10MK, the V-funnel flow time of the remaining nine SCGC combinations accomplishes the requirement for a desirable V-funnel flow time. The minimal V-funnel flow duration is 8.40 s for SCGC without the accumulation of CBA and MK as substitutes for GGBFS alone and jointly. As shown in Fig. 2, the V-funnel flow time is improved as the content of MK and CBA as a substitute for GGBFS separate and together rises, which cause the decrease in the flow of SCGC. This decrease in V-funnel flow time is owing to the extremely small particle size of MK and CBA, which increases their specific surface area relative to GGBFS. The presence of varying amounts of MK and CBA in the V-Funnel test for SCGC has a notable impact on the flow time. This suggests that the viscosity of the mixture and the possibility of segregation are altered. Greater concentrations of MK, as a consequence of its elevated reactivity and small particle size, often lead to longer flow times, suggesting heightened viscosity and a more viscous mixture. The heightened viscosity contributes to a decrease in the likelihood of segregation, guaranteeing a more consistent and stable mixture. Conversely, adding more CBA often leads to longer flow times. This is because the irregular shape and larger size of CBA particles increase internal friction, as well as the viscosity of the mixture. Nevertheless, an overabundance of CBA may also result in workability problems and possible segregation if it is not well balanced with other constituents. In general, adding both MK and CBA in the right quantities may improve SCGC’s stability by increasing viscosity and decreasing segregation. However, it is important to carefully analyse the amounts of these additives to ensure that the ideal flow features and workability of the concrete mix are maintained. Bheel et al.,³ presented that the V-funnel flow time of fresh mixture is decreased as the quantity of WSA and MHA as a substitution for MK separate and together rises in the production of SCGC. A comparable investigation was conducted by Memon et al.⁷⁵, Dhiyaneshwaran et al.⁷⁶, and Bheel et al.².

Fig. 2.

V-Funnel Flow of GGBFS-Based SCGC.

T₅₀ flow

Figure 3 illustrates the T₅₀ flow assessment, which is executed to check the flow of mixture of SCGC production qualities, comprising the use of MK and CBA as a substitute for GGBFS. For this test, the slump flow time should be attained to 50 cm in diameter when performing the fresh mix of SCGC. It has been observed from the experimental outcome that the slump flow time of fresh mixture is lying in the ranges from 2.8 to 6.15 s when MK and CBA are used in varying quantities to replace GGBFS separately and combined in SCGC mixture. According to EFNARC³⁶, the slump flow time is noted within the acceptable range (2–5 s) for all SCGC combinations except MK15, MK20, CBA15, CBA20, and 10CBA10MK. The minimum slump flow time of 2.80 s is observed at the SCGC mixture without the accumulation of CBA and MK as substitutes for GGBFS separately and in combination. As illustrated in Fig. 3, the slump flow time of SCGC improves as the proportion of MK and CBA used to replace GGBFS alone and jointly increases in the production of SCGC. This increase in the T₅₀ flow time is attributable to the increased paste volume achieved by using MK and CBA in place of GGBFS alone and together in SCGC. Moreover, the T50 time characteristic provides information about the concrete mixture’s viscosity and flow properties. A shorter T50 time implies lower viscosity and improved flowability, while a longer T50 time predicts greater viscosity and decreased workability. Using MK and CBA separately as GGBFS substitutes tends to increase the T50 flow time. MK, owing to its pronounced reactivity and the presence of small particles, substantially elevates the viscosity and diminishes the flowability. Likewise, CBA, because of its irregular and bigger particles, causes a rise in internal friction, resulting in increased viscosity and longer T50 times. The combined influence of MK and CBA on the T50 flow time is cumulative, leading to a mixture with increased viscosity and extended T50 durations. Using this combined substitution may improve the stability of the concrete mix while reducing the danger of segregation. However, it is important to carefully balance the proportions to ensure that the workability of the mix remains sufficient. Therefore, whereas each material alone enhances the T50 flow time, their combined use necessitates optimisation to maximise their advantages while balancing the compromise in flowability. Bheel et al.,² itemized that the T₅₀ flow time is amplified by inclusion the content of Groundnut shell ash and MK by the weight of FA alone and together rises in the production of SCGC. The analogous trend was perceived in Memon et al.⁷⁵, Bheel et al.,³, Keerio et al.,⁶⁵, and Dhiyaneshwaran et al.⁷⁶.

Fig. 3.

T₅₀ Flow of GGBFS-Based SCGC.

Mechanical properties

Compressive strength (CS)

The CS of SCGC containing CBA and MK as substitutes for GGBFS is measured in Fig. 4. Nevertheless, the compressive strength is 57.40 MPa, 59.86 MPa, and 57.10 MPa at 5%, 10%, and 15% of GGBFS substituted with MK, respectively, which is more than CM, and thus the strength of SCGC mixture is reached. After 28 days, 52.50 MPa at 20% MK as a substitution for GGBFS is less than SCGC prepared with GGBFS as the only binding agent in the mixture. It has been detected that the accumulation of 5–15% of MK as a substitution for GGBFS in the production of SCGC is increased the CS, however, increasing the amount of MK as a GGBFS substitution ingredient in SCGC decreases the strength. Additionally, the CS is optimized at 56.80 MPa, 58.76 MPa, and 56.40 MPa when 5%, 10%, and 15% of GGBFS are substituted with CBA, correspondingly, while the minimum strength is seen at 51.25 MPa when 20% of GGBFS is substituted with CBA after 28 days. The outcomes indicate that using CBA as a substitute for GGBFS augmented the CS of the SCGC combination up to a point where it begins to drop. This aspect is correlated to Memon et al.,⁷⁵, who demonstrated an increase in the crushing strength of SCGC by substituting SF for FA up to 15% in the combination at 28 days. Furthermore, the highest CS is 57.25 MPa, 59.40 MPa, and 56.98 MPa at 2.5CBA2.5MK, 5CBA5MK, and 7.5CBA7.5MK, consecutively, while the minimum CS is 52.20 MPa at 10CBA10MK as a substitution for GGBFS in the production of SCGC after 28 days. It has been perceived that the use of CBA and MK up to 15% as a substitution for GGBFS combined in the production of SCGC is increased the CS. This enhancement in strength is related to the extremely small size of CBA and MK particles in comparison to GGBFS, which makes the pores smaller and the particles more densely packed, producing the concrete denser and more stable. Bheel et al.,³ described that the 10% of MK substituted with WSA and MHA combined in SCGC augmented the CS. A similar observation was done by Keerio et al.⁶⁵ and Bheel et al.,².

Fig. 4.

CS of GGBFS-based SCGC.

Splitting tensile strength (STS)

It was achieved on SCGC specimens’ addition with MK and CBA as a substitution for GGBFS alone and jointly after 28 days as directed in Fig. 5. The highest STS is noted by 5.75 MPa and 5.63 MPa when using 10% of MK and CBA as a substitution for GGBFS, while the least strength is measured by 5.18 MPa and 5.0 MPa at 20% of CBA and CBA as a substitution for GGBFS in the production of SCGC after 28 days consistently. The experimental result has shown that the STS is enhanced when the usage of CBA and MK up to 15% as a substitution for GGBFS alone and further addition of CBA and MK in the SCGC, the STS gets declined. Bheel et al.,² assured that the accumulation of 10% FA substituted with MK in the production of SCGC boosted the STS at 28 days. Memon et al.,⁷⁵ notified that the usage of 10% of FA replaced with SF in the production of SCGC augmented the STS after 28 days. Moreover, the greatest STS is estimated by 5.68 MPa at 5CBA5MK and the lowest strength is 5.12 MPa at 10CBA10MK as a substitution for GGBFS in the production of SCGC on 28 days correspondingly. From Fig. 5, it has been shown that the consumption of MK and CBA combined up to 15% as a substitution for GGBFS boosted the STS of SCGC. This enhancement in strength is associated with the MK and CBA possessing a higher specific surface area as compared to GGBFS in the production of SCGC. A analogous study was performed by Bheel et al.², where the STS of SCGC is augmented when the consumption of MK and GSA combined up to 10% as a substitution for FA after 28 days. Similar kinds of findings were observed in^3,65,72,76.

Fig. 5.

STS of GGBFS-based SCGC.

Flexural strength (FS)

It was performed on the prism samples made of SCGC introduction of MK and CBA as a substitution for GGBFS alone and together for 28 days as exhibited in Fig. 6. The optimal flexural strength is measured by 6.20 MPa at 10% of MK while the least flexural strength is recorded by 5.50 MPa at 20% of MK as a substitution for GGBFS in the production of SCGC on 28 days, consistently. The result has shown that the FS is enhanced when the usage of MK as a substitution for GGBFS alone up to 15% and more accumulation of MK in the SCGC, the strength starts declined. In the same way, the optimum flexural strength is computed by 6.0 MPa when the consumption of 10% CBA as a substitution for GGBFS and the minimum value is seen by 5.35 MPa at 20% GGBFS replacement with CBA after 28 days consistently. The assessment results demonstrated that the using of CBA as a substitution for GGBFS to 15% in the production of SCGC increased the flexural strength, and after 15% of CBA, it begins to deteriorate. The finding of the experimental was provided by Bheel et al.², wherever the usage of 10% FA as substituted with MK in the production of SCGC augmented the flexural strength for 28 days. Bheel et al.³ declared that the accumulation of 10% WSA as a substitution for MK in the production of SCGC enhanced flexural strength at 28 days. A comparable finding was done by Memon et al.,⁷⁵. Furthermore, the greatest flexural strength is estimated by 6.12 MPa at 5CBA5MK while the minimum strength is acquired by 5.42 MPa at 10CBA10MK as a substitution for GGBFS in SCGC on 28 days consistently. The outcome has been discovered that using CBA and MK as combination material substituted with GGBFS up to 15% in the production of SCGC improved the FS. The increased FS is related with the very fineness of CBA and MK particles, which plugs the voids produced by water in the production of the SCGC matrix. The usage of MK and CBA as the finer particles decrease the microcracks in SCGC and generate a significantly denser SCGC that results in improving the strength of the concrete⁷⁷. Bheel et al.² described that the combined consumption of 10% FA substituted with GSA and MK augmented the FS of SCGC at 28 days. The comparable findings were carried out by^{3,65,72,73,78}.

Fig. 6.

FS of GGBFS-based SCGC.

Durability of SCGC

Permeability of SCGC

It was executed on the SCGC samples incorporating CBA and MK as a substitution for GGBFS alone and jointly at 28 days as indicated in Fig. 7. The permeability of SCGC was assessed by 20 mm, 17 mm, 14 mm, and 10 mm at 5%, 10%, 15%, and 20% of GGBFS substituted with MK at 28 days, correspondingly. The permeability is decreased when the amount of MK as a substitution for GGBFS grows in the production of SCGC. According to Keerio et al.,⁶⁴, the permeability is lowered by up to 15% when MK is used as a substitution for cement for 28 days. The comparable findings were provided by Bheel et al.⁷⁹. Furthermore, the permeability of SCGC is calculated by 22 mm, 19 mm, 16 mm, and 12 mm at 5%, 10%, 15%, and 20% of GGBFS substituted by CBA at 28 days, correspondingly. As the extent of CBA as a substitution for GGBFS in SCGC rises, the permeability of the SCGC decreases. Guneyisi et al.⁸⁰ carried out a similar study in which the permeability decreased as the MK extent of concrete increases. Furthermore, the permeability of the SCGC is calculated by 21 mm, 18 mm, 15 mm, and 11 mm at 2.5CBA2.5MK, 5CBA5MK, 7.5CBA7.5MK, and 10CBA10MK as a substitute for GGBFS after 28 days is smaller as compared to SCGC without the presence of CBA and MK. According to Fig. 7, the permeability is decreased, while the content of CBA and MK combined as a substitution for GGBFS rises in the production of SCGC. The water penetration depth in SCGC specimens with MK and CBA substitutes for GGBFS, as determined by BS EN 12390-8, offers important information on the long-term durability and permeability of the concrete. A lower water penetration depth usually suggests a more compact and less porous concrete structure, which is directly linked to improved durability. The replacement of GGBFS with MK and CBA might have an impact on the microstructural properties of the concrete. Because of its small particle size and strong reactivity, MK may fill empty spaces and aid in the creation of more gel-like substances, thereby decreasing the concrete’s porosity and permeability. When appropriately treated, CBA may also function as a filler substance, resulting in a further reduction in pore volume and enhancing the overall density of the matrix. These changes in the microstructure result in a concrete that is more resistant to water penetration, which is critical for ensuring its long-term durability. Decreased water infiltration results in a lower vulnerability to phenomena such as freeze–thaw cycles, chloride ion intrusion, and other types of chemical attack. These processes may gradually weaken the structural integrity of the concrete. Hence, the water penetration test’s depth, as specified in BS EN 12390-8⁷⁰, acts as a vital indicator of the concrete’s performance in hostile surroundings. The correlation between total permeability and the ability of the SCGC with MK and CBA substitutes to survive environmental stressors and preserve its structural integrity over a long period of time is evident. Existing research supports this association, highlighting the significance of low permeability in the construction of long-lasting concrete buildings. Bheel et al.⁷⁹ did related work in which the permeability is reduced when the doses of combined GGBFS and MK by the mass of cement rise in concrete. A related finding was agreed by Bheel et al.² where the usage of FA substituted with GSA and MK increases alone and jointly in the production of SCGC is reduced the permeability at 28 days.

Fig. 7.

Permeability of GGBFS-based SCGC.

Sulfate attack test

Sulfate attack is one of the main aspects of durability that can be modified by the independent and combined substitution of GGBFS with MK and CBA as shown in Fig. 8. Sulfates are typically found in water, especially when the soil contains a high amount of clay. The abundance of sulfates in rainwater is likely owing to air pollution and the presence of sulfates in seawater. This sulfate attack test was conducted on prism samples of SCGC combined with MK and CBA as a substitution for GGBFS, and the difference between the beginning and final values is shown in Fig. 8. However, the change in length of SCGC samples with 5%, 10%, 15%, and 20% substitution of GGBFS with MK and CBA separate and combine owing to sulfate attack is within the acceptable limit (i.e., 0.1% maximum length change) as defined by ASTM for pozzolanic material. Nevertheless, a 15% substitution of GGBFS with MK resulted in the smallest change in length. Sample break-down was not observed in this investigation, which could be attributed to the dense matrix of SCGC, which has a low water-to-binder ratio and a lower specific surface area of metakaolin. Khatib and Wild⁸¹ informed that the integration of MK in composite reduced sulfate attack growth. The concrete comprising 10% or additional MK displayed no expansion, whereas the concrete comprising 0% or 5% MK did not start growing until the age of 180 days⁸¹. Similarly, at 180 days, a 15% replacement of GGBFS with CBA resulted in the smallest change in length. Specimen breakdown was not observed in this investigation, which could be attributed to the dense matrix of SCGC with a low water-to-binder ratio. According to Sajjad Ali Mangi et al.⁸². The efficacy of concrete including CBA versus concrete with no CBA under 5% sodium sulfate (Na₂SO₄) exposure is similar up to 90 days. For short-term exposure, there is no discernible effect of Na₂SO₄ solution on concrete with and without CBA⁸². Furthermore, at 180 days, the MK7.5CBA7.5 combined replacement of GGBFS with MK and CBA resulted in the smallest decrease in length. Courard et al.⁸³ conducted a similar type of study on the sulfate resistance of MK in matrix by using ASTM C 1012. Within a short period of time, the PC mortar prisms exhibit expansion, and the accumulation of 10% MK leads to a 3.7% decrease in expansion compared to CM⁸³. According to Kasemchaisiri and Tangtermsirikul⁸⁴, the highest expansion of CM was SCC, followed by SCC-BA 10%, SCC-BA 20%, and SCC-BA 30%, correspondingly. The dense matrix greatly reduced SO4^2- ion migration into the concrete, indicating that increasing bottom ash concentration results in greater sodium sulfate resistance. The pozzolanic reaction of bottom ash reduces this expansion in concrete⁸⁵. Despite their higher porosity, the expansion of SCC with bottom ash (BA) reduces as the BA concentration rises. This suggests that sodium sulfate resistance can be attained by the pozzolanic reaction of BA, which takes precedence over porosity rise. Sulfate resistance improvement is helpful for SCC in sulfate-exposing structures such as underground structures, drilled piles, and undersea structures, among others⁸⁵. Different research^81–85 discovered nearly the same trend of sulfate resistance. The ratios of MK and CBA used as GGBFS replacements directly correlate with the extent of length change in SCGC specimens subjected to sulfate attack. Metakaolin, due to its high reactivity and small particle size, aids in the production of a compact and stable geopolymer structure. This structure effectively prevents the penetration of sulphate ions and minimises the development of expansive substances such as ettringite and gypsum. Consequently, an increase in the fraction of MK often results in a decrease in length changes, indicating enhanced resistance to sulfate attack. Moreover, CBA may serve as a substance that fills in gaps, enhancing the structure of the pores and reducing the capacity of sulfate to pass through. The efficacy of CBA depends on its proportions and particle dimensions. It is essential to achieve an appropriate balance between MK and CBA, since an excessive quantity of either might result in negative consequences, such as decreased workability or inadequate geopolymerization. The relationship between material selection and better durability is shown by emphasising the significance of proportioning MK and CBA to produce a geopolymer matrix that is both dense and impermeable. By carefully adjusting the mix’s ingredients, adding the right amount of MK to make sure it has enough pozzolanic activity, and using CBA to increase the toughness, the resulting SCGC may show small changes in length when exposed to sulfate attack. This optimisation guarantees that the concrete retains its structural integrity and durability in situations with high levels of sulphates. Therefore, understanding the relationship between MK and CBA amounts and their impact on the duration change due to sulphate attack is crucial for the creation of long-lasting SCGC combinations.

Fig. 8.

Sulfate attack test of GGBFS-based SCGC at 180 days.

Sustainability assessment

Embodied carbon (EC)

In order to determine the embodied carbon dioxide emissions associated with GGBFS-based SCGC, an environmental impact assessment is conducted for all aspects pertaining to the combination of different GGBFS replacements with MK and CBA in the combination. Table 3 displays the EC levels for all GGBFS-based SCGC components, sourced from prior research. Equation (1)⁸⁶ is used to determine the EC for all SCGC based on GGBFS. Equation (1) represents the unit volume weight (kg/m³) and total EC for any GGBFS-based SCGC, denoted by the symbols Ee, CO_2e, Wi, and i. Moreover, the symbols Ei and CO_2i are used to denote the EC content of the physical constituents, as shown in Table 3.

$$C{O}_{2e}={\sum }_{(i=1)}^n(W_i\times C{O}_{2i}),$$ 1

Table 3.

EC of GGBFS based SCGC components.

Components	EC (kg ${\text{CO}}_2$/kg)	References
GGBFS	0.07	87,88
FA	0.0139	89
CA	0.0408	90
CBA	0.015	65,91
MK	0.33	92,93
SH	0.625	94
SS	0.445	95,96
SP	0.01	97
Water	0	92

The study assessed the environmental impact of utilizing GGBFS^98–100 based SCGC mixed with MK and CBA as substitutes for GGBFS by evaluating the EC of each mix. Figure 9 presents the results and compares them to the reference mix. The Fig. 9 illustrates the significant carbon footprint of sodium silicate compared to MK, CBA, GGBFS, CA, FA, SP, SH, and water. Additionally, the embodied carbon of SCGC blended with various GGBFS substitutes, both separate and combined, is shown in Fig. 9. Furthermore, the figure indicates that sodium silicate emits the high quantity carbon, followed by SH, coarse and fine materials. However, the graph does not show the impact of CBA, SP, and water in the SCGC, suggesting that their contribution to carbon intensity is minimal. Besides, the EC of MK is greater than GGBFS and CBA. Equation (1) was useful to find out the total of carbon of all SCGC built on the information revealed in Fig. 9. Figure 9 reveals that the EC of SCGC is recorded by 2.93%, 5.86%, 8.78%, and 11.71% at MK5, MK10, MK15, and MK20 which is greater than that of reference mixture of GGBFS based SCGC at 28 days separately. It has been perceived that the embodied carbon of SCGC is getting higher with accumulation of MK as substitution for GGBFS in SCGC. This enhancement in EC is associated to the EC factor of MK is more than that of GGBFS. Moreover, the EC of SCGC combined with CBA5, CBA10, CBA15, and CBA20 as substitution for GGBFS is calculated by 0.59%, 1.18%, 1.78%, and 2.37% which is lesser than CM of GGBFS based SCGC at 28 days respectively. This reduction in EC of GGBFS based SCGC is owing to the EC factor of CBA is lesser as associated to the GGBFS. Furthermore, the EC of GGBFS based SCGC is noted by 1.17%, 2.34%, 3.50%, and 4.67% at 2.5MK2.5CBA, 5MK5CBA, 7.5MK7.7CBA, and 10MK10CBA which is higher than that of reference mix of GGBFS based SCGC at 28 days consistently. From the findings, it is noted that the EC of GGBFS based SCGC is reducing with addition of CBA alone and the addition of MK is increased the EC of SCGC alone and together with CBA.

Fig. 9.

Embodied Carbon of GGBFS based SCGC.

RSM modelling and optimization

The development of response surface models and ANOVA

RSM is employed for the construction and development of response surface models, which are then evaluated applying ANOVA. For RSM modelling and optimization, the fresh, mechanical, and durability features of SCGC were considered with the inclusion of 0 to 20% MK and CBA alone or together as a GGBFS substitute. In addition, quadratic models were deemed more appropriate for the slump flow test, V-funnel flow test, and T50 flow test, CS, STS, FS, permeability, and sulfate attack test. Moreover, each of these responses is contained in Eqs. (2)–(9). Equations described in relationships of coded constituents may be utilized to predict the impact of changing parameter values. By default, the greatest levels of the constituents are represented as + 1 and the lowest values as − 1. Using the coefficients of the factors, the coded equations may be used to assess the variables' relative significance^98–100. A and B are variables of input (MK and CBA). The ANOVA findings are shown in Table 4.

Table 4.

ANOVA outcomes.

Response	Source	Sum of Squares	Df	Mean Square	F-Value	p-value > F	Significance
Slump flow test	Model	59,496.78	5	11,899.36	423.23	< 0.0001	Yes
	A-MK	5285.39	1	5285.39	187.99	< 0.0001	Yes
	B-CBA	6705.01	1	6705.01	238.48	< 0.0001	Yes
	AB	56.34	1	56.34	2.00	0.1873	No
	A2	840.05	1	840.05	29.88	0.0003	Yes
	B2	20.80	1	20.80	0.74	0.4098	No
	Residual	281.16	10	28.12
	Lack of Fit	281.16	5	56.23
	Pure Error	0.000	5	0.000
	Cor Total	59,777.94	15
V-funnel flow test	Model	30.37	5	6.07	1299.45	< 0.0001	Yes
	A-MK	1.67	1	1.67	357.73	< 0.0001	Yes
	B-CBA	1.17	1	1.17	249.30	< 0.0001	Yes
	AB	0.19	1	0.19	41.20	< 0.0001	Yes
	A2	0.38	1	0.38	80.36	< 0.0001	Yes
	B2	0.18	1	0.18	38.45	0.0001	Yes
	Residual	0.047	10	4.674E−003
	Lack of Fit	0.047	5	9.348E−003
	Pure Error	0.000	5	0.000
	Cor Total	30.41	15
T50 flow test	Model	18.96	5	3.79	264.55	< 0.0001	Yes
	A-MK	1.05	1	1.05	73.29	< 0.0001	Yes
	B-CBA	1.07	1	1.07	75.01	< 0.0001	Yes
	AB	0.056	1	0.056	3.89	0.0770	No
	A2	0.031	1	0.031	2.19	0.1693	No
	B2	0.090	1	0.090	6.31	0.0308	Yes
	Residual	0.14	10	0.014
	Lack of Fit	0.11	5	0.022	3.59	0.0937	No
	Pure Error	0.031	5	6.250E−003
	Cor Total	19.10	15
Compressive strength	Model	133.64	5	26.73	82.98	< 0.0001	Yes
	A-MK	84.80	1	84.80	263.30	< 0.0001	Yes
	B-CBA	88.55	1	88.55	274.92	< 0.0001	Yes
	AB	53.55	1	53.55	166.25	< 0.0001	No
	A2	42.00	1	42.00	130.40	< 0.0001	Yes
	B2	39.41	1	39.41	122.35	< 0.0001	Yes
	Residual	3.22	10	0.32
	Lack of Fit	3.22	5	0.64
	Pure Error	0.000	5	0.000
	Cor Total	136.86	15
Splitting tensile strength	Model	0.75	5	0.15	27.43	< 0.0001	Yes
	A-MK	0.43	1	0.43	77.87	< 0.0001	Yes
	B-CBA	0.49	1	0.49	90.63	< 0.0001	Yes
	AB	0.28	1	0.28	52.17	< 0.0001	No
	A2	0.25	1	0.25	45.33	< 0.0001	Yes
	B2	0.22	1	0.22	39.38	< 0.0001	Yes
	Residual	0.055	10	5.460E−003
	Lack of Fit	0.055	5	0.011
	Pure Error	0.000	5	0.000
	Cor Total	0.80	15
Flexural strength	Model	1.15	5	0.23	91.21	< 0.0001	Yes
	A-MK	0.74	1	0.74	292.49	< 0.0001	Yes
	B-CBA	0.80	1	0.80	314.87	< 0.0001	Yes
	AB	0.49	1	0.49	193.64	< 0.0001	No
	A2	0.42	1	0.42	167.25	< 0.0001	Yes
	B2	0.30	1	0.30	117.68	< 0.0001	Yes
	Residual	0.025	10	2.527E−003
	Lack of Fit	0.025	5	5.054E−003
	Pure Error	0.000	5	0.000
	Cor Total	1.18	15
Water penetration depth	Model	311.23	5	62.25	1204.32	< 0.0001	Yes
	A-MK	40.38	1	40.38	781.18	< 0.0001	Yes
	B-CBA	29.85	1	29.85	577.44	< 0.0001	Yes
	AB	0.83	1	0.83	16.09	0.0025	Yes
	A2	1.574E-003	1	1.574E−003	0.030	0.8649	No
	B2	2.08	1	2.08	40.30	< 0.0001	Yes
	Residual	0.52	10	0.052
	Lack of Fit	0.52	5	0.10
	Pure Error	0.000	5	0.000
	Cor Total	311.75	15
Sulfate attack test	Model	3.579E-003	5	7.159E−004	15.89	0.0002	Yes
	A-MK	2.893E-003	1	2.893E−003	64.22	< 0.0001	Yes
	B-CBA	2.771E-003	1	2.771E−003	61.52	< 0.0001	Yes
	AB	2.289E-003	1	2.289E−003	50.81	< 0.0001	Yes
	A2	4.884E-004	1	4.884E−004	10.84	0.0081	Yes
	B2	1.766E-003	1	1.766E−003	39.19	< 0.0001	Yes
	Residual	4.505E-004	10	4.505E−005
	Lack of Fit	4.505E-004	5	9.009E−005
	Pure Error	0.000	5	0.000
	Cor Total	4.030E-003	15

In ANOVA, the significance level was established at 95%, meaning that any prototype component with a possibility lesser than 5% is measured statistically important. All of the produced approaches are statistically significant because their probabilities are below 0.05. The slump flow test model’s model terms A, B, AB, A², and B² have statistical significance. A, B, AB, A², and B² are also key model terms for the V-funnel flow test, T50 flow test, CS, STS, FS, permeability, and sulfate attack test. The coefficient of determination, often referred to as R², is a crucial performance metric. It quantifies the model’s goodness of fit to the observed data and can range between 0 and 1, or be presented as a percentage from 0 to 100. A higher R² value indicates a better fit of the model to the empirical data, while a lower value suggests a weaker fit. Table 5 presents the R² value along with other evaluation factors for the models. R² values are observed by 99.53%, 99.85%, 99.25%, 97.65%, 93.20%, 97.85%, 99.83%, and 88.82% for slump flow test, V-funnel flow test, T50 flow test, CS, STS, FS, permeability, and sulfate attack test. Furthermore, "Adeq. Precision" computes the signal-to-noise ratio. A ratio greater than four is preferred. Precision values for the slump flow test, V-funnel flow test, T50 flow test, CS, STS, FS, permeability and sulfate attack test is 60.19, 112.03, 47.28, 22.78, 14.82, 25.69, 98.78, and 12.18, respectively, according to Table 5. These results show that the models are capable of precisely anticipating responses.

$$\mathbf{S}\mathbf{F}\mathbf{T}=+577.85-94.30\times \mathbf{A}-108.97\times \mathbf{B}-11.24\times \mathbf{A}\mathbf{B}-25.83{\mathbf{A}}^2+4.02\times {\mathbf{B}}^2$$ 2

$$\mathbf{V}\mathbf{F}\mathbf{T}=+13.11+1.68\times \mathbf{A}+1.44\times \mathbf{B}-0.66\times \mathbf{A}\mathbf{B}-0.55\times {\mathbf{A}}^2-0.37\times {\mathbf{B}}^2$$ 3

$$\mathbf{T}\mathbf{F}\mathbf{T}=+5.41+1.33\times \mathbf{A}+1.38\times \mathbf{B}-0.35\times \mathbf{A}\mathbf{B}+0.16{\mathbf{A}}^2+0.26\times {\mathbf{B}}^2$$ 4

$$\mathbf{C}\mathbf{S}=+52.26-11.94\times \mathbf{A}-12.52\times \mathbf{B}-10.96\times \mathbf{A}\mathbf{B}-5.78{\mathbf{A}}^2-5.53{\mathbf{B}}^2$$ 5

$$\mathbf{S}\mathbf{T}\mathbf{S}=+5.13-0.85\times \mathbf{A}-0.94\times \mathbf{B}-0.80\times \mathbf{A}\mathbf{B}-0.44{\times \mathbf{A}}^2-0.41\times {\mathbf{B}}^2$$ 6

$$\mathbf{F}\mathbf{S}=+5.43-1.12\times \mathbf{A}-1.19\times \mathbf{B}-1.05\times \mathbf{A}\mathbf{B}-0.58{\times \mathbf{A}}^2-0.48\times {\mathbf{B}}^2$$ 7

$$\mathbf{W}\mathbf{P}\mathbf{D}=+10.99-8.24\times \mathbf{A}-7.27\times \mathbf{B}-1.37\times \mathbf{A}\mathbf{B}-0.035{\times \mathbf{A}}^2-1.27\times {\mathbf{B}}^2$$ 8

$$\mathbf{S}\mathbf{A}\mathbf{T}=+0.11+0.070\times \mathbf{A}+0.070\times \mathbf{B}+0.072\times \mathbf{A}\mathbf{B}+0.020{\times \mathbf{A}}^2+0.037\times {\mathbf{B}}^2$$ 9

Table 5.

Constraints for model verification.

Model Validation Constraints	SFT	VFT	TFT	CS	STS	FS	WPD	SAT
Std. Dev	5.30	0.068	0.12	0.57	0.074	0.050	0.23	6.712E−003
Mean	626.56	11.76	4.97	55.11	5.32	5.70	15.13	0.086
C.V.%	0.85	0.58	2.41	1.03	1.39	0.88	1.50	7.79
PRESS	908.91	0.11	0.32	16.15	0.15	0.067	1.57	1.713E−003
-2 Log Likelihood	91.27	-47.97	-30.04	19.76	-45.48	-57.81	-9.52	−122.24
R-Squared	0.9953	0.9985	0.9925	0.9765	0.9320	0.9785	0.9983	0.8882
Adj R-Squared	0.9929	0.9977	0.9887	0.9647	0.8980	0.9678	0.9975	0.8323
Pred R-Squared	0.9848	0.9963	0.9834	0.8820	0.8142	0.9433	0.9949	0.5748
Adeq Precision	60.195	112.029	47.281	22.786	14.824	25.693	98.784	12.182

Here, SFT, VFT, TFT, CS, STS, FS, WPD, and SAT are Slump Flow Test, V-Funnel Flow Test, T₅₀ Flow Test, Compressive Strength, Splitting Tensile Strength, Flexural Strength, Water Penetration Depth, and Sulphate Attack Test respectively.

The “Residual against Normal” diagram and the “Predicted against Actual” diagram, as shown in Figs. 10, 11, 12, 13, 14, 15, 16 and 17 for the eight parameters, are utilized as effective approach for more rapidly analyzing the condition and applicability of generated response models (SFT, VFT, TFT, CS, STS, FS, WPD, and SAT). All charts display the linearity of the data sets along the fit line to illustrate the accuracy of the created approaches. The arrangement of data facts on residual normal plots suggests that the error components have a normal distribution, which is desired. If 95% of the points lie between − 2 and + 2, which is true for all approaches^101,102, residuals are often dispersed.

Fig. 10.

Slump flow test plots.

Fig. 11.

V funnel flow test plots.

Fig. 12.

TFT test plots.

Fig. 13.

Compressive strength test plots.

Fig. 14.

Splitting Tensile strength test plots.

Fig. 15.

Flexural strength test plots.

Fig. 16.

Water Penetration Depth test plots.

Fig. 17.

Sulphate attack test plots.

The effect of the relationships between the output responses and input factors is shown using 2D diagram and 3 D graphs. Figures 18, 19, 20, 21, 22, 23, 24 and 25 depict response surface plots for the SFT model, VFT model, TFT model, CS model, STS model, FS model, WPD model, and SAT model. In this case, both 2 D plots and 3 D graphs show how the MK and CBA interact. The graphs' color coding illustrates the size of the response as well as the various input feature values under consideration. Figure 18a and b show the 2D and 3D plots for the slump flow test. The data showed that the control combination has a high intensity of slump at up to 15.33% MK and 4.71% CBA as GGBFS replacement material. Moreover, the slump flow of geopolymer GGBFS-based concrete diminishes when more MK and CBA alone and merge in matrix. The surface responses for permeability, and sulfate attack test are all similar. The V-funnel flow test of geopolymer GGBFS-based concrete rises when the proportions of MK and CBA in matrix. The T50 flow test produced similar surface reactions. Furthermore, Fig. 20a and b show the 2D diagram and 3D graphs for CS, respectively. According to the graphs, a high concentration of CS has been stated as a GGBFS replacement material, up to 15.33% of MK and 4.71% of CBA. Furthermore, by including up to 15.33% MK and 4.71% CBA in the concrete, the strength value is substantially increased. This is owing to the void-filling and densification properties of MK and CBA as a pozzolanic material combination. The similar behaviour of response surface plots was observed for STS, and FS respectively.

Fig. 18.

SFT plots.

Fig. 19.

VFT plots.

Fig. 20.

TFT plots.

Fig. 21.

Compressive strength plots.

Fig. 22.

Splitting Tensile strength plots.

Fig. 23.

Flexural strength plots.

Fig. 24.

Water Penetration depth test plots.

Fig. 25.

Sulphate attack test plots.

Optimization

It is employed to analyse the ideal values of the independent variables that result in the highest possible output level. This is performed by creating quality objectives (input and output characteristics) with varied conditions and degrees of priority to achieve the intended purpose. The optimization is assessed via the desirability (0 ≤ dj ≤ 1). The desirability rate is near to one which indicates the superior results^103,104.

Table 6 illustrates the objectives and prerequisites for optimizing this scenario. However, the usage of MK and CBA as a substitute for GGBFS has been restricted to a range of 0 to 20% separate and combined, so that the system can select the appropriate amount required to fulfil the stated purpose. According to the optimization findings, the highest values for SFT, VFT, TFT, CS, STS, FS, WPD, and SAT could be obtained by 639.15 mm, 12.38 s, 5.23 s, 57.08 MPa, 5.50 MPa, 5.93 MPa, 13.31 mm, and 0.081%, respectively. Given the nature of the considerable response value variability, the optimization’s desirability is assessed to be 69.10%. Figures 26 illustrate 3D diagram for desirability.

Table 6.

Optimization goals and outcomes.

Factors		Input factors				Responses (Output Factors)
		MK (%)	CBA (%)	SFT (mm)	VFT (Sec)	TFT (Sec)	CS (MPa)	STS (MPa)	FS (MPa)	WPD (mm)	SAT (%)
Value	Min	0	0	550	8.40	2.80	51.25	5.0	5.35	10	0.059
	Max	20	20	752	13.10	6.20	59.86	5.75	6.20	24	0.11
Goal		Range	Range	Min	Max	Max	Max	Max	Max	Min	Min
Optimization Results		15.33	4.71	639.15	12.38	5.23	57.08	5.50	5.93	13.31	0.081
Desirability					0.691 (69%)

Fig. 26.

3D Plot of desirability.

Conclusions

This study examined the fresh, mechanical properties and durability of SCGC integrating the CBA and MK as a substitution for GGBFS alone and jointly. The following key points are drawn from this investigation:

• The slump flow is decreased as the amount of GGBFS is replaced with MK and CBA alone and jointly rises in the production of SCGC grows. This drop-in slump flow is associated with the very fineness of CBA and MK particles and the specific surface area of GGBFS is lower as compared to MK and CBA.

• The V-funnel flow time is boosted when the consumption of CBA and MK as a substitution for GGBFS alone and together rises in the production of SCGC that result in dropping the fluidity of SCGC.

• The slump flow time is enhanced as the consumption of CBA and MK as a substitution for GGBFS alone and combined increases in the production of SCGC. This increase in the T₅₀ flow time is attributable to the increased paste volume achieved by using MK and CBA in place of GGBFS alone and together in the production of SCGC.

• The peak CS is recorded by 59.86 MPa, 58.76 MPa, and 59.40 MPa at 10% of MK, 10% of CBA, and 5CBA5MK, and the lowest CS is noted by 52.50 MPa, 51.25 MPa and 52.20 MPa at 20% of MK, 20% of CBA and 10CBA10MK as a substitution for GGBFS in SCGC after 28 days, consistently. It has been perceived that the use of CBA and MK up to 15% as a substitution for GGBFS combined in SCGC is augmented the CS.

• The maximum STS is recorded by 5.75 MPa, 5.63 MPa, and 5.68 MPa at 10% of MK, 10% of CBA, and 5CBA5MK, and the lowest strength is found by 5.18 MPa, 5.0 MPa, and 5.12 MPa at 20% of MK, 20% of CBA and 10CBA10MK as a substitution for GGBFS in SCGC after 28 days, consistently. Moreover, the use of CBA and MK up to 15% as a substitution for GGBFS combined in SCGC is increased the STS.

• The maximum FS is recorded by 6.20 MPa, 6.0 MPa, and 6.12 MPa at 10% of MK, 10% of CBA, and 5CBA5MK, and the lowest FS is noted by 5.50 MPa, 5.35 MPa and 5.42 MPa at 20% of MK, 20% of CBA and 10CBA10MK as a substitution for GGBFS in the production of SCGC on 28 days, consistently. It has been perceived that the use of CBA and MK up to 15% as a substitution for GGBFS combined in the production of SCGC increased flexural strength.

• The permeability plummeted as the consumption of MK and CBA alone and jointly as a substitution for GGBFS in the production of SCGC after 28 days.

• The minimum change in length is recorded by 0.075 mm, 0.059 mm, and 0.062 mm at 15% of MK, 15% of CBA, and 7.5MK7.5CBA as a replacement for GGBFS while the maximum change in length is calculated by 0.096 mm, 0.10 mm and 0.11 mm at 20% of MK, 20% of CBA and 10CBA10MK as a substitution for GGBFS in SCGC at 180 days correspondingly.

• The embodied carbon of GGBFS based SCGC is getting decrease when the concentration of CBA rises while the EC of GGBFS based SCGC is getting higher with addition of MK as substitution for GGBFS alone or combine with CBA in SCGC.

• Response surface models predict GGBFS-based self-compacting geopolymer concrete’s fresh, mechanical, and durability characteristics using ANOVA. Models have R² values ranging from 88 to 99%. Using MK and CBA, desirability factor is 69.10%.

Author contributions

Naraindas Bheel: Conceptualization; Data analysis; Investigation; methodology; Writing - Original Draft; Writing - Review & Editing. Mamdooh Alwetaishi: Formal analysis; Funding Acquisition; visualization; Writing - Review & Editing. Idris Ahmed Jae: Formal Analysis; methodology; Writing - Review & Editing. Agusril Syamsir: Conceptualization; Formal analysis; methodology; Writing - Review & Editing. Ahmed Saleh Alraeeini: Formal analysis; Data analysis; visualization; Writing - Review & Editing. Sahl Abdullah Waheeb: Conceptualization; Formal analysis; visualization; Writing - Review & Editing. Loai Alkhattabi: Formal analysis; data curing; visualization; Writing - Review & Editing. Omrane Benjeddou: Conceptualization; Formal Analysis; and Writing - Review & Editing.

Funding

The authors extend their appreciation to Taif University, Saudi Arabia, for supporting this work through project number (TU-DSPP-2024-32).

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Competing interests

The authors declare no competing interests.