Four of the five substantive barrier categories examined in the model were significantly associated with living-wall implementation barriers. Integration and complexity barriers had a coefficient of β = 0.303, and awareness and economic barriers had a coefficient of β = 0.278, with both paths also significant at p ≤ 0.001. Maintenance and technical barriers differed from the other categories: their coefficient was β = 0.097 and their path was not statistically significant, with p = 0.227. These coefficients quantify relationships within the fitted model and do not establish that the associated factors caused particular project outcomes.

The researchers treated the identified barriers as formative constructs, meaning that each constituent indicator was expected to make a distinct contribution to its broader category. They assessed collinearity among those indicators using variance inflation factors. All reported VIF values were below 3.5, the study’s stated threshold for checking that the indicators retained distinct contributions. The model also demonstrated internal consistency: Cronbach’s alpha and composite reliability met the required thresholds, while average variance extracted values exceeded 0.50, or 50% of the variation specified in the assessment criteria.

This was an expert-survey study focused on the Kingdom of Saudi Arabia’s building industry, and its findings therefore concern the barriers represented by the surveyed experts and the measures included in the model rather than a direct comparison of completed living-wall projects. The results indicate that the modeled implementation problem was associated with a broad set of operational, behavioral, resource, structural, integration, complexity, awareness and economic factors, while the maintenance-and-technical category did not show a statistically significant path in the reported model. The authors present the findings as a basis for policy guidance intended to support living-wall adoption in Saudi Arabia and potentially other Gulf nations, but the supplied results do not evaluate the effects of any particular policy intervention.

Subject of Research: Expert-survey analysis of living-wall implementation barriers in Saudi Arabian construction projects

Article Title: Barriers to living walls implementation in Saudi Arabian sustainable construction projects: A PLS-SEM approach

Article References: Alaboud, N. S. (2026). Barriers to living walls implementation in Saudi Arabian sustainable construction projects: A PLS-SEM approach. Scientific Reports. https://doi.org/10.1038/s41598-026-76473-1

Image Credits: AI Generated

DOI: 10.1038/s41598-026-76473-1

Keywords: living walls, Saudi Arabia, sustainable construction, PLS-SEM, implementation barriers, expert survey, structural equation modeling, building industry, Barriers, living, walls, implementation

The study, published in Frontiers in Built Environment by Bonga PraiseGod Khuzwayo and colleagues at the University of South Africa, Durban University of Technology and partner institutions, tackles a familiar blind spot in construction engineering. Structural performance, construction productivity and environmental impact are usually assessed separately, often by different specialists using incompatible units and assumptions. That fragmentation makes it hard to trace why one material was chosen over another. The team’s answer is a multicriteria decision analysis procedure in which every criterion has an explicit engineering definition, a clear direction of preference and an auditable data source, all documented to three decimal places.

The framework rests on three calculated indicators. Structural Efficiency is defined as a common required axial design load of 30 kilonewtons divided by the self-weight of a one-square-metre wall panel, a deliberately narrow measure of demand relative to weight rather than a statement about structural safety. Production Efficiency comes from laboratory trials in which a three-person crew built 1.5-square-metre test panels, measuring completed wall area per hour. Embodied Energy sums renewable and nonrenewable primary energy over the product stage of the life cycle, covering raw material supply, transport to the factory and manufacturing, expressed in megajoules per square metre of wall.

Four alternatives entered the contest: conventional fired clay brick masonry, conventional hollow concrete block masonry, a ten-hole fired clay brick from Brickworks Limited, and a KEBE K190 vertically perforated clay brick. The two manufacturer-specific products came with independently verified Environmental Product Declarations, while the conventional systems relied on literature-based reference values. The authors were unusually candid about data provenance, classifying every input as experimentally derived, literature based, manufacturer reported, calculated or transferred. Because no direct productivity trials existed for the two branded bricks, they were assigned the conventional brick’s measured output and flagged explicitly as transferred analytical assumptions rather than measurements.

The pivotal step is the weighting. Rather than letting the researchers decide that energy matters more than speed, the framework uses entropy weighting, a technique borrowed from information theory that derives weights from the statistical dispersion of the data itself. Because Embodied Energy varied most sharply across the four alternatives, it captured the lion’s share of the discriminatory information, earning a weight of 0.706. Structural Efficiency received 0.216 and Production Efficiency just 0.078, the latter dragged down by three alternatives sharing the same transferred productivity value. The authors stress that these weights describe this particular dataset, not any universal hierarchy of engineering priorities.

The results were unambiguous. Conventional hollow concrete block masonry ranked first with an overall score of 0.984, combining the fastest measured construction rate of 1.027 square metres per hour with the lowest embodied energy of 295.650 megajoules per square metre. The KEBE K190 vertically perforated brick followed at 0.844, helped by the highest structural efficiency in the set thanks to its low self-weight. The Brickworks ten-hole brick scored 0.731, while conventional fired clay brick masonry trailed at 0.485, burdened by the heaviest wall and the highest embodied energy of 726.660 megajoules per square metre.

What makes the finding more than a one-off calculation is the battery of robustness checks. The researchers perturbed each of the least certain inputs, the transferred productivity values and the manufacturer-specific energy figures, independently by ten percent, and repeated the entire entropy calculation in all eight scenarios. No adjacent alternatives swapped positions. They then ran the same decision matrix through two entirely different aggregation methods, TOPSIS and VIKOR, and both reproduced the identical ranking. The ordering, in other words, is not an artefact of one mathematical recipe, though the authors caution that agreement among methods sharing the same inputs validates the aggregation, not the underlying data.

The study is equally notable for what it refuses to claim. The authors insist the framework is a screening tool for preliminary engineering decisions, not a substitute for code-based structural verification, fire and thermal assessment, or regulatory compliance. The structural indicator says nothing about compressive strength, slenderness or wall stability. The environmental assessment covers only the product stage and excludes transport to site, construction, maintenance and demolition. Embodied carbon, a distinct indicator from embodied energy, was left out entirely. The alternatives also differ in geographic origin and manufacturing technology, so the ranking is presented as a controlled methodological comparison rather than a market-wide verdict on any product.

For practitioners, the proposed workflow is a two-stage gate. Mandatory requirements such as compressive resistance, fire performance and durability must be verified first, as non-negotiable constraints that no environmental score can compensate for. Only then do feasible alternatives enter the entropy-weighted ranking, and any decision should be accompanied by the data vintage, system boundary and uncertainty statement so it can be audited and updated. The framework is also modular: cost, thermal resistance or embodied carbon can be added later, but adding any criterion or alternative forces a complete recalculation, because entropy weights are conditional on the full decision matrix.

The broader significance lies in the transparency the method demands. As buildings come under increasing pressure to cut their life-cycle energy demand, material choices that once rested on habit or single-metric comparisons are being subjected to the same auditable, data-driven scrutiny as any other engineering calculation. This study shows that when structural weight, build speed and embodied energy are placed in one matrix and weighted objectively, the traditional pecking order of masonry materials can change dramatically, and, more importantly, everyone can see exactly why.

Subject of Research: An entropy-weighted multicriteria decision framework integrating structural efficiency, construction productivity and embodied energy for masonry unit selection

Article Title: An entropy weighted multicriteria decision framework for masonry unit selection

Article References: Khuzwayo, B. P., Elsaigh, W., Rankine, R., & Walker, M. (2026). An entropy weighted multicriteria decision framework for masonry unit selection. Frontiers in Built Environment, 12, Article 1942111. https://doi.org/10.3389/fbuil.2026.1942111

Image Credits: AI Generated

DOI: 10.3389/fbuil.2026.1942111

Keywords: masonry, embodied energy, entropy weighting, multicriteria decision analysis, concrete blocks, clay bricks, environmental product declarations, construction productivity, structural efficiency, TOPSIS, VIKOR, sustainable construction

Semi-flexible pavements occupy an interesting middle ground in road engineering. They combine a porous, open-graded asphalt skeleton with a cement-based grout that is poured into the voids, producing a surface that behaves partly like a flexible asphalt layer and partly like a rigid concrete one. The result is a pavement that resists rutting under heavy loads and resists fuel spills and heat better than conventional asphalt, which is why it is popular for bus lanes, intersections, airports and heavy-duty industrial yards. The catch is the grout: a high-strength cement slurry demands a substantial cement content, and cement production is responsible for a significant share of global carbon dioxide emissions. Replacing even a fraction of that cement with an industrial by-product is therefore an attractive sustainability lever.

The research team, led by Muhammad Imran Khan of Imam Mohammad Ibn Saud Islamic University in Riyadh together with collaborators at Universiti Teknologi PETRONAS, the University of Tabuk and Kabul University, set out to quantify exactly how much cement could be swapped for waste marble dust without destroying the mechanical performance the grout needs. Rather than relying purely on trial-and-error laboratory work, they built a dataset of 72 experimental mixtures spanning water-to-cement ratios from 0.35 to 0.50, marble dust contents from 0 to 25 percent, measurements of grout flow, and compressive strengths recorded at both 7 and 28 days of curing. That structured dataset became the training ground for five different machine learning algorithms.

The five contenders represent a broad cross-section of supervised learning. Artificial neural networks, loosely inspired by the wiring of biological neurons, are famously flexible function approximators. Decision trees split the data through a cascade of simple yes-or-no questions, random forests average the votes of many such trees, support vector machines carve the input space with carefully placed boundaries, and k-nearest neighbor models simply look up the most similar past cases. Each model was asked to predict the 28-day compressive strength of the grout from the mixture variables, and their performance was scored using the mean absolute error, the coefficient of determination known as R-squared, and the adjusted R-squared, which penalizes needless complexity.

The surprise winner was not the flashiest algorithm. The humble decision tree, a model that essentially learns a flowchart of rules, achieved the highest predictive accuracy, with an R-squared of 0.994 and a mean absolute error of just 0.266 megapascals. In practical terms, the model could anticipate the strength of an unseen grout mixture to within roughly a quarter of a megapascal on average, a margin tight enough to be genuinely useful for mixture design. The result is a reminder that in materials informatics, the best model is not always the most elaborate one; when the underlying relationships in a well-controlled dataset are relatively clean, an interpretable model can match or beat its more sophisticated rivals.

Accuracy alone, however, was only half the goal. Modern machine learning in engineering increasingly demands explainability, and the researchers turned to two well-established tools to open the black box. SHapley Additive exPlanations, or SHAP, borrows ideas from cooperative game theory to assign each input variable a fair share of credit for every individual prediction, revealing which factors push the expected strength up or down. Partial dependence plots complement this by tracing how predictions shift as one variable changes while the others are held in place. Together, these methods let the team read the model’s reasoning the way one might read a lab notebook.

The explainability analysis produced a striking conclusion: the single most influential parameter governing the predicted 28-day strength was the 7-day compressive strength itself. That may sound circular, but it is not. The 7-day measurement is an experimentally observed snapshot of how the grout’s hydration and hardening are progressing at an early age, so it carries rich information about the mix’s intrinsic quality and its response to curing. A grout that is already strong at one week is very likely to be strong at four weeks, and the model captures that correlation with high confidence. For engineers, this suggests that early-age testing could serve as a powerful early-warning indicator, allowing underperforming mixtures to be flagged and discarded long before the standard 28-day verification is complete.

The second lesson from the model was more sobering for anyone hoping that marble dust is a free lunch. Increasing the water-to-cement ratio and increasing the marble dust content both reduced the 28-day compressive strength. Marble dust is largely inert or only weakly reactive compared with cement clinker, so as its share rises, there is simply less clinker available to form the binding phases that give the grout its strength, while additional water dilutes and weakens the matrix further. The physics and chemistry are straightforward, but the machine learning models quantified the trade-offs precisely and revealed where the sweet spot lies.

Using the trained models as an optimization engine, the team identified mixture proportions that push cement replacement surprisingly far. At a water-to-cement ratio of 0.40 and with a modest 0.5 percent superplasticizer dose, the optimization allowed up to 32 percent cement replacement with waste marble dust while still meeting the performance targets for the grout. The superplasticizer is the quiet hero here: by dispersing cement particles and improving workability without extra water, it creates headroom for the powder-rich substitution. The analysis also showed that the replacement strategy loses efficiency at higher water-to-cement ratios, meaning there is a genuine window of mixture design within which marble dust substitution is both technically viable and environmentally rewarding, and outside of which it is not.

The broader implications extend well beyond one pavement technology. The study demonstrates a data-driven template for sustainable mixture design: run a disciplined experimental campaign, train a family of models, validate them rigorously, explain them with SHAP and partial dependence analysis, and then use them to search a design space that would be prohibitively expensive to explore with physical trials alone. Applied to semi-flexible pavements, the workflow shows that a stone-industry waste product can displace nearly a third of the cement in the grout under the right conditions, cutting material costs, diverting landfill waste and lowering the embodied carbon of the road surface. As infrastructure owners worldwide face pressure to decarbonize, studies like this one suggest that the roads of the future may be paved, quite literally, with industrial by-products, guided by algorithms that know exactly how far the substitution can go before the pavement pays the price.

Subject of Research: Machine learning optimization of waste marble dust cementitious grouts for semi-flexible pavements

Article Title: Predictive modeling and optimization of eco-friendly cementitious grouts containing waste marble dust for semi-flexible pavements

Article References: Khan, M. I., Khan, N., Almutairi, O. E., Rafiq, W., & Paktin, H. (2026). Predictive modeling and optimization of eco-friendly cementitious grouts containing waste marble dust for semi-flexible pavements. Scientific Reports. https://doi.org/10.1038/s41598-026-75077-z

Image Credits: AI Generated

DOI: 10.1038/s41598-026-75077-z

Keywords: semi-flexible pavement, waste marble dust, cement replacement, machine learning, decision tree, SHAP, partial dependence plots, compressive strength, water-cement ratio, sustainable construction, grout optimization, circular economy

Expanded perlite is a lightweight mineral aggregate created by heating perlite, a volcanic glass, until it expands into a porous, low-density material. While its natural lightness makes it an attractive candidate for reducing the overall weight of building structures, its high porosity presents a significant engineering challenge. The inherent voids within the perlite particles create a trade-off between density and mechanical strength, a dilemma that has historically limited its application in construction. The study conducted by researchers from Hamad Bin Khalifa University in Qatar and Al-Isra University in Jordan aims to resolve this tension by systematically evaluating how different processing variables influence the final properties of cementitious composites incorporating this material.

The experimental design of the study was rigorous, involving the production of eighteen distinct formulations to map out the complex relationship between particle size, matrix composition, and conditioning protocols. The researchers utilized three series of expanded perlite particle sizes, designated as S1 with a diameter of 4.0 mm, S2 at 2.0 mm, and S3 at 0.8 mm. Each particle size series was combined with six different matrix formulations, labeled a through f, to create a comprehensive matrix of test specimens. This extensive combinatorial approach allowed the team to isolate the effects of specific variables on the mechanical behavior of the resulting composites, providing a detailed dataset for optimizing material performance.

To assess the long-term stability and practical viability of these materials, the specimens were subjected to two different conditioning regimes over a 28-day period. One group was maintained under standard ambient conditions, simulating typical curing environments found in many construction sites. The second group underwent a modified protocol, involving seven days of conditioning at an elevated temperature of 45 degrees Celsius, followed by 21 days under ambient conditions. This low-temperature post-demoulding conditioning was explored as a potential method to enhance material properties without the high energy costs associated with conventional high-temperature curing processes, aligning with the broader goal of resource efficiency in sustainable construction.

The results of the mechanical testing revealed significant variations in performance across the different formulations. Under ambient conditioning, the formulation designated as S3d, which utilized the smallest 0.8 mm perlite particles, emerged as the top performer in terms of compressive strength. This specific composite achieved a compressive strength of 1.904 MPa, with a bulk density of 720 kg/m³ and a peak load capacity of 4.76 kN. These figures indicate that finer particle sizes can contribute to a denser, stronger matrix by allowing for better packing and interlocking of particles within the cementitious binder, thereby improving the overall structural integrity of the composite.

Conversely, the study identified that formulations S3c and S2d provided the lowest-density alternatives among the tested samples. While these options may not match the peak strength of S3d, their reduced mass makes them highly suitable for applications where weight reduction is a primary design constraint, such as in the construction of upper-story partitions or lightweight roofing systems. The overall density range across all eighteen formulations spanned from 528 to 752 kg/m³, demonstrating the material’s versatility in achieving various weight targets depending on the specific requirements of a construction project.

The introduction of the 45-degree-Celsius conditioning protocol yielded nuanced results that challenge the assumption that higher temperatures always improve material performance. While the elevated temperature conditioning generally resulted in a reduction in bulk density, it also led to a decrease in compressive strength and post-peak deformation capabilities. This finding suggests that the thermal treatment may have altered the hydration kinetics of the cement matrix or the microstructure of the perlite particles in a way that compromised their mechanical robustness. The formulation S1d, which used the largest 4.0 mm particles, was identified as the highest-efficiency option after this specific conditioning process, highlighting the complex interplay between particle size and thermal history.

The concept of strength-density efficiency, which measures the mechanical output relative to the material’s mass, proved to be a critical metric in evaluating the composites. The study found that no single formulation was optimal across all performance objectives, with S3d excelling under ambient conditions and S1d showing superior efficiency after 45-degree-Celsius conditioning. This lack of a universal best performer underscores the importance of tailoring material specifications to specific application scenarios. For architects and engineers designing sustainable buildings, this data provides a valuable framework for selecting the appropriate perlite composite based on the specific environmental and structural demands of a project.

The implications of these findings extend beyond individual building components to the broader context of smart urban infrastructure. By developing resource-efficient, non-load-bearing building materials, the construction industry can reduce the demand for heavy, energy-intensive traditional concrete, thereby lowering the embodied carbon of new developments. The lightweight nature of these composites also facilitates easier transportation and installation, potentially reducing construction time and labor costs. As cities strive to meet stringent environmental targets and improve the resilience of their infrastructure, the adoption of such innovative materials could play a pivotal role in achieving sustainable urban growth.

Despite the promising results, the researchers emphasize that further investigation is required to fully validate the contribution of these composites to sustainable urban development. Future studies need to explore the thermal performance, long-term durability, and comprehensive environmental impacts of expanded perlite composites. Understanding how these materials behave under varying climatic conditions and over extended periods will be crucial for their widespread adoption in critical infrastructure. Nevertheless, this study establishes a solid experimental basis for the continued development of lightweight construction materials, offering a clear path toward more sustainable and efficient building practices in the era of smart cities.

Subject of Research: Experimental evaluation of expanded perlite cementitious composites for sustainable construction

Article Title: Development and performance evaluation of resource-efficient expanded perlite composites for sustainable building and smart urban infrastructure

Article References: Ashraf, U., Basha, O., Hzami, A., Khattab, T., & Abu-Rayash, A. (2026). Development and performance evaluation of resource-efficient expanded perlite composites for sustainable building and smart urban infrastructure. Discover Sustainability. https://doi.org/10.1007/s43621-026-05013-8

Image Credits: AI Generated

DOI: 10.1007/s43621-026-05013-8

Keywords: expanded perlite, lightweight concrete, sustainable construction, smart cities, cementitious composites, building materials, resource efficiency, urban infrastructure, mechanical properties, green building, Development, performance

The choice of an L16 orthogonal array is central to the efficiency of the experimental program. Orthogonal designs, rooted in the work of Genichi Taguchi and classical factorial design theory, allow the independent estimation of main effects for multiple factors using a fraction of the full factorial run count. With five mix parameters under investigation—slag content, metakaolin content, sodium hydroxide molarity, the sodium silicate to sodium hydroxide ratio, and the alkali-to-binder ratio—a full factorial exploration at even modest resolution would require dozens to hundreds of paste batches. The L16 array compresses this into sixteen carefully balanced runs in which each factor level appears an equal number of times and factor levels are distributed orthogonally across columns. This balance means that, on average, the influence of one factor is not confounded with the linear influence of another, enabling analysis of variance to attribute portions of observed variance to individual parameters. The trade-off, as the authors themselves note, is that certain two-factor interactions cannot be fully disentangled from main effects within a resolution-limited array, a limitation that is well understood in design-of-experiments practice and typically addressed through follow-up experiments at the identified optimum.

The finding that the alkaline activator ratio exerted the greatest influence on compressive strength aligns with the fundamental chemistry of geopolymerisation. The sodium silicate to sodium hydroxide ratio controls the availability of dissolved silica in the activating solution, which in turn governs the Si/Al ratio of the forming gel network. Higher soluble silica content generally promotes denser, more cross-linked aluminosilicate gels and, in calcium-rich systems, modifies the composition of C-(A)-S-H phases toward higher chain length and greater mechanical integrity. Dissolved silica also moderates the highly caustic environment of sodium hydroxide, affecting the rate at which precursor particles dissolve and the degree to which dissolved species polycondense into stable networks. Meanwhile, the identification of the alkali-to-binder ratio as the dominant factor for flowability is consistent with rheological behavior in alkaline-activated pastes: the total volume of liquid relative to solid determines the spacing between particles and the viscosity of the suspending medium, and alkaline activator solutions are substantially more viscous than water, amplifying their effect on workability.

The economic dimension of the optimization deserves particular attention because it is frequently neglected in laboratory studies of alternative binders. Metakaolin, produced by calcining kaolinitic clay at moderate temperatures, carries an embodied energy cost and a market price considerably higher than that of fly ash or slag, which are industrial byproducts available at low cost near their points of generation. The study’s identification of metakaolin dosage as the dominant contributor to material cost therefore has direct practical consequences: even modest increases in metakaolin content can shift the economics of a mix substantially, which is why the optimal formulation converged on a low metakaolin fraction of ten percent. This cost sensitivity also highlights a general principle in sustainable materials engineering—the environmental and economic benefits of byproduct utilization depend heavily on local supply chains, transport distances, and competing demands for the same materials from other industries, factors that a laboratory cost model can only approximate.

The curing regime established in the preliminary single-factor experiments also merits scientific context. Geopolymerisation is a thermally activated dissolution-condensation process, and low-calcium fly ash systems in particular require elevated temperatures to achieve meaningful reaction within practical timeframes. Temperatures in the range of 60 to 80 degrees Celsius are commonly reported as favorable for fly ash activation, with the 60 degrees Celsius, 24-hour condition identified here as optimal balancing reaction extent against the risk of excessive water loss or gel coarsening at higher temperatures. The presence of slag in the ternary system partially relaxes this thermal requirement, since calcium-rich systems gain strength appreciably at ambient temperature, but heat curing remains beneficial for achieving early strength in blended systems. This thermal dependence has implications for field application, where ambient-temperature curing or modest heat treatment would be required, and it represents one of the practical hurdles that distinguishes geopolymer concretes from conventional Portland cement systems in cold-weather construction.

The microstructural observations reported through X-ray diffraction and scanning electron microscopy provide the mechanistic underpinning for the mechanical results. X-ray diffraction in geopolymer systems typically reveals the characteristic broad hump associated with short-range ordered aluminosilicate gel, superimposed on residual crystalline phases such as quartz and mullite inherited from the fly ash. Shifts in the position of this hump and changes in its relative intensity reflect alterations in gel composition and quantity. Scanning electron microscopy, in turn, allows direct visualization of gel morphology, unreacted particle remnants, and the pore network. The reported features of enhanced C-(A)-S-H and N-A-S-H gel formation, increased matrix densification, and reduced porosity in the optimized mixtures are exactly what reaction mechanism models would predict for a well-balanced ternary system: slag-derived calcium gel fills space and provides early binding, while fly ash and metakaolin-derived sodium aluminosilicate gel contributes to long-term framework development, and the coexistence of the two gel types in an intermixed morphology is widely associated with superior mechanical performance in hybrid systems.

The statistical methodology underlying the decision model also reflects current best practice in multi-criteria optimization. Entropy weighting is an objective technique that assigns greater weight to criteria exhibiting greater variability across the experimental alternatives, on the reasoning that a parameter that discriminates strongly between options carries more information for decision-making. Combining this with subjective weighting, which encodes engineering judgment about the relative importance of strength, workability, and cost, produces a synthetic weighting scheme that tempers purely data-driven assignment with domain expertise. This hybrid approach guards against the failure modes of either extreme: purely subjective weights can be arbitrary and difficult to defend, while purely entropy-based weights can overemphasize criteria whose variability is an artifact of the chosen experimental levels rather than of genuine practical importance. The convergence of the multi-objective analysis on a mixture with moderate values of most parameters—fifty percent slag, ten percent metakaolin, six molar sodium hydroxide, an activator ratio of one to one point five, and an alkali-to-binder ratio of zero point three five—is characteristic of Pareto-optimal solutions in multi-objective problems, where the best compromise rarely coincides with the extreme values that maximize any single objective.

The reported maximum compressive strength of 115.87 megapascals under one experimental condition is notable in the context of published geopolymer literature, where ambient-cured fly ash systems typically achieve strengths in the range of 30 to 70 megapascals and heat-cured or slag-rich systems can exceed 100 megapascals. The high value is consistent with the combination of substantial slag content, heat curing, and an activator chemistry favoring dense gel formation. However, the authors’ transparency about the replication structure of their testing is scientifically important. The three compressive strength specimens per mixture were subsamples drawn from a single batch rather than products of independent batch preparations, which means that the reported variability reflects within-batch homogeneity rather than between-batch reproducibility. Between-batch variation, arising from differences in mixing efficiency, temperature, activator preparation, and precursor sampling, is typically larger than within-batch variation, so the true uncertainty surrounding each mixture’s mean strength is likely wider than the within-batch statistics alone would suggest. The observation that repeat testing broadly preserved the relative ranking of selected mixtures is reassuring for the optimization conclusion, since ranking-based decisions are more robust to absolute strength uncertainty than threshold-based ones, but independent replication at the batch level remains a desirable confirmation step.

From a sustainability perspective, the study contributes to a growing body of evidence that industrial byproducts can substitute for Portland cement clinker at high replacement levels without sacrificing mechanical performance. Cement production is estimated to account for roughly seven to eight percent of global carbon dioxide emissions, driven both by the calcination of limestone and by the fossil fuel energy consumed in clinker kilns. Geopolymer binders eliminate the calcination step entirely, and while the alkaline activators—particularly sodium silicate—carry their own embodied emissions, life cycle assessments generally report substantial net reductions for geopolymer concretes relative to ordinary Portland cement equivalents, with the magnitude depending on activator dosage and transport distances. Ternary systems that incorporate slag, itself a finite byproduct facing increasing demand, raise questions about long-term resource availability, which strengthens the case for including calcined clays such as metakaolin in precursor blends, since clay reserves are abundant and calcination can be performed at temperatures far below those required for clinker production.

The workability findings also carry practical significance for construction practice. Flowability, typically assessed in paste systems through mini-slump or flow table tests, serves as a proxy for the fresh-state behavior of the corresponding concrete, where workability governs placement, compaction, and finishability. The strong influence of the alkali-to-binder ratio on flowability identified in the effect analysis suggests that contractors seeking to adjust workability in the field would have a more effective lever in the liquid content than in the activator composition, though such adjustments must be balanced against the strength consequences of diluting the alkaline environment. The interplay between workability and strength is a recurring theme in geopolymer technology, and the multi-objective framework applied here offers a structured way to navigate it rather than relying on iterative trial and error, which the authors correctly identify as laborious and inefficient for multi-parameter systems.

Finally, the methodological integration demonstrated in this study—combining screening-level single-factor experiments, an orthogonal array for efficient multi-factor assessment, analysis of variance for effect attribution, and a hybrid weighted decision model for multi-objective selection—constitutes a transferable template for other researchers working on complex binder systems. The approach is not limited to geopolymers; it applies equally to alkali-activated materials, limestone calcined clay cements, and other multi-component systems where formulation space is large and performance criteria are competing. The explicit acknowledgment of limitations, including the interaction confounding inherent in the chosen array and the subsample nature of the mechanical replicates, models the kind of methodological candor that allows subsequent studies to build incrementally on reported results, refining the optimum through targeted confirmation experiments and extending the framework to durability indicators, long-term strength development, and full-scale concrete applications where aggregate packing and interfacial transition zone effects introduce additional layers of complexity.

Subject of Research: Optimization of a ternary geopolymer paste through orthogonal experimental design, ANOVA and entropy-based multi-objective decision analysis

Article Title: Optimization of a ternary geopolymer paste through orthogonal experimental design, ANOVA and entropy-based multi-objective decision analysis

Article References: Murugeswari, S., & Meena, T. (2026). Optimization of a ternary geopolymer paste through orthogonal experimental design, ANOVA and entropy-based multi-objective decision analysis. Frontiers in Built Environment, 12, Article 1934627. https://doi.org/10.3389/fbuil.2026.1934627

Image Credits: AI Generated

DOI: 10.3389/fbuil.2026.1934627

Keywords: geopolymer, fly ash, GGBS, metakaolin, orthogonal experimental design, ANOVA, entropy weighting, multi-objective optimization, compressive strength, sustainable construction, XRD, SEM

The material in question is ecological porous concrete, a deliberately sponge-like formulation used in green infrastructure projects such as vegetated retaining walls, slopes, and pavements. Unlike conventional concrete, EPC contains interconnected voids that allow water, air, and roots to pass through. That porosity is its ecological selling point, but it also creates an engineering headache: the same connected pores that admit roots also expose them to the cement matrix, and cement chemistry is brutally hostile to plant life.

The problem stems from the fundamental reactions of cement hydration. When Portland cement reacts with water, it releases calcium hydroxide and other alkaline compounds that push the pore solution well above pH 11, sometimes approaching 13. At those levels, root cells are damaged, nutrient uptake collapses, and microbial communities essential for healthy soil struggle to survive. Meanwhile, the dense cementitious regions that give concrete its strength tend to clog the very pores that vegetation needs. The result is a persistent trade-off between mechanical performance and biological viability, a compromise that has limited EPC adoption despite growing demand for green urban infrastructure.

Wang and colleagues approached the problem by treating watermelon vine as what they call a dual-functional additive, meaning it works in two ways at once. Mechanically, the fibrous plant matter acts as reinforcement distributed through the matrix, helping to bridge microcracks and maintain strength even as the mix is engineered for high porosity. Chemically, the vine’s organic constituents interact with the alkaline pore environment. Watermelon plants are known to accumulate oxalates, and the researchers propose that organic acids from the vine neutralize hydroxide ions in situ, buffering the pore solution toward a pH plants can tolerate.

Determining the right recipe for such a multi-objective material is a genuine optimization challenge, because adding more vine lowers pH but can also compromise strength and porosity. The team therefore built an integrated statistical framework combining an orthogonal array design, response surface methodology, and entropy-weighted TOPSIS analysis. Orthogonal arrays allow many mix variables to be tested efficiently in a reduced number of experiments, response surface methodology then models how each factor influences the outcomes and locates optima in the multi-dimensional design space, and TOPSIS, a multi-criteria decision technique, ranks candidate mixes by their overall closeness to an ideal that balances strength, porosity, and pH simultaneously. Entropy weighting assigns objective importance to each criterion based on the spread of the experimental data rather than subjective assumptions.

This analysis yielded a clear verdict: watermelon vine content was the dominant factor controlling pore pH. At an optimal dosage of 3 percent by weight of binder, the pH dropped to 7.6 within 14 days, a value comfortably within the neutral range where roots and soil microbes can function. No conventional EPC formulation comes close to that figure without external chemical treatment, which makes the result striking given that the neutralizing agent is essentially farm waste.

The optimized mix itself was tailored for vegetation. The researchers specified coarse aggregate in the 16 to 20 millimeter range, which creates the large connected voids characteristic of porous concrete, supplemented with 20 percent fly ash replacing part of the cement. Fly ash, a byproduct of coal combustion, both reduces the carbon footprint of the binder and contributes to longer-term strength through pozzolanic reactions. The water-to-binder ratio was set at a low 0.26, alongside the 3 percent vine addition. Together these choices produced a compressive strength of 10.4 megapascals and an effective porosity of 27.7 percent, a combination that sits squarely in the useful window for load-bearing yet plantable concrete. Critically, the strength remained stable when tested at 60 days, indicating that the organic additive did not degrade the matrix over time, a common worry when biological materials are embedded in cement.

To understand what was happening at the microscopic scale, the team turned to scanning electron microscopy. The images revealed a dense interfacial transition zone, the region of cement paste that surrounds aggregate particles and typically represents the weakest link in concrete microstructure. Within this zone, the researchers observed crystalline products whose morphology is consistent with calcium oxalate, supporting the proposed acid-base neutralization mechanism in which oxalate-rich vine tissue reacts with the alkaline cement environment to form stable minerals. The formation of such crystals may even contribute to densifying the interface, helping explain how the material retains strength despite its open, porous architecture.

The biological endpoint was tested the most direct way possible: by planting seeds in the concrete and watching what happened. Germination trials confirmed that vegetation could establish itself on the optimized material, achieving greater than 80 percent coverage within 60 days. For applications such as vegetated slope stabilization, green walls, and ecological restoration of hardened landscapes, that figure translates into rapidly greening surfaces that can reduce erosion, moderate urban temperatures, and support biodiversity. The fact that this is achieved with a locally sourced waste stream rather than imported chemical additives strengthens the economic case, particularly in agricultural regions where watermelon cultivation generates vine residues seasonally and disposal is a recurring cost.

The broader significance of the work lies in its demonstration that industrial and agricultural waste streams can be paired in unexpected ways to solve structural problems. Porous, plantable concrete has existed for decades, but the alkalinity barrier has kept it confined to niche projects. If a dried, shredded vine can pull pore chemistry into the neutral range while reinforcing the matrix, the calculus for sustainable construction materials changes considerably. The researchers positioned watermelon vine as a viable, low-cost functional resource, and the convergence of their evidence, from statistical optimization to microscopy to living plants, offers a template for how future green building materials might be designed: not by adding functionality after the fact, but by building biology directly into the chemistry of the material itself.

Subject of Research: Use of watermelon vine agricultural waste as a dual-functional additive to strengthen and de-alkalinize ecological porous concrete for vegetation

Article Title: Watermelon vine as dual-functional additive in ecological porous concrete

Article References: Wang, Y., Zhuang, J., & Ma, L. (2026). Watermelon vine as dual-functional additive in ecological porous concrete. Scientific Reports. https://doi.org/10.1038/s41598-026-74981-8

Image Credits: AI Generated

DOI: 10.1038/s41598-026-74981-8

Keywords: watermelon vine, agricultural waste, ecological porous concrete, alkalinity reduction, compressive strength, porosity, fly ash, response surface methodology, TOPSIS, vegetation establishment, calcium oxalate, sustainable construction

The binder, described in a study published in Frontiers in Materials, combines three unlikely ingredients: ground granulated blast-furnace slag (GGBS), a glassy by-product of iron manufacturing; carbide slag, a calcium-rich residue from acetylene production; and biochar made by pyrolyzing corn straw at 500 degrees Celsius. The researchers call the mixture a biochar-modified alkali-activated slag, or BAS, binder. Unlike conventional cement, it contains no clinker, the calcium-silicate ingredient whose production accounts for a substantial share of global carbon dioxide emissions from limestone calcination and fuel combustion. Because dredged silt treatment involves large volumes of relatively low-value material, replacing clinker in this application could have an outsized environmental payoff.

To find the best formulation, the team used response surface methodology, a statistical optimization technique that models how multiple variables interact to influence an outcome. They ran a 17-experiment Box-Behnken design varying the GGBS content between 5 and 15 percent, the carbide slag between 4 and 10 percent, and the biochar between 0 and 10 percent of total specimen mass, measuring unconfined compressive strength at 7 and 28 days. The resulting quadratic models fit the data remarkably well, with R-squared values of 0.9905 for the 7-day strength and 0.9962 for the 28-day strength. Among the linear terms, GGBS dominated, followed by carbide slag and then biochar, indicating that the reactive slag precursor drives most of the strength development.

The optimization module recommended 13.26 percent GGBS, 8.25 percent carbide slag, and 4.30 percent biochar by total specimen mass, predicting an unconfined compressive strength of 6.32 megapascals at 28 days. A confirmation test measured 6.25 megapascals, a relative error of just 1.12 percent. When the three dosages were normalized, they produced an internal dry-mass ratio of 51.38 parts GGBS to 31.96 parts carbide slag to 16.66 parts biochar, which the team then held constant across all subsequent tri-axial experiments. This two-stage approach, screening the composition by unconfined strength and then testing the optimized blend under confinement, is what distinguishes the study from earlier work that stopped at simple strength measurements.

The chemistry behind the binder is straightforward in principle. In calcium-rich precursors like GGBS, alkaline activation dissolves reactive glassy phases and promotes the precipitation of calcium-silicate-hydrate and calcium-aluminosilicate-hydrate gels. These products coat mineral particles, bridge particle contacts, and refine the pore network, which increases both the bonded component of strength, captured as cohesion, and the frictional resistance between particles. Carbide slag, dominated by portlandite and calcite, supplies both the alkalinity and the calcium needed for this reaction without requiring concentrated sodium hydroxide or sodium silicate solutions, making it more practical for large-volume earthwork.

The biochar plays a subtler role. Its porous structure, surface functional groups, and water-retention capacity can modify particle packing and the internal moisture environment during hydration. Previous studies have shown that biochar’s effect on cementitious composites depends heavily on particle size, dosage, and dispersion, with excessive amounts diluting the cementitious phase or introducing weak porous inclusions. The optimized 4.30 percent dosage appears to sit in a productive middle range, though the researchers note that their data do not isolate the biochar contribution directly.

For the tri-axial testing program, the team prepared cylindrical specimens 39.1 millimeters in diameter and 80 millimeters tall from dredged silt collected at a river-channel construction site in Jinan, China. The silt, classified as a low-liquid-limit clay with a natural water content of 48.5 percent, was dried, sieved, and homogenized before mixing. Specimens were cured at 20 degrees Celsius and at least 95 percent relative humidity, then vacuum-saturated and immersed in water for 12 hours before testing, a deliberately harsh conditioning that simulates field exposure. Consolidated undrained tri-axial tests were run at confining pressures of 50, 100, and 150 kilopascals, and the results were interpreted using the Mohr-Coulomb failure criterion to extract cohesion and internal friction angle as total-stress parameters.

The results show a clear dosage response. At 18 percent water content and 28 days of curing, cohesion rose from 169.4 kilopascals at 5 percent BAS to 181.9 kilopascals at 10 percent and 195.2 kilopascals at 15 percent, while the friction angle increased from 38.5 to 42.3 to 42.5 degrees. Notably, the friction angle gained most of its improvement between 5 and 10 percent dosage and then plateaued, while cohesion continued climbing, suggesting diminishing returns on the frictional component at higher binder contents. Water content produced a non-monotonic response, with both parameters peaking at 18 percent, close to the 18.1 percent compaction optimum measured for the center-point mixture, and declining at both 15 and 21 percent. This inverted-U pattern likely reflects competing effects of wetting, ion mobility, compaction efficiency, and pore-solution dilution.

Curing age told a story of front-loaded strength gain. BAS cohesion increased from 190.2 kilopascals at 7 days to 195.2, 196.4, and 196.8 kilopascals at 28, 60, and 120 days, with roughly nine-tenths of the total change complete by 60 days. The friction angle followed the same trajectory, rising from 41.5 to 43.3 degrees over the same period. Compared with ordinary Portland cement tested under identical conditions, BAS performed remarkably well: at 15 percent dosage, BAS achieved 195.2 kilopascals cohesion and 42.5 degrees friction angle versus 191.7 kilopascals and 42.1 degrees for cement, arithmetic differences of under 3 percent. Against alkali-activated slag without biochar, BAS cohesion was 16.40 percent higher and the friction angle 12.73 percent higher; against lime, the gaps widened to 23.62 percent and 17.40 percent respectively.

The authors are careful to frame their conclusions within the boundaries of what was actually measured. The reported parameters are total-stress values applicable to the tested silt, preparation procedure, and confining-pressure range, not general design parameters, and pore-water pressures were not measured, so effective stresses were never calculated. No life-cycle assessment, carbon accounting, leaching testing, or microstructural analysis was performed, meaning no net environmental benefit can be quantified from this study alone. Untreated silt could not serve as a tri-axial control because it disintegrated during immersion. Still, the demonstration that a clinker-free binder built from steel slag, carbide residue, and corn straw can match cement in saturated shear performance marks a meaningful step toward converting two waste streams and an agricultural residue into lower-carbon geomaterials for infrastructure, pending the durability and environmental validation that future work must provide.

Subject of Research: Biochar-modified alkali-activated slag stabilization of urban dredged silt for infrastructure reuse

Article Title: Tri-axial shear performance of urban dredged silt stabilized with a biochar-modified alkali-activated slag binder

Article References: Wen, K., Zhan, Y., Feng, Q., Wei, M., Tang, Z., Wei, Z., & Zhao, Y. (2026). Tri-axial shear performance of urban dredged silt stabilized with a biochar-modified alkali-activated slag binder. Frontiers in Materials, 13, Article 1955304. https://doi.org/10.3389/fmats.2026.1955304

Image Credits: AI Generated

DOI: 10.3389/fmats.2026.1955304

Keywords: dredged silt, alkali-activated slag, biochar, soil stabilization, tri-axial shear strength, GGBS, carbide slag, clinker-free binder, geotechnical engineering, response surface methodology, sustainable construction, Mohr-Coulomb

The review followed the PRISMA guidelines, the gold-standard protocol for systematic literature reviews, and searched both the Web of Science and Scopus databases. The team started with a Boolean keyword search that yielded 683 results from Web of Science and 603 from Scopus, then removed duplicates to arrive at 1,064 unique records. Twenty additional papers were harvested from internet searches and reference-list browsing. After screening titles, abstracts, and full texts using the Rayyan web tool — with two reviewers working independently and a third verifying the final selection — only 24 studies survived the gauntlet. That small final sample, the authors note, reflects how young and specialized this research niche really is, not a failure of the search strategy.

What did those 24 studies reveal? The researchers used bibliometric analysis with VOSviewer software to map keyword co-occurrence networks and publication trends, then layered on qualitative content analysis. The keyword network showed that machine learning and additive manufacturing had the strongest link, with link strengths of 19 and 17 respectively, followed by artificial intelligence, deep learning, 3D printing, and 3D concrete printing. Four distinct thematic clusters emerged: AI methods for design generation (dominated by generative adversarial networks and deep learning), AI-driven optimization logic (machine learning, topology optimization, generative design, path planning), assessment and fabrication intelligence (computer vision, buildability, deep GANs), and the broader additive manufacturing context itself.

The first major research area is AI-driven design and form generation. At the conceptual stage, machine learning approaches help architects rapidly evaluate numerous design alternatives against predefined architectural and construction requirements. Generative models such as GANs and variational autoencoders support automated layout generation, producing unique spatial configurations and building designs that align with structural and functional demands. Some researchers have even experimented with large language models: one study showed that ChatGPT could generate codes and polynomial equations from descriptive language prompts, which were then used to build parametric models — a framework that promotes collaboration between architects and AI. Visualization tools like Veras and other AI-enhanced programs can convert simplified digital models into diverse architectural interpretations in minimal time.

Beyond creativity, AI plays a crucial technical role in 3DCP through predictive material modeling. Machine learning algorithms such as XGBoost have been reported to accurately forecast essential mix-design parameters, including slump flow, compressive strength, and overall structural strength, which facilitates more dependable and optimized printability of cement-based materials. Deep learning models can generate novel component geometries while simultaneously evaluating their printability. Semi-supervised models have been used to design micro-lattice and cellular structures for large-scale additive manufacturing, assessing both manufacturability and structural feasibility. Artificial neural networks extend these capabilities into sustainable building design by optimizing construction processes, predicting project performance, estimating costs, and supporting environmentally responsive design decisions — effectively becoming decision-support systems rather than mere predictive tools.

The second research area is optimization of material and structural performance. Here, machine learning has emerged as the dominant AI technique, facilitating data-driven optimization throughout the entire design-to-construction pipeline. Studies report that ML can help reduce material consumption, improve printability, and increase construction efficiency — potentially contributing to more sustainable construction practices. One notable example is a CNN-based framework developed by Ozalp and Aldemir that automatically estimates the structural capacities of 3D-printed concrete walls with optimal material parameters, performing design evaluations autonomously rather than requiring manual calculations. Another study applied explainable ensemble machine learning combined with deep generative adversarial networks to predict bond failure modes and strengths in 3DCP, using data augmentation to overcome limited experimental datasets. A probabilistic machine learning framework developed by Ma and colleagues reliably predicted bond strength and reinforcement embedment requirements with uncertainty estimates, improving both structural safety and construction efficiency.

The third research area is fabrication monitoring and assessment. Because 3DCP uses no formwork, printing defects must be detected before subsequent layers cover them or make the mistake irreversible. Computer vision systems can evaluate temporal changes in layer thickness and surface texture during printing, using image-derived indicators such as entropy standard deviation and maximum layer thickness deformation to assess buildability in real time. Deep learning models can spot defects, anomalies, and process changes by analyzing data from sensors and cameras, helping keep geometry consistent and preventing structural problems. Generative models like variational autoencoders and GANs help simulate defects and improve automated quality control. One particularly striking study introduced a tower-crane-based printing system controlled by deep reinforcement learning, which improves extruder toolpath control while assessing the accuracy and feasibility of large-scale, multi-storey printing — effectively making assessment part of an intelligent control loop rather than a separate inspection stage.

Despite these advances, the review identifies serious challenges that currently limit real-world implementation. At the material level, the concrete mixture must simultaneously satisfy conflicting requirements related to pumpability, extrudability, buildability, shape retention, open time, and early-age strength, with variations in water content, aggregate characteristics, and admixture dosage all playing determining roles. At the fabrication level, dimensional deviations can result from material deformation, inconsistent extrusion, nozzle movement, layer-height variation, and inaccuracies in robotic positioning — and these deviations can accumulate through early layers to influence final geometry and structural performance. Structural challenges include weak or variable interlayer bonding, anisotropic behavior, reinforcement integration, and uncertainty regarding long-term durability. A geometry that is computationally optimized may not necessarily be printable or structurally reliable under construction conditions.

The authors also flag a deeper systemic problem: most of the reviewed evidence comes from simulations, laboratory experiments, and controlled prototypes rather than full-scale construction sites. Assessment accuracy is strongly dependent on the quality and diversity of available data, and few public 3DCP datasets exist. Data augmentation methods like bootstrap resampling and deep generative adversarial networks are being used to make small datasets more robust, but synthetic data raises questions about whether artificially generated samples adequately represent real material variability and construction-site conditions. The review also notes the absence of standardized digital workflows connecting BIM, structural analysis, material data, slicing, robotic control, and quality verification, which restricts practical implementation.

Looking ahead, the researchers outline several promising directions: developing open benchmark datasets that combine material composition, printing parameters, environmental conditions, and observed printing defects; creating closed-loop fabrication control that integrates computer vision and sensor data to automatically adjust extrusion rate, nozzle speed, layer height, and tool paths in real time; building multi-objective design optimization frameworks that simultaneously consider structural performance, material consumption, embodied carbon, operational energy, printability, and construction cost; and connecting architectural models, structural simulations, and robotic control through AI-supported BIM and digital twin workflows. The authors suggest that large language models integrated with BIM platforms may eventually automate G-code generation, reinforcement detailing, and construction planning. If these gaps are addressed, AI-driven 3DCP could reshape architectural design, affordable housing, and sustainable construction — reducing material usage, improving printability, and automating quality control in ways that traditional methods simply cannot match.

Subject of Research: AI applications in 3D concrete printing for architectural design and construction

Article Title: Artificial intelligence driven transformation of 3D concrete printing in architecture: a systematic review

Article References: Premachandra, S., Silva, N. R., Mudunkotuwa, D. Y., Makavita, C. D., & Udawattha, C. (2026). Artificial intelligence driven transformation of 3D concrete printing in architecture: a systematic review. Frontiers in Built Environment, 12, Article 1914666. https://doi.org/10.3389/fbuil.2026.1914666

Image Credits: AI Generated

DOI: 10.3389/fbuil.2026.1914666

Keywords: artificial intelligence, 3D concrete printing, additive manufacturing, machine learning, generative design, computer vision, systematic review, sustainable construction, digital fabrication, architecture, deep learning, construction automation

The starting point was an uncomfortable truth about steel slag. This calcium-rich by-product of steelmaking looks ideal on paper, packed with lime, calcium silicates and iron-bearing phases, yet it hydrates painfully slowly at ambient temperature. Earlier long-term measurements showed that pure steel slag pastes reached only around 0.6 to 1.0 megapascals of compressive strength after seven days and just 2.8 to 3.1 megapascals after twenty-eight days, needing a full year to climb above twelve megapascals. The reactive ingredients are there, but they linger largely unconverted, leaving calcium-rich phases stranded instead of knitting them into the continuous calcium-silicate-hydrate gel that gives concrete its strength. Simply dumping steel slag into a cement mix, in other words, mostly dilutes it.

The researchers’ strategy was chemical bookkeeping. Steel slag supplies calcium and an alkaline environment, but it is starved of reactive silicon and aluminium. Phosphorus slag, a glassy discard from yellow phosphorus furnaces that is rich in both lime and silica, can donate the missing silicate species. Spontaneously combusted coal gangue — mining waste that has accidentally baked itself in self-sustaining fires — contains disordered, partially dehydroxylated clay minerals brimming with reactive silicon and aluminium that can consume calcium hydroxide and form calcium-aluminosilicate-hydrate gels. But each helper carries a penalty: phosphorus slag is notorious for delaying setting, with prior studies recording initial setting times stretched by up to 49 percent, while coal gangue’s porous, water-hungry particles sap the workability of fresh mortar.

To find the sweet spot, the team ran a progressive single, binary and ternary mixing programme, systematically varying both the steel slag fraction and the ratio of phosphorus slag to coal gangue, with no chemical activators added. The results were striking. Replacing 30 percent of cement with steel slag alone dropped the 28-day compressive strength from 48.99 to 32.39 megapascals, an activity index of just 66 percent. Binary blends with either phosphorus slag or coal gangue lifted the strength to roughly 40 megapascals. But the ternary blend with a steel slag to phosphorus slag to coal gangue ratio of 7:1:2 — mixture P-12 — reached 44.89 megapascals, an activity index of 91.63 percent and a full 38.59 percent stronger than the single-slag mixture.

Tellingly, the strength did not rise monotonically with any single ingredient. At the same 70 percent steel slag level, mixtures with phosphorus-to-gangue ratios of 2:1 and 1:1 both landed near 39.5 megapascals while the 1:2 ratio soared to 44.89. Too much phosphorus slag prolonged setting — mixture P-13, with the highest phosphorus content at that level, took 1,050 minutes to set initially, nearly six hours longer than P-12 — while too much coal gangue dragged down flowability through the water absorbed by its rough particles. The composition window mattered more than the quantity of any individual waste, and P-12 threaded it neatly, holding a respectable flow of 195 millimetres alongside its strength record.

The microstructural detective work, combining thermogravimetric analysis, X-ray diffraction with Rietveld refinement and scanning electron microscopy, produced perhaps the study’s most counterintuitive insight: the strongest mixture was not the one that consumed the most calcium hydroxide or produced the most bound water. The coal-gangue binary blend actually showed greater portlandite consumption and slightly more bound water than P-12, yet it was 13.79 percent weaker. Its hydration products clustered unevenly, leaving open pores and weakly connected zones between particle clusters.

P-12, by contrast, built a denser, more continuous microstructure. Electron images revealed gel-like calcium-silicate-hydrate, needle-like ettringite crystals and aggregated hydrates interwoven around residual particles, with several grains partly coated and bridged by reaction products. Energy-dispersive spectroscopy confirmed the coexistence of calcium, silicon, aluminium and sulphur in the hydrate regions — exactly the elemental recipe for mixed calcium-silicate-hydrate and calcium-aluminosilicate-hydrate gels together with sulphate-bearing phases. The message is one cement chemists are increasingly voicing: strength depends not on maximizing any single hydration indicator but on how effectively the hydrates are distributed through the pore space, coating particles and knitting them into a load-bearing network.

The team also pushed the winning recipe harder, stretching cement replacement to 35 and 40 percent. Strength fell to 35.85 and 30.52 megapascals respectively, and cement efficiency — strength delivered per gram of cement — declined by about 14 and 21 percent. The reason is structural: clinker hydration creates the early skeleton and supplies the calcium hydroxide that drives secondary reactions in the wastes. Starve the system of clinker and even a perfectly balanced waste blend cannot fully compensate. Notably, even at 40 percent replacement the cement efficiency remained slightly above that of plain cement, showing the wastes were still pulling chemical weight, but the numbers mark 30 percent as the practical ceiling for this composition.

The sustainability arithmetic drives the point home. Replacing cement cuts unit-mass emissions and cost — every blended mixture emitted roughly 540 kilograms of CO2 per tonne against 735 for plain cement — but the single-slag mixture’s strength collapse actually raised its carbon and cost per megapascal above the ordinary cement benchmark, a vivid demonstration that green concrete is worthless if you need more of it. P-12 avoided that trap entirely, posting the lowest strength-normalized global warming potential, 12.04 kilograms of CO2 per tonne per megapascal, and the lowest strength-normalized cost, 6.14 yuan per tonne per megapascal. Its material cost was just 0.77 percent higher than the single-slag mix, yet it delivered 38.59 percent more strength.

The broader lesson reaches beyond one laboratory recipe. As cement producers worldwide hunt for clinker substitutions, the temptation is to maximize waste loading and celebrate the reduced emissions on paper. This study shows why that logic fails and how to fix it: wastes are not interchangeable fillers but chemically distinct participants whose calcium, silica and alumina budgets must be jointly tuned so their hydration products interlock rather than segregate. The authors caution that long-term durability, volume stability, leaching behavior and field-scale trials remain to be demonstrated, and that raw-material variability across slag sources will require ongoing optimization. But as a proof of concept, turning steel-furnace residue, phosphorus discard and self-burned mining rubble into a binder that rivals conventional cement — at a quarter of the cost savings and more than nine-tenths of the strength — is a compelling glimpse of how construction’s raw materials could come from its own refuse heap.

Subject of Research: Steel slag-based low-cement binders improved by ternary blending with phosphorus slag and spontaneously combusted coal gangue

Article Title: Improving steel slag-based low-cement binders with phosphorus slag and spontaneously combusted coal gangue: mechanical performance and hydration microstructure

Article References: Shang, Y., Zhang, Y., Wang, Q., Nehdi, M. L., Zhang, L., & Liu, W. (2026). Improving steel slag-based low-cement binders with phosphorus slag and spontaneously combusted coal gangue: mechanical performance and hydration microstructure. Case Studies in Construction Materials, 25, Article e06625. https://doi.org/10.1016/j.cscm.2026.e06625

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06625

Keywords: steel slag, phosphorus slag, coal gangue, low-cement binder, cement replacement, compressive strength, hydration, microstructure, calcium-silicate-hydrate, CO2 emissions, sustainable construction, supplementary cementitious materials

The material in question is concrete slurry waste, the alkaline sludge generated when truck mixers and batching plants are washed out at ready-mixed concrete facilities, along with the surplus concrete that comes back when suppliers overorder. Traditionally treated as a disposal headache because of its high pH, moisture content and inconsistent composition, this slurry actually contains residual unhydrated clinker minerals and cement hydration products. Earlier work by the same researcher demonstrated something remarkable: when the slurry water is subjected to pressurized wet carbonation with supercritical carbon dioxide, each kilogram of the resulting solid can lock away approximately 130 grams of CO2 in stable carbonate form. That turns a waste stream into a carbon sink, but the open question was whether the carbonated powder could do useful chemical work inside a geopolymer, or whether it would simply dilute the mixture.

Geopolymers are binders built not on Portland cement clinker but on aluminosilicate precursors such as fly ash, metakaolin or ground granulated blast-furnace slag, known as GGBS, which are dissolved and reassembled by a strongly alkaline activator solution. In slag-rich systems, the abundant calcium drives the formation of calcium-sodium aluminosilicate hydrate gels, abbreviated C-(N)-A-S-H, which knit into a dense load-bearing network remarkably quickly. The new research, conducted by Sangrak Sim, asked what happens when carbonated concrete slurry waste, or CCSW, replaces 25, 50 or 75 percent of the GGBS by mass in such a system, with the activator-to-binder ratio held constant at 0.6 and a fixed sodium silicate to sodium hydroxide solution ratio of 2.0.

The carbonation step itself proved transformative. Uncarbonated slurry waste showed pronounced X-ray diffraction peaks from anhydrous clinker phases and portlandite, the crystalline calcium hydroxide left over from hardened cement. After ten minutes in a pressurized reactor at 40 degrees Celsius and 100 bar of CO2, those peaks vanished and calcite dominated instead. Perhaps more surprisingly, the carbonation process also shattered the particles: the median particle diameter dropped from 8.89 micrometers for the raw waste to 7.48 micrometers for the carbonated product, with no grinding involved, because the fine calcium carbonate crystals formed during carbonation break the material apart. Fine particles pack better and dissolve faster, which matters enormously in a reactive binder.

The mechanical results traced a clear dose-response curve. The pure GGBS control paste reached compressive strengths of 46.7 megapascals after one day, 73.9 after seven and 87.9 after 28 days of humid curing. At a 25 percent replacement level, the carbonated waste paste was nearly indistinguishable at early ages, hitting 46.4 megapascals on day one and finishing at 81.9 megapascals after 28 days, only about 6 megapascals below the control. Doubling the replacement to 50 percent cost roughly 23 megapascals of 28-day strength, while at 75 percent the paste managed just 4.7 megapascals at one day and 43.1 at 28 days, with almost no strength gain in the final three weeks. The pattern points to dilution: every kilogram of waste displaces a kilogram of highly reactive slag, and calcite, whatever its virtues, cannot rebuild the gel network on its own.

Yet the spectroscopic evidence revealed that the waste is far from inert. Fourier-transform infrared spectroscopy, deconvoluted across the 800 to 1200 per centimeter window where silicon-oxygen-aluminum stretching vibrations live, allowed the researcher to calculate a degree of geopolymerization for each paste. The control climbed from 49.1 percent at one day to 72.8 percent at 28 days. The 25 percent blend reached 64.4 percent, the 50 percent blend 62.3 percent, and even the 75 percent blend managed 56.2 percent, all rising steadily with curing age. Meanwhile, the calcite-related peak area actually shrank over time in some blends, dropping from 3.05 to 2.43 in the 50 percent mixture between days one and 28. That decline hints that under the fierce alkalinity of the activator, calcite surfaces partially dissolve, releasing calcium ions that feed the surrounding gel chemistry rather than sitting passively in the matrix.

Thermogravimetric analysis reinforced this picture with quantitative mass-loss fingerprints. Carbonate decomposition losses at 28 days rose predictably with waste content, from 1.41 percent in the control to 5.44 percent in the 75 percent blend, confirming the calcite inventory. But the mass loss between 300 and 600 degrees Celsius, a region complicated by overlapping dehydroxylation of hydroxyl-bearing gels and the decomposition of poorly crystalline carbonate species, also increased with waste content at every age. Combined with scanning electron microscopy, which revealed two distinct gel chemistries around waste particles, a calcium-poor aluminosilicate gel formed from the waste’s amorphous phases and a calcium-rich gel nucleated near carbonate surfaces, the data sketch a material playing three roles at once: a partial reactive precursor, a fine filler, and a nucleation scaffold.

Pore structure measurements tied these chemical stories to physical performance. Mercury intrusion porosimetry showed that at one day, the 25 percent blend actually had the densest pore structure of all mixtures, its fine carbonate particles physically plugging capillary spaces and seeding gel precipitation. But by seven days the control had overtaken it, its ongoing slag reaction churning out enough binding gel to fill pores faster than any filler could. In the high-replacement blends, pore refinement with age was visible but limited, because the reactive slag reservoir had simply been drawn down too far. Densification, in other words, is ultimately paid for in reactive calcium, silicon and aluminum, and the carbonated waste can only subsidize that budget, not replace it.

The study is candid about its limits. The experiments were run on pastes, so aggregates, interfacial transition zones, workability and long-term durability in real concrete remain untested, and the indicative carbon accounting, which weighs roughly 0.052 kilograms of CO2-equivalent per kilogram of slag against 130 kilograms of CO2 fixed per ton of carbonated waste, deliberately excludes the energy costs of pressurized carbonation, drying, gas compression and activator production. A full life-cycle assessment is still needed before the approach can claim genuine climate benefits. Even so, the core finding stands: at moderate replacement levels, a CO2-sequestering waste powder can shoulder a quarter of the binder workload in a slag geopolymer while preserving most of its strength. In an industry staring down carbon border taxes and trillion-dollar decarbonization bills, the idea that yesterday’s concrete sludge, inflated with captured carbon, could help build tomorrow’s foundations is exactly the kind of circular-economy alchemy the built environment needs.

Subject of Research: Use of carbonated concrete slurry waste as a partial precursor replacement in GGBS-based geopolymer pastes

Article Title: Effect of carbonated concrete slurry waste incorporation on the geopolymerization, mechanical performance, and microstructural evolution of GGBS-based geopolymer pastes

Article References: Sim, S. (2026). Effect of carbonated concrete slurry waste incorporation on the geopolymerization, mechanical performance, and microstructural evolution of GGBS-based geopolymer pastes. Case Studies in Construction Materials, 25, Article e06608. https://doi.org/10.1016/j.cscm.2026.e06608

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06608

Keywords: geopolymer, GGBS, concrete slurry waste, carbonation, CO2 sequestration, C-(N)-A-S-H gel, compressive strength, calcite, FT-IR spectroscopy, microstructure, sustainable construction, alkali-activated materials