The material at the center of the study, rubberized alkali-activated concrete, or RuAAC, combines two waste streams in a single matrix. The binder forms through a chemical reaction between an alkaline activator solution and aluminosilicate precursors, avoiding the high-temperature clinker process entirely. In this work, the binder comprised ground granulated blast furnace slag and Class F fly ash in a one-to-three ratio, activated by a blend of sodium silicate and 12 molar sodium hydroxide. The team tested three rubber gradations, a fine-graded fraction spanning 0 to 0.8 millimeters, a medium-graded fraction of 1 to 3 millimeters, and a wide-range blend covering 0 to 3 millimeters, each incorporated at 5, 10, and 20 percent replacement of fine aggregate by volume. Crucially, the rubber was used exactly as received from the tire recycling plant, with no chemical pre-treatment, a deliberate choice to preserve the environmental and economic case for the material, since soaking and rinsing procedures consume reagents, water, and generate effluent of their own.

The results on strength tell a story of trade-offs. The control mix, with no rubber, reached an impressive 72.3 megapascals at 28 days, reflecting the rapid reaction kinetics of the slag-rich binder. Adding rubber reduced compressive strength across the board, with losses ranging from 21 to 35 percent at 5 percent replacement, 30 to 46 percent at 10 percent, and 36 to 56 percent at 20 percent. But the gradation made an enormous difference. The fine-graded mixes retained the highest strengths at every dosage, reaching 57.1 megapascals at just 5 percent rubber, while the wide-range mixes performed worst, falling to 31.7 megapascals at 20 percent. The reason lies at the microscopic interface between the hydrophobic rubber particles and the hydrophilic alkali-activated gel. Residual zinc stearate from the tire vulcanization process repels the polar activator, entraining air voids at the interface and creating weak interfacial transition zones that act as flaws under load. Larger particles create larger, more critical defects, while sub-millimeter particles distribute the damage into smaller, less consequential discontinuities.

One of the most intriguing findings concerned early-age behavior. At just one day, the rubberized mixes were 40 to 72 percent weaker than the control, but this deficit narrowed substantially by 28 days. The rubber particles, it turns out, delay rather than suppress the alkali-activation reaction. The fine- and medium-graded mixes more than doubled their strength between one and seven days, recovering much of the early loss as the matrix densified around the inclusions. Statistical analysis using two-way analysis of variance confirmed this interpretation: the interaction between rubber content and gradation was significant at one day but not at later ages, meaning the gradation-specific early damage faded as the reaction progressed. By contrast, the medium-graded mix at 20 percent replacement gained almost no strength after seven days, suggesting that failure in that mix was governed by the weak interfaces around its uniform 1 to 3 millimeter particles rather than by the still-reacting matrix.

Where rubber genuinely shines is in deformation capacity. The control concrete failed in classic brittle fashion, with a sharp post-peak drop in the stress-strain curve. The rubberized mixes, by contrast, softened gradually and kept carrying load at strains approaching 0.8 percent or more. The ductility index, a measure of how much a material can deform before failure, rose by 44 percent in the fine-graded mix at 20 percent replacement compared with the control. The elastic rubber particles store strain energy and delay the coalescence of microcracks into macrocracks, transforming a sudden fracture into a controlled, gradual failure. This exchange of strength for toughness is precisely the property profile needed for applications such as crash barriers, pavement slabs, and industrial floors, where energy absorption and crack resistance matter as much as raw load capacity.

The study also broke new ground by measuring direct shear strength, load-slip response, and shear toughness of RuAAC for the first time at the material level, a gap with real safety implications. Shear failures in structural elements such as flat slabs, corbels, and deep beams occur with little warning, and design codes estimate shear capacity as a function of the square root of compressive strength using coefficients calibrated to Portland cement concrete. The team found that shear strength declined by up to 45 percent with rubber incorporation, less severely than flexural strength, which fell by as much as 67 percent, because confining stresses developed during shear testing provided mechanical interlock that partially compensated for the weak rubber-matrix bonding. Notably, at 5 percent fine-graded rubber, the shear toughness index was 25 percent higher than the control, suggesting that modest rubber dosages can genuinely improve energy dissipation along crack planes.

The durability results delivered some of the study’s most surprising wins. Water absorption in the fine-graded mix at 20 percent replacement was 9 percent lower than the control, and its capillary absorption depth at six hours was 39 percent lower, because the densely dispersed hydrophobic particles interrupt continuous capillary pathways and force water to take longer, more tortuous routes through the matrix. Bulk electrical resistivity, a proxy for resistance to chloride ingress and rebar corrosion, was 12 percent higher than the control in the same mix. Abrasion resistance told a dramatic gradation-dependent story: the fine-graded mix at 20 percent lost 29 percent less mass than the control after 500 tumbling cycles, while every medium-graded mix disintegrated completely, with the 20 percent medium-graded specimen failing by just 200 cycles. The uniform 1 to 3 millimeter particles, poorly bonded to the matrix, acted as discrete weak points that accelerated crack propagation under surface wear.

The authors distilled their findings into two practical application regimes, both built on fine-graded rubber. A mix with 5 percent fine crumb rubber retained 57.1 megapascals of compressive strength, a 25 percent higher shear toughness index, and better abrasion resistance than the control, making it suitable for pavement slabs, industrial floors, curbs, and precast panels. A mix with 20 percent fine rubber offers lower strength, 46.1 megapascals, but a 44 percent higher ductility index, reduced density, and improved workability, pointing toward crash barriers and non-structural cladding. The researchers caution that the reduced elastic modulus, between 14.2 and 19.2 gigapascals compared with 24.3 for the control, means deflection rather than strength will likely govern member sizing, and that shear-governed elements require member-scale testing before RuAAC can be deployed in them.

What makes this work resonate beyond the laboratory is its reframing of a familiar sustainability equation. Simply dumping waste into concrete is not enough; the form of the waste determines whether the resulting material is an engineering asset or a liability. By demonstrating that sub-millimeter tire rubber can simultaneously cut cement demand, divert landfill waste, improve ductility, and even enhance durability, the study offers a credible pathway for rubberized cement-free concrete to move from non-structural novelty toward genuine structural relevance. The authors call for microstructural examination of the rubber-matrix interface, extended durability testing covering chloride penetration, carbonation, freeze-thaw and creep, full-scale element testing, and a life cycle assessment quantifying embodied carbon per unit of strength. Until then, the message is clear: if you want tires to strengthen the future of concrete, grind them fine.

Subject of Research: Effect of recycled crumb rubber particle size and replacement level on the mechanical and durability properties of cement-free alkali-activated concrete

Article Title: Effect of recycled rubber particle size and replacement level on the properties of rubberized alkali-activated concrete

Article References: Elbaz, Y., Mwafy, A., El-Maaddawy, T., & El-Hassan, H. (2026). Effect of recycled rubber particle size and replacement level on the properties of rubberized alkali-activated concrete. Results in Engineering, 32, Article 113326. https://doi.org/10.1016/j.rineng.2026.113326

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.113326

Keywords: alkali-activated concrete, crumb rubber, waste tires, geopolymer, compressive strength, shear strength, ductility, durability, water absorption, abrasion resistance, sustainable construction, fly ash slag binder

Researchers Anuoluwapo Sola Kolade, Damilola Oyewumi Oyejobi and Bolanle Deborah Ikotun assembled a database of 40 concrete mixes drawn from 20 peer-reviewed experimental studies published between 2016 and 2026, following a PRISMA-inspired screening procedure that began with 65 records. Twenty of the mixes relied on commercial sodium silicate activation, while twenty used activators synthesized from waste materials, most commonly waste glass powder and rice husk ash. The team deliberately distinguished between sole waste-derived activator systems, which eliminate commercial alkalis entirely, and hybrid systems that supplement the waste-derived activator with additional sodium hydroxide. Every mix was rebuilt from its published proportions into a harmonized cradle-to-gate life cycle inventory, using standardized environmental coefficients from the Inventory of Carbon and Energy database and process-specific literature, so that the environmental comparison rested on consistent assumptions rather than the heterogeneous reporting of the original studies.

The feedstocks at the heart of the analysis are abundant. The world generates an estimated 100 to 130 million tonnes of waste glass annually, while roughly 190 million tonnes of rice husks are produced each year, corresponding to a potential rice husk ash yield of 26 to 34 million tonnes. Both materials can contain more than 70 percent amorphous silica, the reactive ingredient needed to build soluble silicate activators. Two synthesis routes were compared: hydrothermal dissolution of the silica-rich waste in alkaline solution at 80 to 100 degrees Celsius, and thermochemical fusion, in which the waste is heated with solid alkalis at 150 to 650 degrees Celsius to form reactive solid silicate powders suitable for one-part, mix-and-go concretes. The thermochemical route proved especially attractive, since a standardized 500-degree-Celsius furnace cycle for two or three hours consumed only 0.072 to 0.108 megajoules per kilogram of product, roughly six times less energy than the hydrothermal coefficients adopted from the literature.

The mechanical verdict was strikingly even. Commercial-activated mixes averaged 41.56 megapascals of 28-day compressive strength across the pooled database, while the waste-derived systems averaged 39.44 megapascals, a difference the researchers’ statistical tests found insignificant, with a small effect size. Within the structural-grade class of 40 to 60 megapascals, the sole waste-derived systems actually outperformed both alternatives, averaging 52.87 megapascals against 50.32 for the commercial controls and 44.34 for the hybrids. The best waste-glass thermochemical formulations reached 59.12 megapascals, and several rice husk ash systems exceeded 55 megapascals, all without any heat curing. The authors attribute this performance to the higher slag contents these formulations carried, which supply calcium-rich phases that form dense C-(A)-S-H-type reaction products even at ambient temperature.

On carbon, the waste-derived systems delivered a decisive win. Average cradle-to-gate global warming potential fell from 173.42 kilograms of carbon dioxide equivalent per cubic metre for commercial activation to 114.82 for the waste-derived mixes, a reduction of about 34 percent that was statistically significant with a large effect size. The 95 percent confidence interval placed the true reduction between 19 and 49 percent. Sole waste-derived systems did best of all, averaging 103.50 kilograms of carbon dioxide equivalent per cubic metre, because hybrid systems still carry the burden of chlor-alkali sodium hydroxide production, an electrochemical process whose energy demand can exceed 20 megajoules per kilogram. Notably, the embodied energy picture was less flattering: waste-derived mixes actually consumed slightly more energy on average than the commercial controls, a difference that fell just short of statistical significance, reminding engineers that carbon savings and energy savings do not automatically travel together.

Contribution analysis made the reason unmistakable. Activator production dominated the environmental profile of every category, accounting for 82 percent of global warming potential in commercial systems and still 75 to 76 percent in the waste-derived ones, and for 65 to 73 percent of embodied energy. Precursors, particularly slag, formed the consistent secondary hotspot, while aggregates, water, superplasticizer and curing each contributed less than 13 percent individually. In other words, the lever that matters most in low-carbon alkali-activated concrete is not the aggregate supply or the mixing water but the chemistry of the activator itself, which is precisely where waste valorization intervenes.

The robustness of these conclusions was tested through sensitivity analyses that perturbed the most uncertain assumptions. Raising the environmental coefficients assigned to waste glass and rice husk ash by 20 percent shifted the sole waste-derived average by only about 2 percent and changed no rankings. Substituting an alternative, lower-impact inventory for commercial sodium silicate reduced the commercial systems’ average carbon footprint substantially, yet the waste-derived systems still held the lowest position. Cutting the hydrothermal processing energy by 20 percent likewise left the ordering intact. Formal t-tests at the 95 percent confidence level, with Welch’s correction applied where variances were unequal, supported the headline comparisons throughout.

When strength was folded into the environmental accounting, the picture sharpened further. Strength-normalized carbon intensity averaged 3.23 kilograms of carbon dioxide equivalent per megapascal for waste-derived systems against 4.60 for the commercial controls, a significant difference with a moderate effect size. A composite sustainability index, which combines carbon and energy burdens per unit of strength, placed seven of the top ten most eco-efficient mixes in the waste-derived category, with the leading formulation achieving 50 megapascals at an index of just 2.23. Pareto analysis, which searches for mixes that no other formulation can beat on both strength and carbon simultaneously, identified three non-dominated systems, all of them sole waste-derived and all activated by thermochemically fused waste glass. One achieved 35 megapascals at a mere 51.52 kilograms of carbon dioxide equivalent per cubic metre, the lowest-carbon solution in the entire database, while another delivered 59.12 megapascals at 85.39.

Perhaps most consequential for real-world procurement, the researchers mapped their reconstructed carbon values onto the Low Carbon Concrete Group’s tiered embodied carbon classification, a framework increasingly used in infrastructure purchasing. Six of the eleven sole waste-derived formulations earned the top Market Beating rating, compared with only three of the twenty commercial mixes, and several waste-derived structural-grade concretes exceeded 50 megapascals while staying within that best band. The commercial systems, by contrast, spread across the ratings and included the only mixes classified in the worst band. The authors caution that these classifications were applied without regional adjustment for electricity mixes or supply chains, and that the cradle-to-gate boundary excludes transport, construction and end-of-life stages, so the results should be read as production-stage benchmarks rather than complete life cycle verdicts.

The path from laboratory promise to industrial deployment still runs through familiar obstacles: feedstock variability, contamination in waste glass streams, the regional dependence of rice husk ash quality, the need for dedicated thermal processing infrastructure, and the absence of the standards and mix-design frameworks that Portland cement enjoys. The study’s authors argue that future work should integrate durability indicators such as chloride migration and carbonation resistance into eco-efficiency assessments, standardize environmental reporting, and explore low-energy synthesis routes including mechanochemical processing. But the central message stands on firm quantitative ground: concrete that matches the strength of commercial formulations while cutting embodied carbon by roughly a third can be built on activators made from the glass bottles and rice husks the world already throws away, and the bottleneck to greener construction may be solved not by reinventing the binder but by reinventing its chemistry.

Subject of Research: Waste-derived alkaline activators for fly ash-slag alkali-activated concrete and their strength and life cycle environmental benchmarking

Article Title: Strength and environmental benchmarking of waste-derived activators in fly ash-slag alkali-activated concrete: A systematic and quantitative assessment

Article References: Kolade, A. S., Oyejobi, D. O., & Ikotun, B. D. (2026). Strength and environmental benchmarking of waste-derived activators in fly ash-slag alkali-activated concrete: A systematic and quantitative assessment. Case Studies in Construction Materials, 25, Article e06602. https://doi.org/10.1016/j.cscm.2026.e06602

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06602

Keywords: alkali-activated concrete, waste-derived activators, sodium silicate, waste glass powder, rice husk ash, fly ash, ground granulated blast furnace slag, life cycle assessment, global warming potential, embodied carbon, geopolymer, low-carbon procurement

The material at the heart of the research is ultra-high-performance alkali-activated concrete, or UHP-AAC. Conventional ultra-high-performance concrete achieves compressive strengths of 120 megapascals or more by packing cement, fine particles, and fibers into an extraordinarily dense matrix, but it typically requires 900 to 1100 kilograms of cement per cubic meter. UHP-AAC replaces that cement entirely with industrial by-products, chiefly ground granulated blast furnace slag and silica fume, which are activated by an alkaline compound rather than by clinker hydration. Previous work has shown that this substitution can cut carbon dioxide emissions by up to 70 percent and production costs by up to 64 percent while still reaching strengths above 130 megapascals.

Designing such mixtures, however, is notoriously difficult. The compressive strength of UHP-AAC depends on a web of nonlinear, interacting variables: the balance between slag and silica fume, the dose of solid alkaline activator, the amount of recycled steel fiber, the water content, the curing regime, and the age of the specimen. Traditional mix design relies on exhaustive laboratory trial batches, each consuming time, labor, and money. The research team, led by K. Behfarnia and A. Ghasemi Marnani at Isfahan University of Technology in Iran, set out to replace much of that experimental burden with data-driven prediction.

What distinguishes this study from earlier machine learning efforts in concrete research is the quality and scale of its dataset. Rather than pooling results from many laboratories, where differences in materials, curing, and testing procedures can muddy the signal, the authors generated 2048 compressive strength records from 256 unique mixture designs in their own laboratory. Each mixture was tested under two curing conditions, wrapped and ambient, at four ages: 7, 14, 28, and 56 days. Seven input features described every record, including curing method, slag content, silica fume content, sodium metasilicate pentahydrate as the solid activator, recycled steel fiber content, water content, and curing age. Measured strengths ranged from about 27.6 to 137.5 megapascals, a wide and well-distributed span that is ideal for training robust models.

The team benchmarked six algorithms: linear regression, decision tree, random forest, k-nearest neighbors, gradient boosting, and extreme gradient boosting. Each model was evaluated with unusual statistical rigor. Because multiple measurements came from the same mixture, the researchers grouped all data from an identical mix design together during train-test splitting, preventing information leakage between related records. They repeated the entire modeling pipeline ten times with different group-based splits, used five-fold group cross-validation during training, and tuned hyperparameters through grid and randomized searches. This design ensured that models were always judged on genuinely unseen mixtures, the fairest possible test of generalization.

The results were striking. Gradient boosting emerged as the clear winner, achieving a coefficient of determination of 99.93 percent on the unseen test set with a mean absolute error of just 0.406 megapascals, meaning its typical prediction missed the measured strength by less than half a megapascal. Extreme gradient boosting followed at 99.83 percent, then random forest at 98.6 percent, decision tree at 97.1 percent, k-nearest neighbors at 95.9 percent, and linear regression at 90.2 percent. Every prediction from the top three models fell within 20 percent of the observed value. Residual analysis across low, medium, and high strength ranges confirmed that the boosting models remained accurate whether the concrete was expected to reach 40 or 130 megapascals.

The gap between simple and ensemble methods carries a technical lesson. Linear regression assumes additive, linear effects, but the study’s own correlation analysis showed that all relationships between the input variables and strength were nonlinear, with correlation coefficients below 0.7. Boosting algorithms, which build decision trees sequentially with each new tree correcting the errors of its predecessors, excel precisely at capturing such interactions. Random forest, which averages many independently grown trees, reduces variance but lacks the sequential error correction that gives boosting its edge. Single decision trees and k-nearest neighbors, meanwhile, fit training data almost perfectly but generalize less well to new mixtures.

Beyond raw accuracy, the researchers applied SHAP analysis, a technique borrowed from game theory that quantifies each feature’s contribution to every individual prediction. The results were physically coherent and, in places, surprising. Curing method proved the single most influential variable: wrapped curing, which seals specimens in plastic sheeting to preserve moisture, consistently outperformed ambient curing by sustaining the dissolution and gel-forming reactions that build strength. Steel fiber content ranked second, reflecting the fibers’ ability to bridge cracks and restrain lateral deformation under compression. Water content and curing age followed, with higher water diluting the alkaline activator and slowing gel formation, while longer curing allowed continued microstructural densification, especially before 28 days.

The finer details of the binder chemistry also emerged clearly from the model. Increasing slag content raised strength because slag is rich in calcium oxide, promoting the calcium-rich binding gels that give the matrix its density. Silica fume, by contrast, showed a negative effect in this design space: because it contains almost no calcium, substituting it for slag lowers the system’s calcium-to-silicon ratio and limits the formation of those critical binding phases. The optimal activator dose sat near 170 kilograms per cubic meter, and the best fiber contents clustered around 230 to 240 kilograms per cubic meter, roughly a 3 percent volume fraction.

The practical implications extend well beyond the laboratory. The authors position their gradient boosting model as a decision-support tool that engineers can use at the mix design stage to screen candidate formulations and estimate strength at different ages without extensive preliminary trials, dramatically accelerating the development of new cement-free mixtures. They caution, appropriately, that predictions are only reliable within the range of materials and proportions tested, and that laboratory verification remains essential for critical structural applications. Still, the combination of a large, internally consistent dataset, rigorous group-based validation, and interpretable machine learning offers a template for how artificial intelligence can help decarbonize one of humanity’s heaviest industries, one optimized mixture at a time.

Subject of Research: Machine learning prediction of compressive strength in ultra-high-performance alkali-activated concrete

Article Title: A comparative machine learning study for predicting the compressive strength of ultra-high-performance alkali-activated concrete

Article References: Behfarnia, K., & Marnani, A. G. (2026). A comparative machine learning study for predicting the compressive strength of ultra-high-performance alkali-activated concrete. Case Studies in Construction Materials, 25, Article e06577. https://doi.org/10.1016/j.cscm.2026.e06577

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06577

Keywords: machine learning, ultra-high-performance concrete, alkali-activated concrete, compressive strength, gradient boosting, SHAP, ground granulated blast furnace slag, silica fume, steel fibers, sustainable construction, cement replacement, mix design