The research, conducted by Fatih Kantarci and Moncef L. Nehdi, centers on geopolymer concrete, an alkali-activated alternative to Portland cement concrete that is synthesized from industrial by-products rich in aluminum and silicon, such as blast furnace slag, metakaolin, and fly ash. When these precursors are mixed with highly alkaline solutions like sodium hydroxide, a chemical process called geopolymerization forms three-dimensional Si-O-Al-O polymeric gels that bind aggregates into a solid mass. Depending on the precursor and activator chosen, geopolymer binders can cut CO2 emissions by up to 80 percent compared with Portland cement, while saving roughly 60 percent of the energy and reducing production costs by about 25 percent. Geopolymer concretes have already found their way into road pavements, precast elements, and fire-resistant construction in the United States, Australia, Europe, and India.

Like most cementitious materials, however, geopolymer concrete is inherently brittle and prone to cracking under moderate loads or shrinkage stresses. The established remedy is fiber reinforcement, which enhances crack resistance, tensile strength, ductility, and impact performance while redistributing stresses within the matrix. Among the many fiber types available, the researchers selected glass fiber for its affordability, ease of manufacture, corrosion resistance, and high tensile strength of 1300 megapascals. The glass fibers used in the study were just 6 millimeters long and 15 micrometers in diameter, with an elastic modulus of 72 gigapascals and, crucially, a melting point of approximately 850 degrees Celsius, meaning they retain structural stability throughout the temperature range examined.

The experimental program began with a careful optimization of the geopolymer mix itself. Blast furnace slag from a local plant, with a specific gravity of 2.84 and a cement-like fineness, served as the primary precursor at a dosage of 400 kilograms per cubic meter. The team varied the sodium hydroxide activator concentration across 10, 12, and 14 molar solutions and tested alkali activator solution-to-binder ratios of 0.50 and 0.60. Compressive strength measurements at 7, 28, and 90 days revealed a clear optimum: strength increased as the sodium hydroxide concentration rose to 12 molar, then declined at 14 molar. The researchers attribute the initial gain to higher alkalinity, which dissolves silicon and aluminum links in the raw precursor to form aluminosilicate gels, while the decline at 14 molar reflects inhibited condensation reactions of silicate species and the precipitation of geopolymer gels that ultimately weaken the matrix. Scanning electron microscopy confirmed the story, showing a dense, compact microstructure with low porosity in the strongest mixes and abundant large pores and cracks in the weakest.

With the optimum production parameters established at 12 molar sodium hydroxide and a 0.60 activator-to-binder ratio, the team incorporated glass fibers at volume fractions of 0.3, 0.6, and 0.9 percent. Notably, the concrete was cured entirely under ambient laboratory conditions at 23 degrees Celsius and 55 percent relative humidity, with no steam or heat curing, removing a major barrier to casting geopolymer concrete on real construction sites. After 90 days of curing, the specimens were exposed to temperatures of 150, 300, 450, 600, and 750 degrees Celsius for one hour in a furnace heated at roughly 2 degrees Celsius per minute, then cooled under two contrasting regimes: gradual air cooling inside the opened furnace, or rapid immersion in room-temperature water, simulating the thermal shock that firefighting operations inflict on burning structures.

The results reveal a nuanced interplay between fiber content, temperature, and cooling method. Glass fibers improved compressive strength in both heated and unheated specimens, with the optimum at 0.6 percent by volume. At this dosage, fibers wrapped in geopolymer gel bond strongly to the matrix, bridging cracks, reducing stress concentrations at crack tips, and retarding crack propagation. The residual compressive strength of the fiber-reinforced samples actually increased up to 150 or 300 degrees Celsius, a phenomenon attributed to polycondensation and further densification of the tetrahedral aluminosilicate gels as moisture evaporates, before declining at higher temperatures. Remarkably, after exposure to 750 degrees Celsius, the water-cooled specimen containing 0.6 percent glass fiber retained a compressive strength approximately 33 percent higher than the plain, fiber-free samples. Beyond 450 degrees Celsius, however, the mismatch in thermal expansion coefficients between glass fibers and the geopolymer matrix generated interfacial stresses and microcracks, while partial softening of the fibers, dehydration of the gels, and thermal phase transformations further eroded strength.

The cooling regime proved to be a decisive variable. Water-cooled samples consistently exhibited lower residual compressive and flexural strengths than their air-cooled counterparts, because the steep temperature gradients during rapid quenching induce thermal shock, microstructural damage, and an elevated risk of explosive spalling. Flexural strength, which is particularly sensitive to crack initiation and propagation, benefited even more visibly from fiber reinforcement, since the three-dimensionally dispersed fibers direct crack paths and transfer stresses through a bridging effect that preserves specimen integrity. At all temperatures, the 0.6 percent fiber content delivered the highest flexural values, and the relative improvement from fiber addition was more pronounced in flexure than in compression, underscoring the dominant role of crack bridging in bending behavior.

Complementary measurements of weight loss and water absorption traced the progressive thermal deterioration of the material. Weight losses remained modest at 150 and 300 degrees Celsius, driven by the evaporation of free and absorbed water, but increased sharply after 450 degrees Celsius as thermal stress generated microcracks, and again at 750 degrees Celsius, where thermo-chemical damage degraded the geopolymer gel itself. The fiber-free air-cooled specimen lost 1.3 percent of its mass at 150 degrees Celsius but 7.1 percent at 750 degrees Celsius, roughly a five-and-a-half-fold increase, while 0.3 and 0.6 percent fiber additions reduced these losses by preserving microstructural integrity. Water absorption told a parallel story: values stayed nearly unchanged up to 450 degrees Celsius thanks to the dense matrix, then climbed as thermally induced shrinkage and thermal-shock microcracking opened new transport pathways. The fiber-free water-cooled sample doubled its water absorption from 4.3 to 8.6 percent after exposure to 750 degrees Celsius. Interestingly, the highest fiber dosage of 0.9 percent proved counterproductive, increasing water absorption because of poor workability, uneven fiber dispersion, and fiber balling, a reminder that more fiber is not always better.

Visual and microstructural examinations completed the picture. Sample surfaces brightened to a light brown up to 600 degrees Celsius and darkened to brown-black at 750 degrees Celsius, a coloration attributed to the gehlenite phase identified by X-ray diffraction, which also detected calcium silicate, calcium oxide, akermanite, and ilvaite. Crucially, the glass fibers did not melt even at 750 degrees Celsius, and no specimen fragmented, chipped, or disintegrated under either cooling regime, although water-cooled samples displayed more surface cracks. Scanning electron microscopy after 750 degrees Celsius showed that the fibrous air-cooled sample retained a dense, compact microstructure, while the non-fibrous water-cooled sample exhibited large cracks and spherical pores, with fiber-matrix debonding and increased microcrack density explaining the measured strength losses.

The study’s conclusions carry practical weight for the construction industry’s decarbonization ambitions. An ambient-cured, slag-based geopolymer concrete reinforced with 0.6 percent glass fiber emerges as a promising candidate for fire-resilient structural applications, provided that workability and fiber dispersion are carefully controlled. The findings also deliver a clear warning for fire engineering: the way a structure cools after a fire matters nearly as much as the fire itself, with rapid water quenching inflicting measurable thermal-shock damage that gradual air cooling avoids. The authors point toward future research on long-term durability under diverse service environments and on extending the approach to other fiber types and cooling scenarios, steps that could help carry geopolymer composites from the laboratory into the load-bearing skeleton of sustainable cities.

Subject of Research: Elevated-temperature resistance of glass fiber-reinforced geopolymer concrete under different cooling regimes

Article Title: Elevated-temperature resistance of glass fiber-reinforced geopolymer concrete under different cooling regimes

Article References: Kantarci, F., & Nehdi, M. L. (2026). Elevated-temperature resistance of glass fiber-reinforced geopolymer concrete under different cooling regimes. Cleaner Engineering and Technology, 34, Article 101311. https://doi.org/10.1016/j.clet.2026.101311

Image Credits: AI Generated

DOI: 10.1016/j.clet.2026.101311

Keywords: geopolymer concrete, glass fiber, elevated temperature, fire resistance, cooling regime, thermal shock, blast furnace slag, compressive strength, flexural strength, sustainable construction, alkali-activated materials, microstructure