Oven Curing Beats Steam for Fly Ash Geopolymer Mortar Strength, Study Finds

A single change in how mortar is cured can swing its strength by more than 20 percent, according to a new laboratory study of fly ash geopolymer mortars. Researchers prepared three mortar systems with the same binder-to-sand ratio of 1:1: an oven-cured geopolymer (GP-OV), a steam-cured geopolymer (GP-ST), and a water-cured ordinary Portland cement control (OPC-WC). Both geopolymer batches used low-calcium Class F fly ash activated by a 12 M sodium hydroxide solution blended with sodium silicate at a mass ratio of 2.0, with an activator-to-binder ratio of 0.40. The only variable within the geopolymer series was the curing regime: dry heat in a laboratory oven at 60 °C for 24 hours, or saturated steam produced in a self-fabricated arrangement inside the same oven at the same nominal temperature. After curing, all specimens were stored at ambient laboratory conditions until testing at 7 and 28 days.

The headline result concerns early-age strength. The oven-cured geopolymer developed 30.52 MPa at 7 days, exceeding the water-cured OPC mortar by 20.4 percent and the steam-cured geopolymer by 22.9 percent. This advantage reflects the well-documented role of moderate heat curing in accelerating geopolymerization in low-calcium fly ash systems, where dissolution of amorphous silica and alumina in concentrated alkaline solution and subsequent polycondensation require substantial activation energy. By 28 days, however, the picture narrowed: the oven-cured geopolymer reached 32.41 MPa while the OPC mortar reached 33.25 MPa, a difference of only 0.84 MPa that fell within the combined standard deviations of the two mixes. The steam-cured geopolymer lagged throughout, attaining just 24.84 MPa at 7 days and 28.03 MPa at 28 days.

The authors were careful about what the 28-day comparison does and does not show. Tukey’s HSD post-hoc testing found the difference between the oven-cured geopolymer and OPC not statistically significant at 28 days (p = 0.081), but they explicitly warn against reading this as statistical equivalence. With only three specimens per group drawn from a single batch, the test has limited power, and a proper equivalence claim would require a predefined margin and a formal equivalence test such as TOST, which was not performed. The large observed effect size (Cohen’s d = −1.98) indicates that a modest true difference cannot be excluded. By contrast, the steam-cured mix was consistently and significantly inferior to both other systems at 28 days (p < 0.001 against OPC and p = 0.001 against the oven-cured geopolymer). At 7 days, all pairwise contrasts were significant, with GP-OV exceeding both OPC and GP-ST at p < 0.001.

A two-way analysis of variance treating curing regime and age as fixed factors quantified how strongly each shaped the strength outcome. Both main effects were large: curing regime accounted for η² = 0.412 and age for η² = 0.422, both well above the conventional threshold of 0.14 for a large effect, and the regime-by-age interaction was also significant (η² = 0.147, p < 0.001). This interaction captures the divergent trajectories visible in the raw data: the oven-cured geopolymer gained most of its strength during the initial week, while the OPC mortar showed more pronounced continued hydration between 7 and 28 days. The authors stress that these inferential results rest on within-batch replicates rather than independently produced batches, and should therefore be interpreted as exploratory, reflecting variability within the investigated batches and only limited representation of between-batch variability.

Transport-related properties told an equally striking story. Water absorption, measured by oven drying to constant mass followed by 24-hour immersion in line with ASTM C642, serves in this study as an indicator of permeable and connected porosity rather than broader durability. At 7 days, OPC absorbed 9.94 percent, the oven-cured geopolymer 10.43 percent, and the steam-cured geopolymer 14.61 percent. By 28 days, absorption fell to 8.58 percent for OPC and 10.10 percent for the oven-cured geopolymer, but rose to 16.37 percent for the steam-cured mix, a 12.0 percent increase over its own 7-day value and a striking 90.9 percent higher absorption than OPC. One-way ANOVA at 28 days confirmed highly significant differences among the mixes (F2,6 = 322.31, p < 0.001), with all pairwise contrasts significant.

Bulk density followed the inverse pattern. OPC achieved the highest densities at both ages, reaching 2203.89 kg/m³ at 28 days, while the oven-cured and steam-cured geopolymers recorded 1884.27 kg/m³ and 1850.78 kg/m³ respectively, roughly 14.5 and 16.0 percent lighter than the control. All three mixes showed small density gains between 7 and 28 days, consistent with gradual hydration or geopolymerization refining pores rather than changing overall packing. Post-hoc testing showed OPC differed significantly from both geopolymer systems, while the two geopolymers were statistically indistinguishable from each other (p = 0.208). The authors offer the interpretation that oven curing yields a more efficient pore structure than steam curing only as a hypothesis, since no pore-size measurements such as mercury intrusion porosimetry were performed.

Why would moist heat underperform dry heat at the same nominal temperature? The authors draw on literature suggesting that in fly ash geopolymer systems cured under humid or steam conditions, the initially formed gel can undergo continued dissolution-reprecipitation in the presence of excess moisture, producing a coarser and less homogeneous pore network over time. Thermal gradients and surface moisture concentration during steam curing may also promote microcracks upon drying and cooling, which remain open during subsequent ambient storage. Importantly, the steam-curing environment in this study was characterized only nominally: no continuous temperature logging, relative humidity measurement, or spatial uniformity mapping was carried out for the self-fabricated setup. The authors acknowledge that the observed differences between GP-OV and GP-ST therefore cannot be uniquely attributed to a single curing variable, since variations in moisture history may also have contributed. Mechanistic explanations remain interpretive, pending microstructural work using SEM, XRD, FTIR, or MIP.

Regression analysis added quantitative structure to the dataset. A multivariate linear model expressing compressive strength as a function of bulk density, water absorption, and age yielded fc = 52.12 − 0.0076ρ − 1.024Wab + 0.203t, with R² = 0.920 and adjusted R² = 0.903. At 28 days, a strong inverse linear relationship emerged between strength and water absorption: fc = 39.46 − 0.737Wab, with R² = 0.970 and adjusted R² = 0.966. Yet the authors repeatedly caution that these equations are empirical correlations specific to the investigated systems, not validated predictive models. The dataset comprised only 18 specimen-level observations across six mix-age groups, and the 28-day bivariate analysis rested on just three independent material groups, making cross-validation impossible. A weaker strength-density relationship at 28 days (R² = 0.506) indicated that density was a less discriminating descriptor than absorption. Pearson correlations across the full dataset showed strength moderately negatively correlated with absorption (r = −0.640) and age moderately positively correlated with strength (r = +0.650).

Failure patterns differed qualitatively. OPC cubes failed suddenly and in a brittle manner, detaching large fragments, whereas both geopolymer mixes failed more gradually, with peripheral material spalling progressively while the core continued to sustain load. The authors confine this observation to visible morphology: without fracture-energy, toughness, acoustic-emission, or post-peak load-displacement measurements, no inference about damage tolerance is drawn. On sustainability, the findings are similarly bounded. Published life-cycle assessments estimate that replacing OPC with fly ash geopolymers can cut embodied CO2 by roughly 25 to 70 percent, but the 12 M NaOH and sodium silicate activators carry a non-trivial carbon burden, and 24 hours of oven heating demands external thermal energy that could substantially reduce, or under carbon-intensive grids negate, the net benefit. No life-cycle assessment was performed in this study, so the environmental case remains prospective.

The scope of the work is deliberately narrow: one fly ash source from the NTPC Kahalgaon plant in India, one activator composition, one binder-to-sand ratio, two curing regimes, and three specimens per condition from single batches. The authors conclude that, for this specific mix design, oven curing at 60 °C offers a more favourable balance between mechanical performance and transport resistance than steam curing at the same nominal temperature, supporting the technical feasibility of oven-cured Class F fly ash geopolymer mortars as potentially lower-carbon alternatives for precast and controlled-curing applications. Future work they identify includes independently replicated batches with logged temperature and humidity histories, formal equivalence testing, microstructural characterization, quantitative fracture measurement, extended durability testing covering freeze-thaw, carbonation, chloride and sulfate exposure, and a life-cycle assessment incorporating industrial waste heat for curing energy. The regression models, they add, require validation against larger, multi-source datasets before any generalization across different fly ash chemistries and activator formulations.

Subject of Research: Effects of oven and steam curing on the mechanical and transport properties of Class F fly ash geopolymer mortar compared with ordinary Portland cement mortar

Article Title: Effect of curing regimes on fly ash geopolymer mortars: a statistical and regression analysis toward sustainable development

Article References: Neupane, U., Sharma, R., Shrestha, S., Gurung, B., Poudel, S., & Thapa, I. (2026). Effect of curing regimes on fly ash geopolymer mortars: a statistical and regression analysis toward sustainable development. Discover Civil Engineering, 3(1), 271. https://doi.org/10.1007/s44290-026-00668-8

Image Credits: AI Generated

DOI: 10.1007/s44290-026-00668-8

Keywords: fly ash, geopolymer mortar, curing regime, compressive strength, water absorption, bulk density, ordinary Portland cement, ANOVA, regression analysis, sustainable construction, precast applications, low-carbon binders