Engineering

Abstract: An acceptable analytic hierarchy process (AHP) consistency ratio measures the coherence of pairwise judgement, not the integrity of a geographic information system (GIS) product. This study audits three relationships in a relative landslide-susceptibility index for the 22.14-km Villa Verde Road corridor, Pangasinan, Philippines: raster-pair and partition consistency, regenerated-overlay encoding and coverage, and factor-score rule consistency. Comparison of the retained slope surface with the former slope-score raster identified 12 cells incorrectly assigned score 4 despite slopes of 25.16°–29.69°. Correcting them changed 12 index values and, with re-derived class breaks, 20 class labels (0.85%). A separate aspect comparison found 1,167 of 2,357 cells (49.51%) differed between the retained raw-derived surface and adopted V1 scores; the raw-derived alternative changed 307 cells under fixed thresholds and 301 after re-derived Jenks breaks. Weight sensitivity tests produced up to 533 changed cells in one-at-a-time scenarios, while mean Dirichlet agreement ranged from 75.43% to 87.27%; Very High-class Jaccard overlap was lower (0.60–0.77). These first-party checks and conditional sensitivity results do not constitute independent workflow, source-data or predictive validation. The derivative-processing cause and intended aspect convention remain unresolved; principal results are conditional on V1.

Abstract: Secondary compression of soils shows a strain roughly linear in log time over several decades, with a slope—the coefficient of secondary compression C_α—that the Mesri law ties to the compression index C_c. Fractional creep elements are widely used for soil and rock creep; we ask how well they, and classical spectral laws, hold that slope within a stated tolerance. No single Scott-Blair springpot or fractional Kelvin–Voigt element yields an exactly constant C_α: the springpot slope grows by 10^α per decade and the Kelvin–Voigt slope decays as t^−α. A uniform retardation spectrum on a finite band [τ₁, τ₂]—the Becker law with an upper cutoff—gives the closed form C_α(t) = σ₀ A ln10 (e^−t/τ₂ − e^−t/τ₁), so the plateau height and tolerance-defined width follow exactly; holding C_α within 1% over N decades needs a band about N + 2.66 decades wide. For the same target slope and four-decade window under one amplitude rule (peak slope = C∗), a near-degenerate springpot and the Becker law need unbounded creep, while a Kelvin–Voigt element of order α ≈ 0.044 needs a bounded long-time compliance about six times that of the finite band. The finite-band element is thermodynamically admissible as a positive superposition of Voigt modes. The results support choosing a creep kernel by the window it sustains and the long-term settlement it implies.

Abstract: Glass fibre reinforced polymer (GFRP) composites are being increasingly used as structural sections owing to their high strength-to-weight ratio and corrosion resistance, making accurate determination of their tensile properties essential for safe and reliable design. This study presents a comprehensive comparison of the tensile behaviour of pultruded GFRP composites using ASTM D3039 flat specimens and ASTM D638 dog-bone specimens. The effects of specimen geometry on tensile properties and failure characteristics were systematically investigated. Fibre characteristics were analysed using scanning electron microscopy (SEM), thermogravimetric analysis (TGA), and heating to extract the fibres, while digital image correlation (DIC) was employed to monitor full-field strain evolution during tensile loading. Compared to flat specimens, dog-bone specimens exhibited lower tensile strength and elastic modulus of 7-11% and 1-22%, respectively. More importantly, the failure mechanisms differ with flat specimens predominantly failing by longitudinal splitting accompanied by edge delamination, whereas dog-bone specimens exhibited transverse rupture. The results suggest that dog-bone specimens tested as per ASTM D638 would not accurately represent the intrinsic tensile behaviour of pultruded GFRP composites. Therefore, only ASTM D3039 is recommended for tensile characterization, to obtain a reliable basis for quality control and structural design with pultruded GFRP composites.

Abstract: This paper examines hotels and commercial buildings around the two Holy Mosques of Makkah and Madinah, which must be demolished in 10–20 years to make room for mosque precinct expansions, even though their reinforced-concrete (RC) frames are designed for 50 years or more. Instead of the current build–demolish–rebuild practice, the proposal is a set of identical, fully demountable steel buildings that can be unbolted, moved, and re-erected on a new site when a precinct is cleared. A reference module (18 storeys and 63.2 m above a permanent RC basement “dock”, on a 3 × 4 grid of 8 m bays) is derived from the stock of recent central-area hotels and analysed in COMSOL Multiphysics as a three-dimensional beam finite-element model with floor diaphragms, S355 sections and bolted connections, under gravity, SBC 301 wind and equivalent-lateral-force seismic actions and in free vibration. It is found that the bare moment-resisting frame satisfies the strength (utilisation 0.80) and seismic drift (0.0076) requirements but not the wind serviceability limit (inter-storey drift 1/209 > 1/400). Demountable CHS X-bracing in the two end bays of each long face reduces the roof sway under wind from 209 mm to 83 mm and the governing drift to 1/654. Over 60 years with three occupations, building once and relocating twice requires about one-third of the structural embodied carbon (195 against 570 kgCO₂e/m²) and about 55% of the structural cost of the RC practice. Design guidelines covering standardisation, the RC dock interface, connection detailing, tolerances, transport and inspection before reuse are formulated for the engineers and authorities involved in the development of the Haramain central areas.

Abstract: Pressurized irrigation networks have improved water use efficiency, but deterministic sizing does not anticipate off-design performance. The pressure reliability index (IF), Todini resilience index (IR), and supplied demand ratio (IQ) were evaluated in two gravity-fed branched networks, Tuncarta (Ecuador) and ARJ-42 (Spain), combining operation (on-demand and rotation), modeling approaches (demand-driven analysis, DDA, and pressure-driven analysis, PDA), and two uncertainty scenarios: stress testing and Monte Carlo simulation. Rotation operation was fully reliable under design conditions but lost reliability at over-demands of 5% (ARJ-42) and 10% (Tuncarta). Under random opening, median reliability was 0.90 in Tuncarta (one-third simultaneity, exceeding the design value) and 1.00 in ARJ-42 (design simultaneity). Under pressure deficits, DDA overestimated delivered flow by up to 45% (Tuncarta) and 83% (ARJ-42) and produced negative resilience values without physical meaning. Both networks exhibited three deterioration zones and, with all hydrants open, supplied 69% and 55% of their reference flow rates. The framework supports design validation and expansion planning.

Abstract: Collapse pits and sinkholes are among the common hidden defects in levee engineering, characterized by high suddenness and difficulty in early identification during their formation process. Based on levee model tests, this paper adopts a method of embedding ice bodies combined with internal heating and melting to simulate the evolution process in which the internal support structure is gradually lost and a cavity is formed. Continuously acquired microtremor signals are collected by deployed high-sensitivity acceleration sensors, and time-frequency energy spectrum features are extracted using wavelet transform. The results show that after the cavity is formed, the energy concentration zone of the acceleration wavelet energy spectrum exhibits a significant low-frequency shift, with an energy frequency shift of 6 Hz (from 8 Hz to 2 Hz). At this point, the ice body has completely melted and the support loss rate reaches 100%, while the ground surface remains continuous without macroscopic collapse, indicating that this method can capture early warning signals before disaster occurrence. On this basis, a risk evaluation index system for collapse pits and sinkholes covering four indicators across three dimensions of "load conditions–dynamic fingerprints–deformation response" is constructed. A game-theoretic combination weighting method combining the analytic hierarchy process and the entropy weight method is adopted to determine indicator weights, and quantitative risk level evaluation is achieved based on fuzzy comprehensive evaluation. Two typical stages, namely before heating and after cavity formation (before surface collapse), are taken as evaluation samples. The evaluation results show that the comprehensive score before heating is 95.0 (Level I, safe), and decreases to 37.9 (Level V, high risk) after cavity formation. Sensitivity analysis shows that after removing the dynamic fingerprint indicators, the comprehensive score at the post-cavity-formation stage rises to 53.8, and the risk level decreases from "high risk" to "medium risk," with a relative increase of 41.9%, directly demonstrating that the dynamic fingerprints extracted based on wavelet energy spectrum are sensitive to the identification of internal cavities in levees, providing technical support for pre-disaster warning of collapse pits and sinkholes. This study provides an approach for risk analysis of levee collapse pits and sinkholes integrating measured indicators and fuzzy comprehensive evaluation.

Abstract: In an effort to create an environmentally concrete industry, "geopolymer concrete" has appeared as a less harmful alternative to "ordinary Portland concrete". In this paper, the Numerical research of flexural behavior of "geopolymer concrete" and "ordinary Portland concrete" beams reinforced with CFRP rebars by using the ANSYS program. The specimens have compressive strengths ranging from 20 to 50 MPa with varying reinforcement ratios. A numerical analysis was conducted using ANSYS to examine various parameters that impact the structural behavior of GPC beams, including longitudinal reinforcement, fiber addition, and opening inclusion. Test results revealed that GPC specimens with CFRP bars displayed higher deflection when contrasted with OPC specimens with CFRP bars. Research results suggest that the inclusion of "steel fibers (SF)" in geopolymer concrete led to improved resistance to load failure in specimens subjected to flexural stress. Moreover, utilizing CFRP bars for beam reinforcement exhibited superior resistance to bending moments compared to BFRP and GFRP bars. Furthermore, reinforcing "geopolymer concrete" beams with openings using FRP and steel bars significantly augmented the failure load of these beams.

Abstract: Accurate building footprint extraction from high-resolution aerial imagery is essential for urban planning, cadastral mapping, disaster management, and digital twin development. Existing deep learning approaches often improve individual architectural components in isolation or rely on increasingly complex designs, at the expense of efficiency and generalization. This study proposes AR2Net, an encoder–decoder architecture integrating residual learning, recurrent convolutional refinement, and channel–spatial attention within a unified framework to improve optimization stability, spatial feature refinement, and building-relevant feature selectivity. A two-stage training strategy is introduced on a large-scale international dataset with OpenStreetMap-derived annotations before fine-tuning on a high-quality regional dataset from Iranian cities. AR2Net is evaluated through ablation experiments and comparisons with five widely used segmentation models (U-Net, U-Net++, DeepLabV3+, HRNetV2, and SegFormer), achieving the best overall performance (IoU = 88.35%, F1-score = 93.31%). A geometric accuracy assessment against manually digitized ground truth across 486,000 buildings yielded a mean positional error of 0.92 m. The extracted footprints were further used to generate LoD1 CityGML building models, illustrating a practical pathway toward 3D cadastral map production. These results demonstrate that AR2Net provides an accurate, geometrically reliable, and scalable solution for automated building footprint extraction.

Abstract: This study comparatively evaluated the mechanical performance of high-strength concrete incorporating locally sourced rice husk ash nano-silica and imported nano-silica as partial cement replacement. The research responds to the heavy reliance of nano-silica-modified concrete production on costly, industrially synthesized, imported material, and examines whether a locally producible alternative derived from an abundant agricultural waste can deliver comparable results. Rice husk ash nano-silica was produced through controlled combustion of rice husk followed by alkaline dissolution and acid precipitation, and its silica content, specific surface area, and specific gravity were determined alongside those of commercially sourced imported nano-silica. Twenty-seven high-strength concrete mixes were proportioned using a particle-packing-based mix design, incorporating each nano-silica type at replacement levels of 0 to 20 percent across three water-binder ratios, and evaluated for workability and for compressive, splitting tensile, and flexural strength at 7, 14, and 28 days. Both nano-silica types achieved an optimum replacement level of approximately 15 percent, beyond which strength declined. At this optimum, concrete containing imported nano-silica achieved a slightly higher 28-day compressive strength than concrete containing rice husk ash nano-silica, though statistical analysis found no significant interaction between nano-silica source and replacement level. The findings indicate that rice husk ash nano-silica is a mechanically viable substitute for imported nano-silica in high-strength concrete production, supporting its consideration as a lower-cost, locally sourced alternative, particularly in rice-producing regions.

Abstract: Closed-form solutions simplify the design of thick orthotropic plates, but their accuracy for plates with a free edge is unknown. This study developed one-term closed-form and multi-term series solutions of Alternative 1 Theory, with three-dimensional (3D) constitutive relations, for uniformly loaded rectangular orthotropic plates with a free edge. Ten support combinations, with the free edge parallel or perpendicular to the stiff direction, were analysed in 1,200 cases (E₁/E₂ = 5–40, a/t = 4–20, b/a = 0.5–3.0). A 3D Ritz solution, with deflections within 0.08% of exact, Lévy-type, and finite element solutions, served as the reference, and the series separated shape error from theory error. The converged series underestimated the 3D deflection by 4–19% at a/t = 4 and by up to 10% at a/t = 20. Its error was largest where the stiff direction spans between a clamped and a simply supported edge; there, it also overestimated the clamped-edge moment by up to 129%. Free-edge orientation governed the response. Plates with the free edge across the stiff direction deflected up to 1,050 times as much as their quarter-turned counterparts, and the one-term solution underestimated their deflection by 11–95% for b/a ≥ 1.0. The one-term formulas were within 10% of 3D elasticity in 319 cases, half of them for four plates with a/t ≤ 5 and b/a ≥ 1.0, partly through a cancellation of shape and theory errors. In thick plates, 3D face stresses at the free edge exceeded the linear values by up to 110%.

Abstract: Spatial estimates of reservoir evaporation from satellite observations are often inaccurate due to biases in raw satellite-derived water surface temperature (Ts_sat) and the need for concurrent in situ measurements of water surface temperature (Ts_m) for correction. This study expands the Turbulent Exchange Approach for Reservoir Evaporation Estimation (TEAREE) by developing transferable satellite-derived Ts_sat correction methods that eliminate the need for in situ Ts_m measurements at the application lakes. Analysis of 627 corresponding Ts_sat and Ts_m from 30 lakes across four climatic zones revealed climate-dependent biases, with a pooled mean bias of +1.53 °C and RMSE of 2.43 °C. Global (Method 1) and a climate-zone-based (Method 2) corrections were developed using 22 lakes and independently validated at eight lakes. Method 1 and Method 2 reduced satellite-derived Ts_sat RMSE from 2.92 °C to 2.15 °C and 1.87 °C, respectively. Incorporating these corrections into TEAREE reduced pooled daily evaporation RMSE from 1.37 mm/day to 0.81 mm/day (41%) and 0.62 mm/day (55%), respectively. Method 2 achieved comparable performance to the site-specific correction benchmark, with statistically equivalent mean evaporation estimates and absolute prediction errors at the pooled scale (TOST p < 0.001; ±5% equivalence margin). These findings demonstrate that climate-zone-based correction improves TEAREE evaporation estimates and supports broader application to lakes and reservoirs where in situ Ts_m measurements are unavailable.

Abstract: The safety of construction workers has been a major issue because of the substantial number of accidents and fatalities caused by ignoring hazards. Notwithstanding all these attempts, not all the approaches adopted by the industry have effectively achieved the required results. Some strategies could not make substantial progress or help tackle the unique issues of construction safety performance. This demonstrates the complexities of maintaining safety within construction sites and the need for more personalised and sturdy remedies that can successfully handle distinct threats. Further research is needed to incorporate teamwork between workers to bolster safety performance. In answer to this demand, the current research paper presents Teamwork based Construction Safety Performance Index Appraisal Kit (C-SPI-AK), which was created utilising the fuzzy synthetic evaluation method to evaluate the construction work practices in different construction working environments to ensure the well-being of workers and the efficacy of safety protocols. The tool will help evaluate six major components, safety culture indicators (Cat1), Safety rules and behaviour indicators (Cat2), Safety practices and engagement indicators (Cat3), Safety communication and inspection indicators (Cat4), Management commitment indicators (Cat5) and Safety meetings and maintenance indicators (Cat6). Furthermore, construction safety performance index appraisal kit will bridge the gap between the conceptual research analysis and the practical implementation resulting in enhanced safety standards.

Abstract: Classical plate theory (CPT)¹ is the default model for the bending of thin orthotropic plates, yet the span–thickness ratio beyond which it can be trusted is typically taken from isotropic experience. This study determines this ratio from three-dimensional (3D) elasticity for the 21 distinct orthotropic rectangular plates with clamped, simply supported, and at most one free edge, under uniformly distributed loading. A Ritz solution of 3D elasticity, validated against exact and finite-element solutions, was compared with CPT over 3,360 cases (E₁/E₂ = 5–40, aspect ratios 0.5–3, span–thickness ratios 4–100). The ratio needed for CPT to be within 5% of the 3D deflection ranges from 10 to 90, and the common threshold of 20 fails in 76% of the 420 combinations. For aspect ratios of 1.5 or more and the 15 plates whose fibres end on supported edges, the limit depends only on those two edges and scales with the square root of E₁/G₁₃. These limits follow a Timoshenko beam strip spanning the fibres, which is first-order shear deformation theory in cylindrical bending: the group averages agree within 0.3% and individual plates within 4.5% (9% in one case). This yields a closed-form rule: the square of the limit equals 0.96, 2.65, or 4.80 times E₁/(εG₁₃) for simply supported, propped, and clamped strips, where ε is the tolerance. The rule holds within 1.2% for graphite–epoxy, glass–epoxy, and glulam plates, and soft simple supports increase the limits by at most 3%. Plates with a free edge across the fibres and no clamped edge opposite it carry the load across the fibres and follow √(E₂/G₂₃) instead. Deflection governs the limit in 88% of cases; moments govern mainly when a free edge crosses the fibres.

Abstract: The flexural capacity of composite cold-formed steel (CFS) beams is often limited by premature instability of the thin-walled section, so that the compressive capacity of the slab is not fully used. This paper examines which measures remove that limit and when a higher concrete grade becomes worthwhile. A nonlinear finite element model, validated against three full-scale tests, was used in a parametric study of thirty beams. The variables were the slab cross-section, the CFS profile, the concrete grade, the type and spacing of plate shear connectors, and stiffening by web stiffeners and a lower-flange plate. Four failure modes were identified, and the toughness followed their hierarchy rather than the ranking of peak loads. At equal self-weight a hollow slab raised the peak load by 37%, and inverting the lips of the channel raised it by 78% while making the connector spacing almost irrelevant. Web stiffeners raised the toughness by up to 124%, against 27% for a lower-flange plate that gave the higher strength. Raising the grade from C30 to C60 was worth 6.6% in peak load while the section buckled locally but 12.9% once stiffened: high-performance concrete pays only after the instability of the steel section is suppressed.

Abstract: The aim of this paper is the analysis of arbitrarily loaded isotropic rectangular Kirchhoff plates supported at all corner points, whereby the edges are arbitrarily supported (simply supported, clamped, or free). Analytical methods commonly used include the single trigonometric series of Lévy, the double trigonometric series of Navier, the power series expansions, etc. In this paper the deflection surface was approximated with the sum of a particular solution to the governing differential equation (GDE) and two single series. The terms of the single series were the product of an unknown function of an independent variable and a trigonometric function of the other independent variable, whereby the trigonometric functions were chosen to vanish at the corner points. On the one hand the terms of the series were required to satisfy the homogeneous GDE, leading to two uncoupled differential equations, one for each unknown function, and so the approximate solution satisfied exactly the GDE. On the other hand the boundary conditions were satisfied only at selected collocation points along the boundary, the number of collocation points in each direction corresponding to the number of terms of the associated series. Results for several plates supported at all corner points with arbitrary edge support conditions were presented and showed good agreement with the exact results, the accuracy increasing the more terms of series were considered.

Abstract: Finite element models of ultra-high-performance concrete (UHPC) deep beams are usually judged by peak load, yet a correct capacity can come with a wrong mechanism. Our earlier shell-element study of these beams reported 104–107% of the measured capacity for normal- and high-strength concrete (NSC, HSC) and 63% for UHPC, ratios that would be 0.32–0.54 if its predictions are applied loads, with flexural failure where the tests failed in shear and a code-derived UHPC tensile strength 38% below the measured value. A 2 × 2 design—layered shell versus solid elements, each with a code-derived and a measured tensile strength—separates the two causes in four beams. The shell model failed in flexure and over-predicted capacity 1.25–2.86 times; its element effect (with elastic transverse shear), −0.61 to −2.06, far exceeded the tension effect, −0.05 to +0.09. The measured tensile strength raised the solid UHPC predictions by 11–25%, yet left them at 0.72–0.73, consistent with missing fibre bridging but not separable from the mesh dependence of the solid model. The solid model came within 5–21% of the capacity of the beams without fibres, yet failed by diagonal cracking where the tests failed in diagonal compression. Peak-load agreement does not validate a deep-beam model.

Abstract: The durability and safety of cathodic protection (CP) in reinforced and prestressed concrete depend on interactions among electrode materials, electrochemical conditions, and structural response. This critical review examines how anode composition, microstructure, conductive layers, concrete resistivity, and current distribution affect protection performance and material degradation. Laboratory studies, numerical simulations, and field evidence are compared for titanium/mixed-metal-oxide anodes, zinc-based galvanic systems, graphite and carbon-modified cementitious materials, polymeric coatings, and contact mortars. Particular attention is paid to interfacial processes, distinguishing initial conductivity and current delivery from electrochemical stability, anode consumption, and demonstrated service life. Accelerated aging and field exposure are treated as complementary, but non-equivalent, evidence of durability. For prestressing steel, the effects of cathodic polarization, strength, stress state, and exposure on hydrogen-assisted damage are examined; meeting protection criteria alone does not establish hydrogen safety. Instant-off potentials, depolarization measurements, and electrochemical impedance spectroscopy are assessed alongside material and structural diagnostics, including their limitations. A three-dimensional P–T–B framework distinguishes local protection performance (P), technological durability (T), and material and structural safety (B). Standardized characterization and long-term validation remain necessary before laboratory findings can inform prestressed-concrete applications under realistic operating and environmental exposure conditions.

Abstract: Offshore wind turbine monopiles are usually analysed with a single eccentric load applied above the mudline, although the real environmental action is distributed along the structure as rotor thrust, tower wind drag and hydrodynamic wave load, and the balance between them changes with the sea state. The consequences of this idealisation for the foundation response have not been systematically examined. This study compares three load cases for a 5 m diameter monopile embedded 25 m in dense sand and supporting the NREL 5 MW turbine: a conventional single hub load, rated wind with a normal sea state (NSS), and extreme wind with an extreme sea state (ESS). Their effective load heights above the mudline were 125.0, 96.1 and 26.7 m (e/D = 25.0, 19.2 and 5.3). A three-dimensional coupled finite-element model in Abaqus, validated against FAST, ADAMS and OpenSees modal frequencies and the PISA Dunkirk field test, and an independent nonlinear p–y model were used. Responses were compared at equal mudline displacements of 0.10 m and 0.50 m (0.1D). In both models, the ESS case needed 3.1–3.5 times the horizontal load of the hub-load case at the same displacement, while carrying only 67–74% of its overturning moment. ESS loading moved the rotation point and the maximum-moment location deeper, reduced the maximum bending moment by 21–25%, and reduced the normalised toe reaction to about 0.3 times the hub-load value. The NSS case remained close to the hub-load case. These trends were consistent in both soil models, whereas absolute stiffness and the depths of the rotation point and maximum moment depended on the soil model. A single hub load is therefore adequate for thrust-dominated operation but misrepresents the load path, bending-moment distribution and soil mobilisation under wave-dominated extreme conditions, with implications for embedded-section design.

Abstract: Various studies have proposed using plasma treatment prior to anaerobic digestion to enhance the degradation of organic dry residue (oDR) and increase biogas production. This disintegration process also significantly affects phosphorus release by disrupting cell walls. Several studies show that the driving disintegration process through plasma implementation occurs due to the production of hydrogen peroxide. Since targeted plasma treatment is still technically demanding, this study proposes to estimate the related P release by applying defined hydrogen peroxide dosage to the test medium. Here, the approach was implemented on washed activated sludge (WWAS) in batch tests using H₂O₂ at two different temperatures: 20 °C and 60 °C. Results show only heating treatment mobilised approximately 23% P as PO₄-P, while only 9% P was mobilised by applying 3.4 g H₂O₂ / g DR at 20 °C. The reason behind this scenario is the low reactivity of H₂O₂ as catalase is active at ambient temperature, which also caused only 3% COD solubilisation. However, complete COD solubilisation was achieved by applying similar dosage of H₂O₂ at 60 °C. In this treatment condition, dissolved P mobilisation increased to 40% while PO₄-P mobilisation reduced by 11%. The reduction of PO₄-P can be occurred by parallel formation of P related precipitants, since metal ions like Ca and Mg released in this study. Additionally, the oxidation of ammonia to nitrate is also observed, resulting in lower pH (4.41). Also, the oxidation of Fe²⁺ to Fe³⁺ could cause FePO₄ precipitation. These observed effects are probably transferrable to activated sludge from other wastewater treatment plants (WWTPs). Summarizing these effects, high temperature developed in intense wet plasma treatment can prevent the intended P recovery from the wet sludge phase or sludge liquor.

Abstract: The variability of thermal protection provided by nominally identical fire-resistant glazing panes is poorly captured by the pass/fail outcome of full-scale fire-resistance tests. We re-analysed 15-s unexposed-surface temperature histories from 60 same-type, full-size panes nested within seven heterogeneous EI 30 partition tests. Between 30 and 45 min, the median within-test interquartile range increased from 1.59 to 5.20 K; an increase was observed in six of seven tests. A complementary maximum-above-median measure exposed individual heated panes even where the interquartile range did not increase. A five-minute differential descriptor identified two early accelerated responses (T2-G2 and T5-G6) in a common observation window, whereas two additional original flags arose only later during longer exposures. Time to a 140 K mean pane-temperature rise ranged from 8.12 to 33.91 min; this descriptor is not an additional fire-resistance rating. The exploratory lower-tail calculation was sensitive to rare components and cannot establish an independent pane-count effect. These findings demonstrate substantial heterogeneity and distinct anomalous trajectories within one nominal glazing product while preserving the original EI 30 classification.