Punching shear failure in flat slabs remains one of the most critical issues in reinforced concrete structures due to its sudden and brittle nature. This research investigates the potential of using Slurry Infiltrated Fiber Concrete (SIFCON) as an alternative to traditional punching shear reinforcement in flat slabs cast from normal-strength concrete (NSC). The fiber type was hooked-end steel fiber with a fraction volume of 6%. Five square flat slab specimens were cast; the first specimen was cast using NSC. Four of them were reinforced with maximum flexural steel to ensure failure by punching shear. In two of these slabs, SIFCON was used as a whole slab, but since it is expensive to use SIFCON for entire slabs, in others, SIFCON was used partially over an extended area in a form similar to a plus sign shear reinforcement and full or partial depth to improve the resistance to punching shear. The data demonstrate that SIFCON, when applied strategically, is a highly effective material for increasing punching shear resistance and improving post-punching behavior and ductility. Providing a realistic alternative to standard reinforcement approaches.
This study investigated the influences of metakaolin on the compressive strength and density of normal-and high-strength concrete. The metakaolin, sourced from a local kaolin mining site in Umuahia and calcined at 800 degrees C in the kiln at the University of Nigeria, Nsukka, was characterised using X-ray fluorescence. Its high silica (67.5 mol%) and alumina (26.5 mol%) contents confirmed strong pozzolanic activity, while particle-size analysis showed a sub-micron to low-micron distribution that enhances reactivity. Incorporating up to 10% metakaolin increased normal-strength concrete strength from 23.8 to 37.6 N/mm2 at 7 days and from 32.8 to 48.1 N/mm2 at 28 days. High-strength concrete also improved, with density rising from 2486 to 2575 kg/m3 and strength increasing from 50.9 to 72.6 N/mm2 at 28 days. Ultracrete 61 superplasticiser further enhanced workability and compressive strength. Although chloride content (1.95 mol%) raises corrosion concerns, the findings underscore the need for careful proportioning and material evaluation to support more durable and sustainable concrete formulations.
This study employs Scheffe's Simplex Lattice Design to predict and optimize the split tensile strength of concrete incorporating reclaimed asphalt pavement (RAP) as a partial replacement for natural coarse aggregates. A {5,2} augmented Simplex Lattice, executed in Minitab 22, generated 21 experimental runs to evaluate the effects of cement, sand, water, natural coarse aggregates, and RAP. Pseudo-components were transformed into real ratios using a matrix-based approach, ensuring accurate mixture proportion representation. A quadratic regression model, with an R2 of 92.89% and a significant F-value of 0.98 (p = 0.049), demonstrated strong predictive accuracy. All main components exhibited significant effects, with low variance inflation factors (VIF approximate to 1.59) indicating minimal multicollinearity. Optimal split tensile strength (3.14 N/mm2) was observed in a sanddominated mixture (Run 18), closely matching the predicted 3.118 N/mm2. Experimental results highlighted that a 75% RAP replacement maximized 28-day split tensile strength (3.05 N/mm2), suggesting RAP's viability as a sustainable aggregate. Residual analysis and a lack-of-fit p-value of 0.032 confirmed model adequacy. These findings offer a robust framework for optimizing concrete mixtures, advancing sustainable construction practices and predictive modeling in civil engineering.
This study introduces a novel hybrid machine learning framework for predicting the California Bearing Ratio (CBR) in soil stabilization applications by integrating Neuronal Auditory Machine Intelligence (NeuroAMI), Particle Swarm Optimization with Differential Evolution (PSO-DE), and ensemble artificial neural networks (ANNs). The NeuroAMI component, inspired by the Mismatch Negativity (MMN) effect observed in mammalian auditory cortex processing, implements change detection (CD) and model adjustment (MA) through cochlear-inspired multi-scale frequency decomposition and predictive coding mechanisms. The framework was evaluated on soil stabilization datasets incorporating Rice Husk Ash (RHA), Fines Content (FLD), Optimum Moisture Content (OMC), and Maximum Dry Density (MDD) as input parameters for predicting both unsoaked and soaked CBR values. Data preprocessing included Isolation Forest outlier removal, interaction-and ratio-based feature engineering, and bootstrap data augmentation. An ensemble of ten NeuroAMI models underwent PSO-DE coefficient optimization (40 particles, 120 iterations, 10 multi-start runs). Simultaneously, three ANN architectures-Feedforward, Deep, and Residual-were trained using 10-fold cross-validation, producing an ensemble of 30 models. Results revealed distinct performance disparities. For CBR unsoaked, the ANN ensemble achieved R-2 = 0.866, while NeuroAMI attained R-2 = 0.778 on the test set but degraded to R-2 = 0.523 on the holdout prediction set. For CBR soaked, the ANN ensemble maintained high accuracy (R-2 = 0.882), whereas NeuroAMI exhibited severe performance degradation (R-2 = -2.667 on the test set, -1.169 on the prediction set). Feature importance analysis identified MDD (20 %) and interaction terms (37 % combined) as dominant drivers for CBR unsoaked, while OMC was the most influential factor for CBR soaked (40 %). Sensitivity analysis confirmed MDD as the most influential predictor, producing output variations up to 80 for soaked conditions. Cross-validation indicated substantial variability across folds (R-2 = 0.55-0.95), reflecting dataset heterogeneity. Overall, results demonstrate that auditory cortex-inspired architectures are fundamentally unsuitable for static soil mechanics prediction, whereas conventional deep ensemble ANNs provide reliable and robust performance (R2 > 0.86) across both soaked and unsoaked conditions. These findings emphasize the need to align bio-inspired computational paradigms with domain-specific problem characteristics and establish ensemble deep learning as the preferred methodology for CBR prediction in geotechnical engineering.
Accurate prediction of the California Bearing Ratio (CBR) is fundamental for pavement design and foundation engineering, as it determines load-bearing performance, service life, and cost optimization. Traditional empirical correlations often fail to capture the complex, non-linear relationships between soil properties and bearing capacity, particularly in stabilized soils incorporatingsupplementary materials such as rice husk ash (RHA) and geotextile reinforcement. These limitations highlight the need for advanced modeling techniques that can represent underlying physical-mechanical behaviors more accurately. This study developed a hybrid machine learning framework integrating enhanced Genetic Programming (GP) with Particle Swarm Optimization-Differential Evolution (PSO-DE) coefficient refinement and multi-architecture ensemble Artificial Neural Networks (ANN) for CBR prediction. Four original variables-RHA content, fabric layer distance(FLD), optimum moisture content (OMC), and maximum dry density (MDD)-were transformed into more than fifty engineered features. Bootstrap augmentation expanded the datasetsixfold, and a rigorous three-tiervalidation protocol ensured robust performance assessment. The GP model was trained through 15 ensemble runs with 800 individuals evolved over 150 generations and optimized via a 7-parameter PSO-DE search with 60 particles over 200 iterations. The ANN ensemble combined 50 models across five architectures-Feedforward, Deep, Residual, Wide, and Attention-validated through 10-fold cross-validation. The ANN ensemble achieved superior performance with R-2 = 0.856, RMSE = 0.415, and MAE = 0.328, while GP achieved R-2 = 0.767, RMSE = 0.521, and MAE = 0.412. Notably, GP outperformed its test set accuracy on the holdout prediction set (R-2 = 0.821), indicating strong generalization. ANN exhibited stable error distribution (+/- 0.4 units), whereas GP showed heteroscedasticity (+/- 1.0 unit). Both models exceed engineering acceptance thresholds (R-2 > 0.75). ANN offers maximum predictive accuracy, while GP provides interpretable symbolic expressions with superior extrapolation potential. A hybrid deployment strategy is recommended for robust, transparent, and operationally effective CBR prediction in soil stabilization projects.
The use of recycled materials in ceramic production is becoming increasingly important in the development of more sustainable and cost-effective materials. This study focuses on the use of recycled glass fiber-reinforced epoxy resin (GFRE) powder, produced by mechanical grinding, as an additive in mullite-based ceramic systems. The goal is to investigate how the addition of this composite powder affects mullite formation and the physical properties of the fired ceramic specimens. The main aim of the study is to determine whether the recycled GFRE powder can act as a reactive filler that promotes mullite formation at lower temperatures, and whether it has any effect on the porosity or other characteristics of the sintered ceramics. X-ray diffraction (XRD) and scanning electron microscopy (SEM) will be used to analyze the phase composition and microstructure of the samples. Using GFRP waste as a ceramic additive offers a novel way to recycle difficult-to-process composite materials. This approach may contribute to the development of more sustainable ceramic technologies and provide a practical solution for reusing industrial composite waste in high-temperature applications.
The research investigated the structural performance of concrete made with ironstone as coarse aggregate. Ironstone is available in abundance in the south-eastern part of Nigeria especially Anambra and Enugu states. In view of the usual errors associated with traditional laboratory experimental procedures, a machine learning approach (Gene Expression Programming) was employed in GeneXpro Tools for the prediction of the compressive and flexural strengths of ironstone concrete. For this purpose, a database consisting of 352 data points was constructed by replacing ironstone with granite chippings and river gravel up to 50% in 1:2:4 concrete at 0.45, 0.5, and 0.55 water to cement ratios (W/C) respectively. The data set was divided into two sets called the training and validation datasets having 70% and 30% of the data respectively. The training data set was used to train the algorithm while the validation data set was used to validate the algorithm. The algorithm accuracy was checked by calculating the six commonly used errors: mean squared error (MSE), root mean squared error (RMSE), mean absolute error (MAE), and relatively root squared error (RRSE), coefficient of correlation (R) and R2 for both data sets. The statistical evaluation shows that the R2 Values are within range specified in the literature (greater than 0.8 and less than 1.0). At 50% replacement with granite chippings at 0.45W/C, optimum results yielding compressive strength of 32.30 N/mm2 and flexural strength of 13.15 N/mm2 was achieved. The accuracy of the algorithm was verified using K-Fold cross validation, plotting scatter and shapely sensitivity test. Thus the developed equation can be used to forecast the 28day compressive and flexural strengths of ironstone concrete.
Production of eco-friendly concrete is a global aim, to tackle this matter; Geopolymers have been identified as viable alternative replacements for ordinary Portland cement due to its excellent engineering properties and low CO2 emissions. The advancement of geopolymer concrete (GPC) represents a significant milestone to make it more applicable and popular. An effort was made to investigate the impact of the blended Metakaolin (MK) and Ground Granulated Blast Furnace 2.5%) as reinforcement additive. For this, an experimental study was carried out to evaluate the effect of PPF on properties of the product. SEM, XRD, and FTIR analyses have been performed to determine the surface morphology and phases. Results show that incorporating PPF effectively reduces the shrinkage and improves its compressive strength compared with the control geopolymer mortar; optimum value was obtained with a PPF ratio of 0.5%.
The integration of industrial by-products into concrete mixtures has gained significant attention as a sustainable approach to improving material performance while reducing environmental impact. This study investigates the effects of black liquor (BL), a lignin-rich by-product of the paper industry, as a novel admixture in concrete. Concrete samples with varying BL concentrations (0-4 wt.%) were prepared and evaluated for mechanical strength, workability, density, porosity, and microstructural modifications. The results indicate that the incorporation of 2 wt.% BL optimally enhances concrete properties, yielding a 25% increase in compressive strength (from 40 MPa to 50 MPa) at 28 days. Furthermore, splitting tensile and flexural tensile strengths peaked at 3.6 MPa and 15 MPa, respectively, at this concentration. Workability improved significantly, as evidenced by a 200% increase in slump value compared to BL-free concrete. Additionally, bulk density reached its maximum at 2.28 kg/L, while apparent porosity exhibited a notable decline to 9%, indicating matrix densification. Scanning electron microscopy (SEM) confirmed the refinement of pore structure and enhanced cementitious bonding at 2 wt.% BL. However, excessive BL content (>= 3 wt.%) led to reduced performance due to increased porosity and disruption of cement hydration. These findings highlight the potential of black liquor as an effective and sustainable concrete admixture, offering enhanced mechanical properties and improved durability while promoting industrial waste reutilization.
Structural fire resistance is a primary aspect of a passive fire engineering design measure that allows a structure to withstand intense fires. Generally, concrete structural elements perform well under these conditions due to their non-flammable nature. However, fire incidents require a deeper understanding of concrete behavior and structural mechanics to improve fire design. Structural elements exposed to fire and heat show reduced strength; this reduction must be evaluated to determine whether to demolish or repair a building based on its condition and capability to support future loads. Evaluating the post-fire strength characteristics, including compressive, tensile, and bond strengths, is essential for determining the structure's safety. Prolonged exposure to high temperatures can degrade concrete properties, particularly the bond strength between rebar and concrete. This paper investigates the bond strength of materials after exposure to fire. The study explores the effects of temperatures ranging from 20 degrees C to 500 degrees C, following the ISO 834 fire curve, on compressive, tensile, and bond strengths. Cylindrical pull-out specimens were heated to specific temperatures and held for 2 hours. Afterward, they were cooled for one day before testing. The results indicate that bond strength decreases by approximately 72% at 500 degrees C, about twice the reduction observed in compressive and tensile strengths.
The natural illitic clay was mixed with powder calcite (22 wt.% in the dry mixture), distilled water, and 3% solution of polyvinyl alcohol to obtain a plastic mass for samples. To determine the correct temperature dependence of Young's modulus, thermogravimetry (TG), thermodilatometry (TDA), and impulse excitation technique (IET) have to be performed in the same temperature regime. The effect of the size of TDA sample on Young's modulus was negligible. Contrary to the TDA sample, the form of the TG sample had a significant effect. When results from a small TG powder sample are substituted into the formula for Young's modulus, relative error up to 10% can result in the temperature region in which changes of some mineral components occur. this study, we showed the importance of using compact samples with the same cross-section for TG, TDA, and IET in order to obtain correct values of Young's modulus during thermal treatment of the illitic
This study investigates the mechanical properties of Hair/Sisal/Glass Fiber Reinforced Polyester plastic (HSGFRP) hybrid composites composed of 15% fiber content by weight, with equal proportions of hair, sisal, and glass fibers, and 85% polyester matrix. However, to ensure accuracy in representing the material composition, the fiber content was recalculated and expressed in terms of volume fraction (vol%), consideringthe differences in densities amongthe fibers and the matrix material. Based on this adjustment, the total fiber content corresponds to approximately 7.67% by volume, with equal volumetric proportions of hair, sisal, and glass fibers. Specimens fabricated using the hand lay-up technique were tested for tensile, flexural, impact, and water absorption properties. The HSGFRP achieved 53% of the tensile strength, 94% of the flexural strength, and 77% of the impact strength of Glass Fiber Reinforced Composite (GFRP), with moderate water absorption properties comparable to glass and glass/sisal fiber materials. The addition of glass fiber enhanced tensile and flexural strength while reducing impact strength. HSGFRP exhibits balanced mechanical properties, making it a viable, lightweight, and costeffective alternative to traditional glass fiber composites for light-load applications, despite a noticeable gap in tensile strength compared to GFRP.
The ability of natural clays to suspend particles makes them a promising candidate for use as dispersants, with potential applications in pigment dispersions. Unlike conventional dispersing agents that stabilise particles through adsorption, clays can maintain particle suspension by forming a three-dimensional network structure due to their charged surfaces. While there are reports of clays effectively dispersing various nanoparticles, to the best of our knowledge, their use as dispersing agents for pigments has not yet been explored. Clay-assisted dispersion presents a valuable opportunity for the coatings industry due to the low cost and minimal toxicity of clays. In contrast, conventional dispersants such as alkylphenol ethoxylates (APEs) are subject to increasing restrictions and scrutiny due to their high toxicity and long-term environmental persistence. Here, we reportthe use ofclays as dispersingagentsfor an organic pigment. Dispersions comprising various clay specimens (smectite, mica, and kaolinite) and Pigment Yellow 138 (PY138), a yellow organic pigment, were prepared via an in-situ grinding process at both low (0.1% w/w) and high concentrations (0.5% to 5% w/w). Among the tested clays, smectite demonstrated superior colloidal stability compared to the control. At low concentrations, smectite produced the most stable dispersions, while at higher pigment concentrations, a critical threshold was observed at approximately 1.0% w/w smectite.
Mg-Al-LDHs with different Mg/Al mol ratio were studied. Adsorption of different organic substances on the layered double hydroxides (LDHs) of different composition was studied. Fatty alcohols, phenols and naphthalene derivatives were used as a sorbates. The changes of structures, taking place because of this process, was explored by X-ray method. It has been found, that at the adsorption of such substances as alcohol and phenols the anion exchange take place. After the sorption the specific surface and inner distance of LDHs were measured. The mechanism of structure change was proposed.
Geotextile fabrics are integral to modern soil stabilization practices, offering benefits such as cost-effectiveness, ease of installation, and environmental sustainability. This paper explores the mechanisms of geotextile reinforcement, including separation, filtration, drainage, and reinforcement, which contribute to their effectiveness in stabilizing expansive soils. Geotextiles reduce swelling pressure, improve load-bearing capacity, and mitigate shrinkage cracks, making them essential in various geotechnical applications. Despite their advantages, challenges such as durability, compatibility with different soil types, and potential clogging issues need to be addressed. Recent technological advancements, including smart and nano-modified geotextiles, and improved manufacturing techniques have significantly enhanced their performance. Hybrid approaches integrating geotextiles with other stabilization methods demonstrate synergistic effects, providing comprehensive solutions to complex geotechnical challenges. This review highlights the critical role of geotextiles in soil stabilization, emphasizing the need for ongoing innovation and careful material selection to maximize their benefits and address existing limitations.
