2031 research outputs found
Sort by
Optimized Feature Selection for Predicting the Number of Casualties in Traffic Crashes
Get PDFTraffic crash prediction remains a critical challenge in transportation safety management, with increasing emphasis on leveraging machine learning techniques for accurate casualty prediction. This study aims to develop an optimized feature selection framework for traffic crash casualty prediction by comparing six selection techniques: Design of Experiments (DOE), Forward and Backward Sequential Feature Selection, Information Gain, Lasso Regularization, and Random Forest (RF) Feature Importance, with subsequent integration using the Borda count method. By analyzing 517,000 UK traffic crash records (2019-2023), 25 machine learning models (linear models, decision trees, ensemble methods, and neural networks) were evaluated across 12 critical attributes. Results demonstrate eXtreme Gradient Boosting (XGBoost)'s superior performance with a Root Mean Square Error (RMSE) of 0.671 and Mean Absolute Error (MAE) of 0.372 using the proposed Borda count integration method while maintaining efficient computation time (11.3 minutes compared to the baseline's 17 minutes). Five factors consistently emerged as the most influential predictors across all selection methods: number of vehicles involved, speed limit, police officer attendance, day of the week, and urban/rural classification, while environmental factors showed lower importance than traditionally assumed. The novel integration of multiple feature selection techniques through Borda count provides a more robust feature subset than any individual method, offering an optimal balance between computational efficiency and prediction accuracy. The framework enables transportation safety authorities to implement more efficient crash prediction systems while providing actionable insights about key risk factors for targeted interventions, especially to support the Highway Safety Manual development. Doi: 10.28991/CEJ-2025-011-04-01 Full Text: PD
Evaluating the Microstructure and Strength of Geopolymer Mud Blocks for Sustainable Architecture
Get PDFThis study investigates the physio-mechanical, microstructural, and durability characteristics of Geopolymer Mud Blocks (GMB) as a sustainable alternative to traditional Soil Stabilized Blocks (SSB). Utilizing locally available Alumino-Silicate Sources (ASS) and Alkali-Activated Materials (AAM), GMB were produced with varying molarity levels (6M, 7M, and 8M) and mix proportions (M1 to M3). Experimental results reveal that compressive strength increased by 10–20% with molarity escalation from 6M to 8M. The highest compressive strength of over 50 MPa, achieved with the M4 mix at 8M, equaled M50-grade concrete, making it suitable for load-bearing walls in earthquake-resistant structures. Durability tests demonstrated less than 10% water absorption, indicating low permeability. Type B6 (6% AAS, 8M, 28 days) exhibited superior performance, attaining the highest compressive strength of 47.32 MPa and prism strength of 33.12 MPa. Additionally, it showed commendable durability metrics, including water absorption at 5.20%, chloride diffusion at 1.87%, acid diffusion at 3.33%, and sulphate diffusion at 1.05%. The dense matrix and minimal porosity of this mix, resulting from the use of distilled water and optimal binder content, significantly enhanced its strength and durability. Type C6 (6% AAS, 8M, 28 days) exhibited the weakest performance, characterized by high porosity, suboptimal matrix quality, and unfavorable durability indicators, such as water absorption (10.33%) and chloride diffusion (4.47%). Type B6 demonstrates the highest effectiveness, providing an optimal balance of strength and durability, whereas Type C6 exhibits the lowest efficiency. GMB exhibited enhanced resistance to acid, sulphate, and chloride attacks with increased molarity. XRD analysis confirmed the geopolymerization process, with significant diffraction peak changes. SEM images revealed denser microstructures with higher molarity, correlating with increased strength. The study concludes that GMBs offer superior strength, durability, and cost-strength efficiency compared to SSBs, promoting sustainable construction practices. Doi: 10.28991/CEJ-2025-011-04-09 Full Text: PD
Crack Pattern Analysis and Reinforcement Strain Development in UHPSFRC-Strengthened RC Joints
Get PDFThis study assesses the effectiveness of ultra-high performance steel fiber-reinforced concrete (UHPSFRC) in improving the seismic behavior of reinforced concrete (RC) exterior joints, emphasizing crack patterns and reinforcement strain development. Three full-scale specimens (a conventional RC control joint and two UHPSFRC-strengthened versions with 800 mm and 1025 mm strengthening lengths) were subjected to reversed cyclic loading to mimic seismic forces. Crack progression and strain distribution were examined through visual observations, strain gauges, and displacement data, offering a detailed evaluation of joint performance. Findings indicate that UHPSFRC enhances shear resistance, reduces crack widths significantly (<0.5 mm compared to >2 mm in the control), and modifies failure modes: the 800 mm length shifts damage to beam flexural failure, while the 1025 mm length increases peak capacity (231.4 kN vs. 185.8 kN) but reverts to joint shear failure. The novelty lies in UHPSFRC's ability to replace transverse reinforcement in congested joint zones, enhancing ductility and easing construction difficulties. This research provides fresh insights into optimizing UHPSFRC application length, delivering practical guidance for seismic retrofitting, and contributing to design standards for robust RC frames in seismic regions. Doi: 10.28991/CEJ-2025-011-04-014 Full Text: PD
Modeling the Compressive Strength of Metakaolin-Based Self-Healing Geopolymer Concrete Using Machine Learning Models
Get PDFMetakaolin-based self-healing geopolymer concrete treated with Bacillus bacteria represents a significant advancement in sustainable construction due to its eco-friendly properties, enhanced durability, and self-healing capabilities. It is a transformative material for sustainable construction. By reducing carbon emissions, utilizing waste, improving durability, and lowering lifecycle costs, it aligns with global goals for environmentally friendly and resilient infrastructure. Continued research and development will further unlock its potential, making it a cornerstone of the future of sustainable construction. In this research project, a study on modeling the compressive strength of environmentally friendly metakaolin-based self-healing geopolymer concrete treated with Bacillus bacteria (BB) has been conducted, analyzed, and reported. Machine learning methods such as the "Group Methods Data Handling Neural Network (GMDH-NN)”, "Generalized Support Vector Regression (GSVR), "K-Nearest Neighbors (KNN)”, "Tree Decision (Tree)”, "Random Forest (RF)” and "Extreme Gradient Boosting (XGBoost)” were applied to model the compressive strength of the self-healing concrete. The GMDH-NN model was created using GMDH Shell 3.0 software, while XGBoost, GSVR, KNN, Tree, and RF models were created using "Orange Data Mining” software version 3.36. The research method also included gathering relevant experimental and field data, categorizing it effectively, and performing initial analysis to identify trends and relationships. A global representative database was collected from literature for different mixing ratios of self-healing concrete corresponding to the compressive strength, with a total of 147 records, which contained Fly Ash (FA), Silica Fume (SF), Metakaolin (MK), and Bacillus Bacteria (BB) considered as the input constituents. The collected records were divided into a training set (75%) and a validation set (25%) based on established requirements. At the end of the modeling exercise, the GMDH-NN produced the best model with an accuracy of 0.99, while the KNN and the GSVR followed closely with accuracies of 0.975 and 0.97, respectively. However, the RF and the Tree models also produced good accuracies of 0.965 and 0.955, respectively. Also, the GMDH-NN and the KNN again outperformed the other methods, producing an R² of 1.00 and 0.99, respectively, while the GSVR, RF, and Tree followed in this order with R² of 0.98, 0.97, and 0.96, respectively. The error indices, such as the overall error, RMSE, MSE, MAE, and SSE, also confirm this order of performance. The sensitivity analysis on the modeling of compressive strength of metakaolin-based self-healing geopolymer concrete treated with Bacillus bacteria produced a metakaolin (MK) impact of 30%, a silica fume (SF) impact of 29%, a fly ash (FA) impact of 27%, and a Bacillus bacteria (BB) impact of 14%. This highlights the dominant role of metakaolin (30%), silica fume (29%), and fly ash (27%) in determining the compressive strength of metakaolin-based self-healing geopolymer concrete. Bacillus bacteria (14%) have a smaller but meaningful impact, primarily contributing to self-healing and long-term durability. These insights can guide material selection, mix design, and process optimization to enhance both strength and durability. Doi: 10.28991/CEJ-2025-011-04-020 Full Text: PD
Effect of Infill Wall Opening Ratio on the Mechanical Characteristics of Reinforced Concrete Frames
Get PDFThis study investigated the influence of infill wall (IW) opening ratios on the mechanical performance of reinforced concrete (RC) frames using a novel numerical model. The proposed model incorporated stiffness degradation and a nonlinear "Gap Element" to simulate the interaction between RC frames and IWs under seismic loading. A 3D finite element model was developed in SAP2000 and calibrated using validated experimental data. Parameters such as IW thickness, opening ratio (0–100%), and opening position (symmetric, asymmetric, corner) were systematically varied to assess their effects on lateral displacement , fundamental period , shear force , and bending moment . The results indicated that increasing the opening ratio significantly reduces frame stiffness, especially beyond 40%, and leads to substantial increases in displacement. Corner openings were found to have the most detrimental impact, while thicker walls (≥220mm) can partially mitigate stiffness loss. However, at ratios above 60%, even thick IWs failed to preserve structural performance. Based on these findings, a limit of 40% opening ratio was recommended for design purposes, and reinforcement was advised for higher ratios. The study provides a practical framework for optimizing the seismic and structural design of RC frames with openings in IWs, contributing new thresholds and modeling strategies for improved performance
Phosphate Adsorption from Aqueous Solutions Using Eggshell and Sacha Inchi (Plukenetia volubilis) Mixture
Get PDFThe use of bioadsorbents for the removal of pollutants is being increasingly investigated worldwide due to their high efficiency and the potential use of various natural sources. The present study introduces a novel approach for phosphate adsorption using sacha inchi cuticle and eggshell mixture. These materials were pyrolyzed (400°C for 20 min) and mixed in a 1:10 (eggshell:cuticle) ratio. An adsorption study was carried out using synthetic phosphate solution concentrations of 0–300 mg/L and adsorbent masses of 0.1–1 g/100 mL. The temperature, pH and stirring were kept constant (25°C, pH:5 and 150 rpm) during the tests. The phosphate adsorption capacity increased as higher phosphate concentrations were used, reaching a maximum of 300 mg/L. However, differences in removal were observed when varying the amount of adsorbent used, reaching equilibrium in approximately 1 h, with a percentage of phosphate removal between 31 and 41%. The adsorption process followed a Freundlich isotherm with a correlation coefficient of 0.97, suggesting a multilayer adsorption process. According to the SEM-EDX results confirmed a high concentration of carbon and oxygen in the sacha inchi cuticle, in that sense, this by-product could be evaluated for the removal of other contaminants from water
Effect of Graphene Oxide on the Performance of Fly Ash Concrete Exposed to Ambient Temperature
Get PDFThe rising global temperatures due to climate change are accelerating concrete deterioration by shortening its service life, which subsequently increases maintenance costs. Therefore, the objective of this investigation is to analyze the graphene oxide (GO) effect on the mechanical characteristics and microstructural properties of fly ash (FA) concrete exposed to ambient temperatures. Concrete specimens were created by employing GO from 0.01% to 0.05% by weight of cement and cured using two distinct methods. These include standard curing in immersed water and for 7 days followed by ambient exposure. The mechanical test showed that GO significantly enhanced compressive strength, with 0.04% GO observed to have increased strength by approximately 16% at 28 days. However, exposure to ambient conditions led to decreased compressive and flexural strength and increased mass loss. The microstructural analysis also showed that ambient-exposed concrete exhibited higher porosity and incomplete hydration. The results showed that the addition of GO enhanced durability by refining the microstructure, reducing porosity, and enhancing thermal stability. Thermal analysis also confirmed that GO minimized moisture loss and improved thermal resistance. Furthermore, Fourier Transform Infrared Spectroscopy (FTIR) validated the improvement in bonding for the GO-FA concrete. These results showed that GO could mitigate the adverse effects of environmental exposure, leading to its identification as an advantageous additive to increase the long-term durability and concrete performance in different temperature conditions
Retrofit Design for Climate Resilient Housing: Strategies for Architectural Adaptation to Climate Change
Get PDFThis study examines design flaws in single-family homes in the UAE, worsened by climate change-triggered rainfall and escalating maintenance requirements. The research focuses on three objectives: identifying existing weaknesses, analyzing building materials and construction methods, and proposing enhanced retrofit design standards. The methodology comprises both secondary data, gathered through literature reviews, and primary data obtained via site visits, participatory observation, and case studies. Examining multiple UAE regions, particularly six case studies in affected housing in Dubai, Sharjah, and Ajman, underscores widespread concerns in resilient housing, revealing deficiencies in drainage, waterproofing, and protective detailing. Notable problems include inadequate drainage slopes, subpar sealing around structural penetrations, and insufficient moisture barriers. These issues compromise structural integrity, inflate maintenance costs, and pose health hazards from mold and poor indoor air quality. By assessing current conditions, the study suggests various retrofit solutions, such as improved water-resistant coatings, slope modifications, drip edges, and overhangs. Findings emphasize rigorous detailing, robust materials, and periodic inspections to mitigate impacts from intensifying rainfall. Additionally, broader urban planning strategies, such as flood risk assessments and upgraded infrastructure, are crucial in minimizing future water intrusion. Collectively, these insights advocate novelty in research and set a blueprint for a fundamental shift in UAE housing design, prioritizing climate resilience, structural longevity, and occupant well-being in an era of rapidly changing environmental conditions. Doi: 10.28991/CEJ-2025-011-03-011 Full Text: PD
Piezometer Time-Lag and Pore Pressure Ratio for Identification of Dam Internal Erosion
Get PDFEarth dams on complex geology without proper foundation treatment often face the risk of seepage problems. Sufficient installation and interpretation of field instruments are essential for monitoring dam behavior. Three indicators are introduced for assessment of seepage behavior: time lag (TL), pore pressure ratio (PR), and trigger water level (HW). The normalized TL reflects the washing out and plugging of rock cracks, as well as the progression of internal erosion. The foundation of the studied dam consisted of foliated rocks that were highly fractured, with the axis of the foliations aligned almost in the upstream-downstream direction, with a possible low-stress zone on the syncline axis. The existing crack easily opened in the concave section of the syncline when the reservoir had risen to a certain elevation, resulting in increased permeability and a higher flow to the downstream area, known as "hydraulic fracturing” (HF). The piezometer TL clearly indicated a shorter response time as the operating period progressed. The study dam showed the possibility of HF in the foundation, as observed during 2003–2024. The progression of HF was also confirmed by the increase in PR levels toward downstream. This revealed that the ongoing progression of HF had occurred at sta.2+700, which agreed well with the location of the slip zone that had occurred in 1993. HWwas activated by the reservoir water level response also decreasing with time from 2003 to 2024, confirming that water infiltration through the rock crack progressed with time. These three indicators could act as good warning indices for seepage problems. This compiled knowledge could be transformed into a flowchart to identify the possible risks of hydraulic fracturing in the dam. If the three indices all showed the same trend, the potential for hydraulic fracturing and internal erosion would be very high. Doi: 10.28991/CEJ-2025-011-03-019 Full Text: PD
IRI Performance Models for Flexible, Semi-Rigid and Composite Pavements in Double-Carriageway Roads
Get PDFPavement Management Systems (PMS) depend upon reliable pavement performance models. In this paper, our aim is to develop International Roughness Index (IRI) prediction models for the heavily trafficked (right-hand) lanes of motorways in the province of Gipuzkoa (Spain) in flexible, semi-rigid, and composite pavements. A deterministic approach was selected, based on the available information in the PMS employed in that province, covering complete pavement structures. Omitting pavement type, the model yielded a determination coefficient (R²) of 0.696 with only three variables: pavement age, cumulative volume of heavy vehicles travelling through the section, and total thickness of bituminous layers. Then, two superior models were generated with pavement type as a variable, yielding R²values of 0.781 and 0.795, respectively. Unlike the opaque features of Machine Learning (ML), the deterministic models captured precise relationships between the variables to a high degree of accuracy. They can moreover be applied to all pavements with bituminous layers, unlike many other models that are only applicable to a single pavement type. Furthermore, the models are presented for freeways where traffic is randomly distributed between lanes; a less widely covered topic in the literature. Doi: 10.28991/CEJ-2025-011-05-01 Full Text: PD