Engineering and Applied Science Research

Electrospinning nanofibers for wound healing using antioxidant from Rang Jued (Thunbergia laurifolia Lindl.) extract via subcritical fluid extraction

2025-08-12T14:11:06+07:00

Rang Jued (Thunbergia laurifolia Lindl.) is a local Thai plant known for its bioactive compounds. In this study, subcritical ethanol extraction was used to extract antioxidant from Rang Jued (RJ) leaves. A total of 15 experiments were designed using Box-Behnken design. Response surface methodology was employed to determine the optimal condition, which yielded the highest DPPH scavenging activity of 94.91% and a total phenolic content of 30.35 mg GAE/g under the conditions of 1.64 g of Rang Jued powder, an extraction temperature of 190 ºC, and an extraction time of 15.14 minutes. Furthermore, the electrospinning technique was used to fabricate antioxidant wound dressing nanofibers. The process was conducted by varying the ratio between PVA and RJ extract, as well as the voltage supply. Scanning electron microscopy (SEM) was used to investigate the morphology of the nanofibers. The average diameter ranged from 248 to 362 nm. The highest antioxidant activity of the nanofibers was observed at 72.25%, using a PVA:RJ extract ratio of 7:3 and a voltage of 40 kV.

Classification of Thailand’s industrial firms under global supply chain disruptions: Integrating resilience and sustainability in industrial performance

2025-07-15T15:58:42+07:00

There has been significant disruption of global supply chains as a result of pandemics, geopolitical tensions, and climate-related events. Firms around the world, including those in Thailand, have been compelled to adapt, with increasing emphasis on resilience and sustainability. While many studies have addressed supply chain performance through these lenses, few have focused specifically on the industrial context in Thailand. Moreover, the direction of industrial transformation in response to global challenges remains insufficiently examined. This study addresses these gaps by proposing a comprehensive framework that integrates conventional supply chain performance indicators with resilience and sustainability dimensions. Data were collected from 98 publicly listed industrial firms in Thailand using their 2023 annual reports. Principal component analysis (PCA) was used to identify key performance patterns, and this was followed by K-means clustering to classify firms based on strategic orientation. The analysis revealed four distinct clusters. The largest group, referred to as resilient enterprises (54 firms), demonstrated a balanced performance across efficiency and resilience. Performance-oriented firms (23 firms) exhibited high customer satisfaction, product quality, and flexibility. Agile and lean operators (6 firms) prioritized operational speed and rapid delivery. Flexibility-centric firms (16 firms) focused on adaptability but faced constraints in financial and inventory performance. Overall, the findings indicate that economic factors such as revenue, profitability, and productivity, together with resilience attributes like agility and flexibility, are the primary variables that differentiate firms. This suggests that Thailand’s industrial firms currently place the most emphasis on operational efficiency and adaptability, while broader environmental and social sustainability concerns remain less prominent in strategic differentiation.

Developing a predictive model for quantity estimation of tie columns and lintel beams in residential construction

2025-07-08T14:22:16+07:00

Accurate construction cost estimation is at the root of any effective project planning, yet it often requires extensive expertise and time-consuming calculations. This paper discusses a predictive equation for estimating the quantity of tie columns and lintel beams in a two-story residential building. In this study, multiple linear regression analysis was employed to identify the significant variables that impact the quantity of those structural elements using 75 sets of residential drawings, all of which featured conventional two-story brick masonry construction with reinforced concrete frames. The formulated equation, where Y represents the total linear meters of tie columns and lintel beams combined, is expressed as Y = 1.834 + 1.243 (brick wall area in m²) - 0.639 (open space area in m²). The equation was checked against fifteen residential designs with detailed estimates. The percentage error was observed to be between -3.58% and 5.37%, which is considered within an acceptable limit for preliminary estimates. This equation could provide a useful tool for cost estimators, offering a much-simplified approach yet yielding reasonable accuracy for preliminary assessments of the structural quantities of buildings. This research highlights the equation's potential for improving efficiency in project planning and cost estimation within its defined scope, with further validation across a wider range of designs recommended to broaden its applicability.

Comparative adsorption study of Pb(II), Fe(II), and Zn(II) using non-chemically activated rubber seed shell biochar and commercial activated carbon

2025-08-18T10:52:17+07:00

The widespread contamination of water sources by heavy metals such as Pb(II), Fe(II), and Zn(II) poses serious environmental and health risks. This study investigated the use of non-chemically activated biochar derived from rubber seed shells, an agricultural waste material, as a sustainable adsorbent for heavy metal removal. Biochars were produced by a two-step carbonisation process at temperatures of 850, 900, and 950 °C, and the physicochemical properties were systematically assessed. The sample carbonised at 850 °C (PRC850) exhibited the most favourable properties, including a high BET surface area (795 m²/g), mesoporous structure, and suitable surface functional groups, as confirmed by SEM, BET, XRD, and FTIR analyses. Initial screening was conducted for Pb(II), Fe(II), and Zn(II) adsorption, and PRC850 demonstrated superior performance, removing up to 98.64% of Pb(II), which was significantly higher than the 85.52% removal rate achieved by commercial-grade activated (CGA) carbon. The adsorption behaviour of Pb(II) was best described by the Langmuir isotherm model, and the pseudo-second-order kinetic model fitted the experimental data well, indicating chemisorption. These findings indicated that rubber seed shell biochar had the potential to serve as a cost-effective and ecologically friendly adsorbent, particularly for Pb(II) removal, while also performing effectively for Fe(II) and Zn(II).

Ceria-modified zeolite: A dual-function approach for effective removal of arsenite from polluted water sources

2025-07-15T16:59:46+07:00

As a carcinogen, arsenic poses a significant threat when it contaminates water sources and agricultural products. In water-based environmental contamination, the most significant arsenic species are arsenate (As(V)) and arsenite (As(III)), with the latter presenting a greater challenge for removal. The development of more efficient adsorbents to successfully remove As(III) from contaminated water is still needed. A novel nanosorbent, ceria supported on Na-P zeolite (CeZ), was created in this study to perform the dual functions of oxidizing As(III) to As(V) and subsequently adsorbing the resulting As(V). CeZ was characterized by XRD, TEM, FTIR, pHpzc, and XANES analyses. Batch adsorption experiments indicated that As(III) removal in the pH range of 3-10 was highly efficient, with a maximum removal capacity of 31.746 mg g^-1, which was best explained by pseudo-second-order kinetics. XANES analysis confirmed that CeZ oxidized As(III) to As(V) on the surface during As(III) adsorption. The hydroxyl groups at the CeZ interface play a key role in As(III) sorption, forming inner-sphere monodentate and bidentate complexes. As(III) removal was effective because the sorption reaction was coupled with the oxidation process. Specifically, the CeO₂ on the Na-P zeolite surface was the main factor responsible for the oxidation of As(III) to As(V) and its sorption. The As(V) in the solution subsequently adsorbed onto the zeolite.

Superior mechanical and tribological properties of Al7075 metal matrix nanocomposites processed through a novel multi-stage casting route

2025-07-03T11:39:24+07:00

This study aims to improve the microstructural features and mechanical performance of Al7075 aluminium matrix composites reinforced with silicon carbide (SiC) nanoparticles and graphite (Gr) through a novel processing route. The proposed method integrates mechanical alloying-assisted semisolid stir casting with die casting, followed by a T6 heat treatment. The Al7075/SiC composite subjected to T6 treatment exhibited superior mechanical properties, including a microhardness of 218 HV, a 0.2% proof stress of 250 MPa, an ultimate tensile strength of 364 MPa, and an elongation of 16%. These enhancements are primarily attributed to synergistic strengthening mechanisms, including grain refinement, Orowan looping, and precipitation hardening. In contrast, the Al7075/SiC/Gr hybrid composite demonstrated a marginally reduced ultimate tensile strength of 254 MPa, representing a 12% decline compared to the Al7075/SiC composite, which was attributed to graphite agglomeration and inadequate interfacial bonding. Wear resistance testing revealed that the SiC-reinforced composite exhibited the lowest material loss, with scanning electron microscopy (SEM) analyses confirming reduced groove depth and plastic deformation. Conversely, the hybrid composite displayed increased surface roughness and porosity, primarily due to graphite-induced defects. These findings indicate that the incorporation of SiC nanoparticles, in conjunction with T6 heat treatment, constitutes an effective strategy for enhancing the structural integrity and mechanical performance of Al7075-based composites. However, further optimization of graphite morphology and dispersion is necessary to fully realize its potential as a solid lubricant in hybrid composite systems.

Evaluation of missing value handling methods in machine learning for emergency department mortality prediction

2025-08-01T13:18:29+07:00

Missing data remains a significant challenge in emergency medicine, particularly in mortality prediction models. This study investigates five distinct missing value handling methods applied to various machine learning algorithms using a dataset of 331,151 emergency department records from a Thai hospital (2016–2021). The study evaluates complete case analysis, zero imputation, mean imputation, k-Nearest Neighbors (kNN) imputation, and MissForest, combined with logistic regression, decision tree, random forest, Light Gradient Boosting Machine (LightGBM), and Extreme Gradient Boosting (XGBoost). The results indicate that XGBoost with zero imputation delivers the best performance, achieving an accuracy of 0.8659, precision of 0.8726, recall of 0.8659, F1-score of 0.8681, and an AUC ranging from 0.8848 to 0.9947 across eight prediction classes. Furthermore, tree-based models demonstrated greater stability across different missing value handling methods, whereas linear models were more sensitive to imputation techniques. These findings suggest that strategic selection of missing data handling approaches can significantly enhance the reliability of mortality predictions in emergency care settings.

A genetic-neural optimization approach for friction stir spot welding of semi-solid metal Aluminum Alloy 5083

2025-06-29T16:16:24+07:00

Friction Stir Spot Welding (FSSW) of Semi-Solid Metal (SSM) Aluminum Alloy 5083 poses challenges due to nonlinear interactions between process parameters and mechanical properties. Traditional optimization methods, such as Response Surface Methodology (RSM), provide statistical modeling but often fail to capture these complexities accurately. This study integrates Artificial Neural Networks (ANNs) with Genetic Algorithms (GAs) and Response Surface Methodology (RSM) to develop a hybrid optimization framework for FSSW parameter selection, aiming to enhance weld strength and hardness while minimizing the number of experimental trials. The ANN model, trained using a feed-forward backpropagation algorithm with the Levenberg-Marquardt learning rule, predicts tensile shear strength and weld hardness based on key parameters: rotational speed, travel speed, and dwell time. GA optimizes these parameters through an evolutionary search, while RSM validates the results and assesses parameter interactions. The optimized parameters 2143.93 RPM, 14.33 mm/min, and 6.58 s yield a shear strength of 5999.99 N. ANN exhibited lower mean absolute error (MAE) and root mean squared error (RMSE) than RSM, confirming superior predictive capability. However, RSM provided statistical validation, ensuring robust insights. The findings highlight the effectiveness of AI-driven optimization in welding applications, reducing experimental trials while ensuring optimal mechanical performance. Future research should explore the integration of deep learning and real-time sensor feedback for further enhancement.

Application of the HEC-HMS model for analysing land use change and hydrological responses across different return periods in tropical flood-prone areas

2025-06-18T18:46:02+07:00

Land use change is a significant environmental concern worldwide today, as it has the potential to increase the frequency of natural disasters, such as floods. The Sg. Segamat Watershed, which is particularly vulnerable to flooding, highlights the importance of hydrological modelling as a crucial tool in disaster mitigation. In this study, the Hydrologic Engineering Center’s Hydrologic Modelling System (HEC-HMS) was utilised to assess the effects of land use change on hydrological responses across various return periods. The analysis examined both the pre- and post-calibration phases under varying land use conditions. Land use data from 2006 and 2011 were used to simulate future scenarios. The findings showed that the expansion of built-up areas and the conversion of forested land to mixed agriculture had a significant influence on flood patterns. Specifically, built-up areas expanded by 2.24%, resulting in increased flood volumes in sub-basins 4, 9, and 12 between 2006 and 2026. Concurrently, forest cover declined by approximately 4.40%, which led to heightened flood peak heights in sub-basins 1, 3, and 11 under all land use conditions. Sub-basin 3 recorded the highest flood peak height, estimated at 1,150.20 m³/s in the pre-calibration phase, 1,165.80 m³/s in the post-calibration, and 1,036.20 m³/s using the initial CN with calibrated parameters. Meanwhile, sub-basin 4 demonstrated the highest flood volume, with estimates reaching 342.10 mm pre-calibration, 366.09 mm post-calibration, and 341.04 mm using initial CN with calibrated parameters. These results clearly demonstrate how land use changes influence hydrological behaviour, emphasising the need for watershed planning and flood risk management. The study highlights the value of hydrological modelling as a tool for enhancing flood mitigation strategies and provides crucial insights for policymakers, planners, government agencies, and local communities.

Enhanced biogas production through co-digestion of tapioca starch wastewater and duckweed in a continuous stirred tank reactor

2025-06-04T17:01:37+07:00

This study investigates the biogas production performance of co-digesting duckweed with tapioca starch wastewater (TSW) in a laboratory-scale continuous stirred tank reactor (CSTR). Duckweed, with its rich nutrient composition, represents an underutilized biomass resource for renewable energy production in Thailand, where the tapioca starch industry constitutes a significant economic sector. The experimental setup utilized a 3-liter CSTR operated at mesophilic conditions (35°C) with a hydraulic retention time (HRT) of 28 days. Initial mono-digestion of TSW at an organic loading rate (OLR) of 0.31 gCOD/L-d resulted in a specific methane production of 0.28 NL-CH₄/g COD removed (NL = liter of gas at 273 K and 1 atm). Subsequent co-digestion with duckweed (1.0 g dry weight per liter of TSW) under identical operational conditions, enhanced methane production to 0.35 NL CH₄/g COD removed—corresponding to a 1.3-fold increase in specific methane production yield. These findings demonstrate that co-digestion of duckweed with TSW significantly enhances methane yield compared to mono-digestion of TSW, offering a promising approach for simultaneous wastewater treatment and renewable energy generation in Thailand's tapioca processing industry.

Chitosan beads with Calamondin essential oil: Antioxidant, antimicrobial, physical, and structural properties

2025-09-16T15:23:17+07:00

Chitosan, a biopolymer renowned for its biodegradability and antimicrobial properties, is frequently enhanced with essential oils (EOs) to improve functionality. This study investigated the impact of incorporating calamondin EO (CmEO) into chitosan beads on their physicochemical properties, antioxidant activity, antibacterial efficacy, and structural characteristics. The incorporation of CmEO improved the sphericity of the beads, ranging from 68.20% to 87.59%, and produced 177–263 beads per 10 mL of solution, depending on the formulation. The addition of CmEO to chitosan beads significantly boosted antioxidant activity, with the highest concentration exhibiting radical scavenging activity of 77.96%. Antibacterial assays revealed that the chitosan beads with CmEO could inhibit the growth of Escherichia coli, Bacillus cereus, Salmonella enteritidis, and Staphylococcus aureus. Structural analysis shows significant changes on the bead surface corresponding to changes in CmEO concentration. These findings highlight the potential of chitosan-based materials enriched with EOs for sustainable food preservation.

Assessment of surface water and groundwater potential under climate change in the Lam Phaniang River Basin

2025-04-18T11:31:21+07:00

The assessment of surface water and groundwater potential under climate change in the Lam Phaniang River Basin was based on SWAT model for evaluating streamflow, and MODFLOW model for evaluating groundwater flow. The SWAT model was well calibrated and validated with daily discharge measured at E.68A during 2010-2015 and 2016-2021, respectively, with R² and Nash-Sutcliffe Efficiency values more than 0.60. The MODFLOW model was also well calibrated and validated with observation data from 16 groundwater wells during May 2021, and 2 groundwater observation wells of Department of Groundwater Resources during 2013-2021, respectively, with r greater than 0.95 and Normalized Root Mean Square Error less than 10%. Future climate analysis (2022-2099) was based on Regional Climate Models (CNRM-CM5, CanESM2, and GFDL-ESM2M), under RCPs 4.5 and 8.5 scenarios. The maximum and minimum temperatures under RCP 4.5 were 34.54 °C and 21.07 °C, respectively, while under RCP 8.5, both temperatures were 35.30 °C and 20.36, respectively. The future mean temperature tended to be higher than present temperature (32.35 °C and 19.28, respectively). The future mean annual rainfall, which were 1,246.37 mm/year and 1,250.01 mm/year under RCPs 4.5 and 8.5, respectively, were lower than the mean annual rainfall recorded between 2002-2021 (1,257.00 mm/year). The surface water under RCPs 4.5 and 8.5 were lower than the present condition (2,175,988,582 m³/year), while the future groundwater supply was increased from the present (424,418,714 m³/year). When annual surface water supply was compared with water demand, no water shortage was expected under present condition, while low to moderate levels of water shortage were identified under RCPs 4.5 and 8.5. When compared annual surface and groundwater supply with the demand, no water shortage was detected under present and future conditions. Finally, the obtained results will be useful for surface and groundwater management, in which water-related problems can sustainably be solved.

Predictive maintenance of industrial milling machine using machine learning techniques

2025-08-19T15:56:04+07:00

Condition-based predictive maintenance of industrial machinery is a key area of research in the present world looking towards Industry 4.0. Machine learning (ML) techniques can have tremendous impact in this aspect because of their robust predictive modeling capabilities. The present paper aims to determine the optimized machine learning technique for the predictive maintenance of an industrial milling machine. The data pertaining to the operating parameters and the failure types of the machine is obtained from a public dataset with 10,000 data points. Five of the most popular classification ML algorithms namely, Artificial Neural Network (ANN), Discriminant Analysis (DA), Naïve Bayes (NB), Support Vector Machine (SVM) and Decision Tree (DT) techniques are implemented for the dataset to determine their optimized hyperparameters for an effective prediction of the machine failure type. DT and ANN were found to be the two best techniques with overall accuracy of 99.15% and 98.8%, respectively, and superior performance metrics of Precision, Recall and F-Measure compared to the other models. The results obtained from the present study may be enriched in the future by incorporating deep learning-based models and hybrid ML and intelligent optimization techniques for effective predictive maintenance of various industrial systems. The present approach can thus be employed in real-time factory settings to realize the targets of Smart Manufacturing and Industry 4.0.

Performance evaluation of travel demand forecasting models for transportation network analysis

2025-06-08T18:31:50+07:00

This paper presented a performance evaluation of travel demand forecasting techniques on transportation networks in Upper Northern Provincial Cluster 2, Thailand. The study compared multiple regression analysis and four-step sequential decision models. The findings revealed that the four-step sequential decision model forecasted person-trip generation in the study area to be 346,506, 373,422, 404,356, and 440,132 person-trips/day for the years 2029, 2034, 2039, and 2044, respectively. In comparison, the multiple regression model predicted approximately 320,245, 328,678, 338,123, and 349,567 person-trips/day for the same years, showing differences of 8.20%, 13.61%, 19.59%, and 25.91%, respectively. This variation can be attributed to the four-step sequential decision model's superior capability in comprehensively considering the impacts of future infrastructure development projects in the area compared to the multiple regression model. While both models forecast total person-trip generation, the four-step model additionally provides spatial distribution, modal allocation, and network assignment of these trips, enabling detailed analysis of traffic volumes on specific corridors. However, when evaluating model development convenience and time requirements, the multiple regression analysis approach offers faster problem-solving capabilities due to its more straightforward development process, while providing reasonably accurate forecasts of total person-trips.

Evaluation of the impact of tool geometry on thrust force and bearing strength of ramie woven composite

2025-06-08T18:29:33+07:00

This research evaluates the impact of tool geometry on thrust force, bearing strength, and failure modes in ramie woven composite materials. Four types of tool geometries were tested: Brad and Spur Drill (BSD), Twist Drill (TWD), End Mill Centre Cut (ECC), and End Mill Centre Hole (ECH). The thrust force measurement results show that the ECC device generated the highest thrust force of 148.17 N with the lowest bearing strength of 78.00 MPa. Conversely, the lowest thrust force was observed in the TWD device at 60.52 N, which coincided with the highest bearing strength of 98.54 MPa. This condition indicates that materials with higher bearing strength tend to be stiffer, reducing the amount of thrust force that can be effectively transferred. The failure modes observed after the bearing tests indicated that all specimens experienced net tension, with damage variations depending on the tool geometry used. These results directly benefit industries utilizing ramie composites, such as the automotive, aerospace, and environmentally friendly manufacturing sectors. By reducing material damage during the machining process and improving mechanical efficiency, this research can help enhance final product quality, extend components' service life, and reduce production costs.

Enhancing the energy efficiency of traditional brick kilns through sustainable insulation with rice husk ash and wood ash

2025-07-29T11:32:30+07:00

This study investigates the use of rice husk ash (RHA) and wood ash (WA) as sustainable thermal insulators to enhance the energy efficiency of traditional brick kilns. As readily available byproducts of agricultural and biomass combustion processes, RHA and WA are low-cost materials that support circular economy practices. Their favorable physical properties, including low thermal conductivity, high porosity, and reactive silica content, make them suitable for use as insulation for kiln walls. Experiments were carried out using a scaled-down downdraft open-top kiln to evaluate the thermal performance across various wall thicknesses and compaction levels. The results indicated that WA, particularly in a 15 cm loose-fill configuration, achieved the lowest heat loss (23.24 MJ) and the highest thermal efficiency (54.42%), representing a 15.09% improvement compared to the control. This approach to insulation also reduced the unit energy cost per brick from 0.00487 to 0.00414 USD, yielding an estimated annual fuel saving of 588 USD and a payback period of 1.47 years (~20 batches). In addition to energy savings, the reuse of RHA and WA reduces landfill waste, mitigates reliance on virgin insulation materials, and contributes to emission reductions, potentially lowering the CO₂ output by up to 750 kg per year for small-scale kilns. These findings confirm that incorporating RHA and WA into kiln construction is a viable, cost-effective strategy for improving sustainability in artisanal and semi-industrial brick production. The results are scalable and adaptable to various geographical and climatic contexts, thereby supporting broader adoption in developing regions.

Characterisation and quality improvement of binder free bio-pellets from the sugar industry residues and grass jelly food wastes for energy purposes

2025-05-01T11:03:07+07:00

Torrefied pelletised biomass wastes can be a sustainable and efficient solid fuel; however, the addition of binders is frequently required to improve the consistency, durability and overall quality of the pellets. Such additional processes can increase production costs, reduce stability on exposure to moisture, increase ash content and may reduce the heating value of the fuel. Therefore, the production of durable binder free bio-pellets would create a sustainable and economically viable route for producing these solid fuels. Herein, the binder free pelletisation of waste sugarcane leaves and/or grass jelly leaves/stalks from the agricultural industry was investigated for use as bio-based solid fuels. Importantly, the resulting pellets had a diameter, length, moisture content, unit density, bulk density, calorific value and durability that met the international standards (DIN 51731, SS 187120 and CEN/TS 14961) and Thailand standard (TIS 2772–2560). However, the ash content for grass jelly leaves/stalks pellets was 16%, while the blended pellets of 50:50 sugarcane:glass jelly was 7%, both of which were higher than that required by the Thailand standard. The calorific value of the pellets was improved via a torrefaction process at 200 – 300^oC for between 15 – 60 min. Yields of pellets dropped with increasing temperature and residence time, while the calorific value increases as the torrefaction severity increases from 16,630 – 26,334 kJ/kg. Crucially, this is the first reported pelletisation and torrefaction of sugarcane leave, grass jelly leaves/stalks or a 50:50 blend of these wastes to yield bio-based fuels with calorific values comparable to coal-like fuel pellets. Moreover, the mechanical strength of obtained pellets was still maintained without the use of an additional binder, thus reducing additional processing steps and potential cost. The optimal operating conditions for torrefaction were 250°C for 30 min, resulting in the greatest integrity, calorific value, enhancement factor and energy yield.

Development of water hyacinth reinforced jackfruit-seed-starch bi-layerd composites for sustainable thermal insulation

2025-06-20T13:28:06+07:00

This study investigates the development of biodegradable composite materials using water hyacinth pulp and jackfruit-seed-starch as a binder, aimed at providing an eco-friendly solution for thermal insulation applications. The composites were fabricated using a compression molding process with varying starch contents (10%, 20%, 30%, and 40%) under controlled temperature and pressure. Mechanical properties such as tensile strength, flexural strength, impact strength, hardness, and thermal conductivity were evaluated. The results indicated that composites with 30% jackfruit seed starch exhibited the best mechanical performance, including tensile and flexural strength, along with favorable thermal conductivity. However, water absorption remained a challenge, with higher starch content leading to increased moisture uptake. The findings highlight the potential of these composites for thermal insulation, particularly in extremely cold countries, where they could serve as a sustainable alternative to conventional materials. Further studies are needed to reduce water absorption and enhance the durability of the composites for long-term use.

Effect of curing methods on compressive strength of pervious concrete containing silica fume and calcium carbonate

2025-03-31T16:41:39+07:00

Concrete curing plays a critical role in the development of compressive strength, particularly in pervious concrete, which is highly susceptible to moisture loss due to its porous structure. This study investigates the effects of different curing methods on the compressive strength of pervious concrete and examines how the incorporation of silica fume (SF) and calcium carbonate powder (CC) influences the curing sensitivity index (CSI). Experimental results indicate that water curing consistently yields the highest compressive strength across all pervious concrete mixes at 7 and 28 days, followed by plastic and air curing. The presence of silica fume increases CSI, making pervious concrete more dependent on curing conditions, particularly under air curing. In contrast, calcium carbonate powder reduces CSI, enhancing curing efficiency and mitigating sensitivity to curing variations. Notably, a ternary blend of silica fume and calcium carbonate significantly lowers CSI at early ages, indicating improved curing resilience. However, at 28 days, the effect of CC in mitigating curing sensitivity diminishes slightly, while SF continues to increase curing dependency. These findings suggest that optimal curing strategies should be tailored to the binder composition, with calcium carbonate powder proving effective in stabilizing curing sensitivity. The results contribute to developing more durable and sustainable pervious concrete mixes with enhanced performance under variable curing conditions.

Optimizing the pork supply chain: A model integrating feed production and pig farming with outsourcing and subcontracting costs

2025-02-24T14:31:18+07:00

This study introduces a mathematical model designed to optimize the vertically integrated pork supply chain by addressing key cost factors, including pig farming, feed production, and outsourcing. The model integrates pig fattening and feed production stages, incorporating essential cost components to synchronize farming schedules with feed production plans while minimizing total costs. Computational experiments, using data from an empirical study of Thailand's vertically integrated pork supply chain, were conducted to evaluate the model's efficiency and sensitivity under varying farm sizes and planning horizons. The results demonstrate that the model effectively identifies optimal solutions for shorter planning periods (up to 14 months). However, extended planning horizons and larger farm sizes significantly affect solution times and the quality of feasible solutions. These findings provide valuable insights and practical strategies for pork production companies seeking to enhance cost efficiency and improve supply chain sustainability. Future research should focus on developing advanced heuristics and exploring additional supply chain dynamics within integrated environments.

Ensemble machine learning-based PM2.5 modeling using hotspot counts (0-1000 km) reflecting Chiang Mai, Thailand’s extreme pollution

2025-05-07T13:25:58+07:00

Persistent local and transboundary smog has critically elevated PM2.5 levels in Northern Thailand over the past decade, resulting in significant health risks. The spatial distribution of hotspot counts, indicative of biomass burning and smoke dispersion, demonstrates a strong correlation with PM2.5 concentration patterns, underscoring the importance of incorporating such data into air quality analyses. This study integrates hotspot data to capture both temporal dynamics and external influences in PM2.5 prediction models. The importance of lagged hotspot counts within 100–1000 km of Chiang Mai—ranked as the world’s most polluted city during the study period—and lagged ground-level PM2.5 is assessed using Lasso regularization. The analysis reveals that the cumulative effects of hotspots extend their influence on air quality in Chiang Mai up to approximately 700 km. Advanced tree-based ensemble machine learning methods, including Random Forest (RF), Gradient Boosting (GB), and Extreme Gradient Boosting (XGBoost), are implemented alongside the Long Short-Term Memory (LSTM) deep learning model to evaluate their predictive performance. This approach provides a novel framework for PM2.5 modeling in Northern Thailand. Five key features with specific day lags were identified for modeling. These include PM2.5 at lag 1, short-range hotspots within 100 km at lags 1 to 3, mid-range hotspots at 200 and 400 km at lags 2 to 4, and long-range hotspots beyond 700 km at lag 5. Incorporating hotspot data improved model performance by approximately 20%, as evidenced by error metrics and residual analysis. Among the models tested, GB outperformed XGBoost, RF, and LSTM, achieving the highest R² (0.97), lowest RMSE (5.49), MAE (2.08), and MAPE (5.8%), along with near-zero MBE and minimal MdAE (0.48). Statistical validation confirmed the model’s reliability with no significant bias.

Banana quality classification using lightweight CNN model with microservice integration system

2025-04-08T14:40:55+07:00

Banana sorting has been performed manually, which often leads to human error due to the high volume and diverse characteristics involved. This paper presents a banana quality classification system using ConsolutechMobileNetV2 (CST-MobileNetV2) to classify banana ripeness into four categories unripe, ripe, overripe, and rotten. A lightweight deep learning model is proposed and integrated with a uniquely designed microservice system to optimize performance while minimizing computational demands. A publicly available dataset containing 13,478 images was used, and the data split into 56% for training, 14% for validation, and 30% for testing. Image normalization and augmentation techniques were applied to enhance the model's robustness. The model's performance was evaluated using a confusion matrix, achieving 98% precision, recall, and F1-score. The proposed model was compared with other deep learning models to benchmark its performance and deployed in different operating systems to evaluate its flexibility and capabilities. The LINE platform was employed as the user interface, enabling practical interaction with users. The system also demonstrated an average response time of 9.25 seconds per image, ensuring efficient processing, delivers high accuracy and scalability making it a practical and efficient solution for automated banana quality classification.

Modeling standard lines for soil compaction testing using artificial neural networks and geometric algorithms

2025-05-13T15:17:30+07:00

Soil compaction testing is crucial for ensuring the stability and durability of infrastructure projects. Traditional methods for generating standard lines, such as averaging and polynomial fitting, often fail to capture the nonlinear relationships and variability in compaction data, leading to inaccuracies in soil property assessments. This study introduces a novel framework that leverages Artificial Neural Networks (ANNs) to dynamically model standard lines for compaction curves, addressing limitations of traditional approaches. Four activation functions—ReLU, Sigmoid, Tanh, and Swish—were evaluated, with Swish emerging as most effective for capturing complex relationships between Dry Density (DD) and Moisture Content (MC). A tolerance-based evaluation framework, incorporating tolerance levels of 2%, 5%, and 10%, was applied to analyze coverage areas. The 5% tolerance level was identified as most balanced, minimizing errors while providing reliable representations of compaction data. The study also introduced the Ray-Casting Algorithm for precise calculation of coverage areas, enabling a new performance indicator based on density of data points within the region. Results demonstrate that the ANN framework, particularly with Swish activation, outperforms traditional statistical methods in accuracy and adaptability. ReLU delivered the best performance, with the lowest prediction and percentage errors (0.1910–0.2005 and 10.66%–11.74%), while effectively explaining over 55% of the data’s variability. Sigmoid showed the weakest results, with high errors and near-zero variance explanation. Tanh performed moderately, balancing accuracy and generalization with reasonable error levels and 44%–54% variance capture. Swish was consistently reliable, with stable errors and over 50% of the variance explained. This research advances compaction testing by addressing variability, operator-induced errors, and nonlinear data patterns, establishing a reliable methodology for generating standard lines. Future work could explore diverse soil types, integrate environmental factors, develop hybrid machine learning models, and improve performance indicators.

Flexural performance of reinforced concrete beams used shredded scrap tire rubber and steel fibers

2025-07-01T15:05:43+07:00

This research experimentally investigates the flexural performance of fibrous reinforced concrete beams containing shredded scrap tire rubber (SSTR) as a substitute for gravel. Six reinforced concrete (RC) beams (1500 × 300 × 200 mm) were prepared with varying steel fibers (SF) (0%, 0.5%, 1%, and 1.5%) and SSTR (0%, 5%, 7.5%, and 10%) by volume of concrete. All samples were tested as simply supported beams under 3-point static loads. The RC beam with natural materials (0% SSTR and 0% SF) exhibited a typical crack propagation pattern, while the addition of 1% SF and 5% SSTR caused cracks to cease, resulting in ductile behavior. The optimal percentages were found to be 1% SF and 5% SSTR. The presence of SSTR reduced the compressive strength due to the impermeability of rubber, which helps absorb load energy, though SF additions improved this. Compressive strength reductions for 5%, 7.5%, and 10% SSTR were 15.37%, 11.64%, and 18.08%, respectively, compared to the control mix. However, the compressive strength of concrete containing 1% SF and 5% SSTR increased slightly by about 2.52%. Flexural strength of concrete prisms also decreased with higher SSTR content and varied SF dosages. Compared to the control mix, reductions in flexural strength for 0.5% SF with 5%, 7.5%, and 10% SSTR were 14.29%, 19.05%, and 19.05%, respectively. These reductions are due to the poor bond between SSTR and the cement matrix. The flexural performance of the reinforced concrete beam improved slightly by 1.71% for the B5 beam, which was made with 5% SSTR and 1% SF, accompanied by a slight increase in deflection (2.24%) and beam weight. The design ultimate loads from BS8110 were lower than the experimental values, with ratios of tested failure load to design ultimate load ranging from 1.99 to 2.78, the maximum ratio achieved by the B5 beam.

Safety of a bridge span with prestressed concrete beams

2025-03-26T09:57:34+07:00

This article assesses the structural performance of a bridge overpass with a new type of reinforced concrete prestressed T-beams (TBN - abbreviation for stressed T-beams). Compared with VTK bridge beams (VTK - abbreviation: high-tech transportation construction), prestressed T-beams have technological advantages in the production process and during installation that allow for increased spacing, from 1400 mm for high-tech transport structure type beams to 2200 mm for prestressed T-beams. The study aims to evaluate the reliability of 24-m long prestressed T-beams and operation of their joints as part of a bridge overpass. Static testing of a bridge overpass was carried out using vehicles, with measurement of true deflections of the experimental beams and investigation of the operation of their joints as a part of the overpass. The calculated deflection values are determined using the finite element method in the Midas Civil software package. Test results showed the true and calculated deflections. The structural coefficients for the first and second test schemes were 0.79 and 0.81, respectively. This meets structural requirements. The structural coefficient is a criterion corresponding to the required load capacity for reliable operation of a structure under various force combinations.

Synthetic flood damage function for direct damage estimation in Loei Town Municipality

2025-01-12T21:30:43+07:00

The reflection between flood damage phenomena and flood characteristics in the flood-affected areas of Loei Town Municipality where flood damage issue is not deeply examined and documented, is desirable for a more accurate damage estimation. Therefore, based on the available empirical dataset collected during 2021-2023 floods, the site-specific flood damage functions and their curves were developed for assessing direct monetary damage to buildings. The replacement cost for household damaged contents was gathered through survey interview of 637 households, in which 75% and 25% of the entire dataset were randomly split for constructing and validating synthetic functions, respectively. The polynomial function was the best fitting method, rather than the other five damage functions (i.e., exponential, Fourier, Gaussian, Rational, and the Sum of Sines), as characterized by the highest R² values of 0.73, while relatively low values of MAE (0.17), MBE (-0.17), and RMSE (0.19) clearly indicated the validity of the synthetic damage function. All relevant data were then entered into the HEC-FIA software for damage estimation based on a structure-by-structure basis. The results revealed that the 2002 flood caused 199,330 USD in damage to Loei Town Municipality, which showed a reasonable agreement with the governmental relief budget (accounted for 56% of the 2002 total flood-relief budget of Loei Province). It is noteworthy that the findings gained from this study may be of assistance when assessing flood damage to buildings for the areas with flood-related data scarcity, in which the depth-damage function/curve developed herein could help pave the way towards more accurate flood damage estimation for risk assessment at local scale. Finally, this study could be technically beneficial for government and local authorities of Loei Town Municipality in making decisions in reducing flood damage and for residents living in flood-prone areas to achieve resilience after flood events.

Enhancing sustainability and key success factors in digital food supply chain management through digital transformation: A fuzzy AHP approach

2025-02-28T10:26:13+07:00

This study investigated the key success factors in digital food supply chain management for Thailand's food retail industry through digital transformation, employing the Fuzzy Analytic Hierarchy Process (FAHP) for the integrated digital food supply chain model. Through an extensive literature review and expert consultations, the critical criteria were identified and ranked in the researched domains of digital strategy, digital operations management, digital customer experiences, and digital organization and culture management. Business alignment is the primary focus within digital strategy, followed by technology investments to enable this alignment and data-driven decision-making. Supply chain visibility is considered the most crucial aspect of digital operations management, and data analytics along with inventory management are integral aspects that lead to operational efficiency. In the realm of digital customer experiences, aspects of customer centricity, convenience and accessibility, and order tracking and communication are the top drivers of satisfaction and loyalty. Finally, leadership and vision are recognized as essential for digital organization and culture management, along with fostering digital skills, capabilities, agility, and adaptability. In addition, this study provides a practical framework for improving Thai food retail organizations through the alignment of engineering, organizational, and consumer expectations.

Durability and mechanical characterization of tapiales with lime and sugar cane fiber

2025-06-15T18:06:09+07:00

Worldwide, technological advances in construction have transformed construction materials, highlighting earth for its advantages. The traditional rammed earth method combines sustainability, efficiency and the creation of durable and attractive structures. Therefore, the objective of this research was to study the durability and mechanical characterization of rammed earth by adding lime and sugar cane fiber. Using an applied methodology and experimental design, preparing samples with the addition of lime in doses of 5%, 10%, 15%, 20%, obtaining an optimum to combine it with sugar cane fiber in doses of 0.5%, 1%, 1.5% and 2%. The fiber was 40 mm long and was randomly placed in the dry rammed soil mixture. The results indicated that the optimal percentage was 15% lime + 1.5% sugar cane fiber, since adding lime increases its mechanical properties and adding sugar cane fiber helps in durability analysis, improving by 10% compared to the initial samples.

Effect of warm deformation parameters on hardness and microstructure of AISI 1020 low carbon steel for near-net shape forging

2025-01-21T14:36:56+07:00

This research aims to present a concept for altering a metal manufacturing process from cold to warm forging thereby reducing unnecessary steps and energy consumption. This will lower costs and increase production profits. The study explores the impact of warm forging process parameters on the hardness and microstructure of low-carbon steel for near-net-shape forging in more than two continuous stages. The material used in this experiment is annealed AISI 1020 carbon steel with chemical additions of 0.01% Ni, 0.03% Cr, and 0.044% Al. The study procedure involves: (i) heating slugs with a height-to-diameter ratio (ho/do) of 2.07 to a temperature range of 200–700 and soaking them for 1 hour. The grain size noticeably increases at temperatures above 500 °C. (ii) The materials were forged at six different temperatures from 200 to 700 with both hardness and microstructure examined at each stage. This was done to determine the recrystallization temperature. The experimental results showed that recrystallization begins at 500 in a warm forging process and becomes more pronounced at 600 to 700 °C. The lowest average hardness value in the transverse direction (TD) occurs at 500 to 700 °C. This suggests that the suitable warm working temperature range should be below 500 °C, as the primary microstructure in the forging process has not yet undergone recrystallization. Our research provides valuable insights for manufacturers aiming to transition from cold to warm forging, emphasizing the importance of precise control over deformation parameters to achieve desired material properties.

Sediment transport analysis in water management system using optimized neural framework

2025-02-27T14:19:12+07:00

In a water management system, sediment prediction is considered a complex process. Conventional sediment prediction techniques are less precise. The sediment predictions through artificial intelligence analysis possess more prediction characteristics than the conventional approaches. One of the divisions of artificial intelligence is the deep learning technique. In this investigation, the deep learning technique is combined with the neural network approach for predicting the sediment parameters of the water management system using a novel Hyena Deep Neural Sediment Framework (HDNSF). Prime contributions are initially, the parameter of the water management system is provided as input to the approach, then the shear stress and transportation stage of the sediments are determined and the outcomes are generated. Thereafter, the mean velocity of the sediment particles, total sediment load, and the rate of total transported sediment are determined for the river system. The outcomes of the water management system are predicted to consist of transported sediment load and rate of transported sediments. Along with that, the determination coefficient of the prediction system is also evaluated. The outcomes of the prediction system and the determination coefficient of the prediction system are compared with recent studies such as Ant Colony Optimization Fuzzy Inference System (ACOFIS), Adaptive Neuro-Fuzzy Inference System (ANFIS), Least Square Support Vector Machine (LS-SVM), and Group Method of Data Handling (GMDH). A plain river water management system is analyzed, including its characteristics. At the end of fifth month, peak sediment load transport and sediment rate was recorded. The peak sediment load was 124695 tons/day and the peak sediment transport rate was 19636m³/s. The coefficient of determination of the proposed HDNSF technique is 0.982.