Engineering Journal (Faculty of Engineering, Chulalongkorn University, Bangkok)
Not a member yet
1223 research outputs found
Sort by
Optimizing Nickel-Cobalt Electrodeposition for Sustainable Agricultural Equipment: Wear Resistance Enhancement through Multi-Response Optimization
Full text linkThe rapid wear of sugarcane harvester blades, caused by continuous abrasion and impact, leads to increased operational costs and reduced efficiency in the sugarcane industry. To address this issue, this study focuses on optimizing nickel-cobalt (Ni-Co) electrodeposition parameters to enhance the wear resistance and service life of harvester blades. While individual parameter adjustments have been studied, a comprehensive multi-response optimization approach for agricultural machinery remains underexplored. This research applied the Grey-Taguchi method to optimize two key parameters: current (0.02 A, 0.04 A, 0.06 A) and cobalt volume (10%, 20%, 30%). A Taguchi L9 orthogonal array was used to design the experiments, while Grey Relational Analysis (GRA) integrated multiple responses to minimize wear volume and control coating thickness. The effects of the parameters were further analyzed using Analysis of Variance (ANOVA) to assess statistical significance. Results showed that current had the most significant impact on wear resistance, contributing 96.72% to the total variation, while cobalt volume played a lesser but still important role. The optimized parameters (0.04 A current and 10% cobalt) led to a substantial reduction in wear volume (0.00177 mm³) and improved coating performance. Confirmation experiments validated these findings. The study demonstrates that the Grey-Taguchi method is an effective approach for optimizing Ni-Co coatings in agricultural machinery, leading to improved blade durability and reduced maintenance costs. Future research should explore additional process parameters and assess the environmental impact of these coatings
Political Spectrum Classification using Natural Language Processing
Full text linkThe political party is a group of politicians who represent the citizens. In particular, each party represents a set of ideologies and policy goals. The difference in ideologies and policy goals of a party can be observed as a relative position compared with the others, which is generally called the political spectrum. However, politicians do not always uphold their ideology at the representative level and the political party level. Therefore, this study aims to identify the political spectrum of political parties using voting results from voting sessions. Then, natural language processing (NLP) is used to classify political speeches from the Minutes of the Sitting sessions into the political spectrum groups. The classifiers are then used to predict whether a politician belongs to which group. The clustering result shows that political parties in Thailand can be clustered into three groups based on the voting session. However, the two of the three groups are very similar; therefore, this work also considers the two-group configuration. With the three-group configuration, the best classifier can archive the F1-score of 67.72%. However, with the two-group configuration, the evaluation shows a better result, with the best F1-score of 73.92%
Estimating Parking Building Spiral Ramp Capacity Using Traffic Microsimulation Technique
No full textThis study aims to determine the traffic flow capacity on spiral ramps within a multi-story car parking building. A spiral ramp flow capacity is a crucial element in parking building design when considering traffic flow efficiency and safety. This study applies a microsimulation modeling analysis technique to estimate the traffic flow capacity on the spiral ramp. In the methodology, modified driving behavior variables are used in the traffic model to reflect the driving behavior within the chosen parking building in Bangkok, Thailand. For planning purposes, having the knowledges of obtaining a robust spiral ramp’s capacity simplifies the process of calculating the number of spiral ramps required for multi-story large-scale parking buildings with improved economic utility, safety, and efficiency in traffic flow. According to this study driving speed and width of the ramp directly affects the traffic flow capacity of the ramp. The spiral ramp’s traffic flow capacity of the small, medium, large, and extra-large width spiral ramp is 836 vehicles per hour, 976 vehicles per hour, 1,055 vehicles per hour, and 1,165 vehicles per hour respectively
Virtual Reality of Labless Welding for Next Industrial Training
No full textThis research presents an innovative Virtual Reality (VR) application designed to replicate the core steps of welding operations for training purposes. The primary goal is to assess whether VR can effectively replicate key welding procedures, such as safety protocols, workpiece preparation, electrode selection, and welding execution, in a virtual environment. The development process involved the integration of multiple technologies, including the Unity engine and 3D modeling, based on extensive research. This study focuses on validating the accuracy of the VR simulation by comparing it with real-world welding processes. The VR welding application provides a safe, accessible, and repeatable method of learning, overcoming geographical and resource limitations. With its potential for future enhancements, such as refining task accuracy and integrating performance metrics, the application offers a dynamic solution in the evolving field of virtual education, particularly for hands-on skills training
Bench-Scale Synthesis of High-Surface-Area Two-Dimensional Siloxene as the Support for Dry Reforming Catalysts
Full text linkDeveloping better supports for catalysts is vital for improving the dry reforming of methane (DRM) process, which turns greenhouse gases into useful industrial products. Siloxene, a two-dimensional material made from silicon, is a promising new option because it has a high surface area, can be tuned for specific tasks, and is very stable. This study focuses on making high-quality siloxene quickly using a specific chemical method at 0°C. The siloxene we created was used to support a nickel (Ni) catalyst, and it performed much better than traditional silica (SiO2) supports. The best catalyst, 5%Ni/3h-Siloxene, achieved a high initial conversion of 96.3% for CO2 and 76.59% for CH4. Over a 10-hour test, the siloxene catalyst was remarkably stable, with only a tiny 2% drop in CO2 conversion. This happened even though it had a lot of carbon buildup (44.49%). In contrast, the standard Ni/SiO2 catalyst, which had only 4.19% carbon, had a much bigger performance drop of 5%. Our findings suggest that the type of carbon that forms is more important than the amount. This makes siloxene a very promising material for long-term DRM applications
Design and performance evaluation of a propeller-based peanut sheller: efficiency and Economic Feasibility Analysis
Full text linkThis study focuses on the design and performance evaluation of a propeller-based peanut sheller intended to assist farmers with limited labor resources by enhancing shelling capacity and minimizing losses during the shelling process. The analysis focused on two key factors: rotor speed (RS) and concave clearance (CC). The sheller’s performance was evaluated based on shelling efficiency (SE) and grain breakage (GB). The results indicated that at an RS of 4.90 m/s (550 rpm), the sheller achieved the highest shelling efficiency of 88.05% and a GB rate of 9.70%. Additionally, adjusting the CC to 10 mm resulted in an SE of 87.98% with a GB rate of 10.94%. Economic analysis indicated that processing a minimum of 29,803.21 kg of peanuts per year would result to a payback period of 0.82 years (approximately 9.84 months). Future research should explore additional factors affecting the feeding rate and cleaning of the propeller-based peanut shelling machine
Hybrid Approaches to Machine Learning for Improved Battery Sales Forecasting: A Case Study in Thailand
Full text linkBattery sales forecasting is a critical component of demand planning in the automotive battery industry, directly influencing production, inventory management, and supply chain optimization. This study presents a comprehensive evaluation of traditional forecasting methods and machine learning techniques to predict monthly sales for a battery manufacturer in Thailand. Utilizing a dataset of monthly sales for the 10 best-selling products from January 2018 to December 2023, the research investigates the performance of traditional models such as Holt’s Linear Trend, Holt-Winters Seasonal, ARIMA, SARIMA, and SARIMAX. Advanced machine learning approaches, including Long Short-Term Memory (LSTM) networks and Artificial Neural Networks (ANN), are also explored. Additionally, hybrid models combining traditional and machine learning techniques are developed to leverage their respective strengths. The study integrates external factors such as economic indicators, industry-specific variables, and lagged data during feature selection to enhance predictive accuracy. Model performance is rigorously evaluated using Mean Absolute Percentage Error (MAPE). The results demonstrate that the hybrid ANN-LSTM model achieves the highest accuracy, with an average MAPE of 8.83%, significantly outperforming individual models, including the best-performing traditional model, ANN, at 9.43%. This research contributes to the field by providing a robust analytics framework that integrates traditional and advanced machine learning methodologies, offering actionable insights for battery sales forecasting and enhancing decision-making processes in the automotive industry
Development and PCA Evaluation of Lunar Mortar Compositions from Lunar Simulant and Potato-Based Materials
Full text linkSpace activities have taken another step, particularly on the Moon, where the establishment of a human exploration base has garnered interest from many space agencies. One of the challenges lies in constructing shelter and accommodation using locally available resources. This study investigates the lunar mortar compositions using Thailand lunar regolith simulants and potato-based materials, including potato starch and fibers, as potential binding agents. Extensive experiments optimized the mortar formulation by varying the ratios of Thailand lunar simulant (TLS-01), potato starch, and fresh/fermented potato fibers. Compressive strength tests evaluated the effects of fiber reinforcement, TLS-01 percentage, potato starch, and heat treatment. Microstructural analysis via SEM revealed the internal structure and cohesion. Principal Component Analysis (PCA) identified major influencing variables on compressive strength and their correlations. The LC-FrF-3 sample, with TLS-01 (39.47%), potato fresh fiber (7.9%), and freezing (-10ºC), exhibited maximum 0.65 MPa compressive strength. SEM showed specimens with dense cohesion and reduced voids, such as LC-FrF-3, had better strength. PCA highlighted 'TLS-01', 'Potato starch', 'Heat', and 'Freeze' as the most significant influencing variables. This research demonstrates the potential of lunar regolith simulants and potato-based materials for developing suitable lunar mortar for construction, contributing to in-situ resource utilization for space exploration
Reanalysis of Vertical Land Motion at Tide Gauge Stations in Thailand Utilizing GNSS Continuous Operating Reference Stations
Full text linkSea level monitoring is critical for coastal management, water resource planning, and climate change studies, particularly in Thailand, where agriculture forms the backbone of the economy. In Thailand, sea level observations primarily rely on tide gauge stations. However, tide gauge measurements are often influenced by vertical land motion (VLM), including land subsidence or uplift. To address this, the Global Navigation Satellite System (GNSS) offers a reliable solution for determining VLM. This study leverages the established network of Continuous Operating Reference Stations (CORS) in Thailand, utilizing co-located GNSS CORS with tide gauge stations in the Gulf of Thailand to quantify VLM at tide gauge stations. The VLM corrections were applied to tide gauge data to refine sea level estimates and provide insights into long-term sea level changes. The findings reveal that sea level changes corrected for VLM demonstrate discrepancies of approximately 4–5 millimeters when compared to sea level changes derived from satellite altimetry. This indicates that GNSS-derived VLM from the CORS network in Thailand is influenced by additional factors that may introduce biases in corrected sea level measurements. These results highlight the importance of addressing these influences to improve the accuracy of sea level monitoring and contribute to more reliable climate and coastal management strategies
Employing Neuroevolution of Augmenting Topologies (NEAT) in Linear Multi-Echelon Inventory Systems
Full text linkReinforcement learning has emerged as a leading algorithmic approach due to its successful applications across various domains. While many implementations favour the model-free approach for its aptitude for handling complex problems, its learning curve tends to be slower. Given the intricacies of the Linear Multi-Echelon Inventory System, a model-based approach might be more fitting, offering faster learning rates. This study seeks to integrate Neuroevolution of Augment Topologies (NEAT) – a hybrid of model-based reinforcement learning and evolutionary algorithms – into such an inventory system. Furthermore, the research delves into hyperparameter tuning, experimenting with seven specific hyperparameters to discern the most efficient combination and understand their interplay. Benchmarking against the model-free Proximal Policy Optimisation (PPO) serves as a measure of NEAT's effectiveness. Findings indicate that when optimally tuned, NEAT can slash total costs by 25.02% compared to PPO. Impressively, NEAT achieves this peak performance in a mere 1,000 generations, significantly outpacing PPO's learning trajectory