Technical University of Malaysia Malacca
Universiti Teknikal Malaysia Melaka (UTeM) RepositoryNot a member yet
10990 research outputs found
Sort by
Effect of filler size on the properties of oil palm empty fruit bunch high-load filler biocomposite
Get PDFThis work aimed to study the effect of filler size on the performance of an empty palm oil fruit bunch (OPEFB)
high-load filler epoxy resin biocomposite (80 vol.% OPEFB and 20 vol.% epoxy resin). The particle sizes of OPEFBs
used to prepare the biocomposites were 60, 80, 100, 120 and 140 mesh. The biocomposite samples were prepared by the
press method. The physical (density, porosity, thickness swelling), mechanical, and thermal properties of the
biocomposite were evaluated. A universal testing machine, thermogravimetric analysis, and scanning electron microscopy were utilized to characterize the biocomposite samples. The results show that the physical, mechanical, and thermal properties of the OPEFB epoxy resin biocomposite are significantly affected by the particle size of OPEFBs. As the particle size was reduced from 0.250 to 0.105 mm, the density improved from 0.974 to 1.101 g/cm3, the porosity decreased from 15.1 to 9.1 %, and the thickness swelling decreased from 14.2 to 7.4 %. The modulus of rupture and modulus of elasticity improved from 9.7 to 22.8 MPa and 1,667 to 2,403 MPa, respectively. Thermal analysis indicated that finer fillers enhanced thermal stability. The OPEFB biocomposite remained stable up to 300 °C. Smaller filler sizes exhibited improved biocomposite properties, which were attributed to better interfacial bonding and uniform dispersion within the matrix. The results of this study demonstrate that the performance of biocomposites can be significantly enhanced by reducing the size of the fillers. These findings indicate that filler size is crucial for high-load filler biocomposites
An efficient irrigation based on hargreaves potential evapotranspiration to improve yield for tomato plantation
Get PDFWith a tropical climate in Malaysia, varieties of vegetables can grow all year round. Nevertheless, during the hot season, watering the plant is challenging, especially for vegetables that are intolerant to heat such as tomato plants. Over-watering or under-watering could decrease the yield and quality of tomatoes. Therefore, in this study, we proposed an efficient irrigation system based on Hargreaves’s potential evapotranspiration to improve the yield and quality of tomato plants in Melaka, Malaysia. Using the Hargreaves equation, the correlation between the surrounding temperature and the amount of water needed by the tomato plants is investigated. Three growth stages are considered: early stage (0−30 days of planting), middle stage (31−76 days of planting), and final stage (77−105 days of planting). Based on a 30-day analysis, on average, tomato plants require 45.5 mL/day, 87.4 mL/day, and 60.8 mL/day respectively for the early, middle, and final stages of growth. A mobile monitoring application is also developed using MIT App Inventor for users to monitor
the temperature, humidity, soil moisture, pH level, status of water pumps, and the amount of water released to
the plants. The proposed system can increase the efficiency of the irrigation process and ultimately, reduce the farming cost
The influence of laser cutting parameters on the heat-affected zone in fast-growing Malaysian wood species
Get PDFWood is a naturally occurring renewable resource widely used in various industries, including in construction, packaging, furniture, and paneling. In Malaysia, 80%
of furniture products are made from wood, making it a crucial material in this sector. Laser cutting is an advanced machining technique that enhances precision and minimizes
material waste, yet its thermal effects, particularly the heat-affected zone (HAZ), remain a challenge. This study investigates how laser cutting parameters—including the laser power, traverse speed, and focus position—affect HAZ formation in two fast-growing Malaysian wood species, Acacia mangium and Azadirachta excelsa. This research seeks to determine the optimal laser settings that minimize HAZ dimensions while maintaining cutting precision. A diode laser cutting system was used to analyze the effects of three
laser power levels (800, 1500, and 2400 mW), three traverse speeds (2, 5, and 10 mm/s), and three focus positions (on-focus, +0.2 mm, and −0.2 mm). We employed statistical analysis, including a two-way ANOVA, to assess the significance of these parameters and their interactions (p < 0.001). The results indicate that a higher laser power and slower speeds significantly increase the HAZ’s width and depth, with Azadirachta excelsa exhibiting a greater HAZ width but shallower penetration compared to Acacia mangium. A slight above-focus position (+0.2 mm) reduces the HAZ’s width, whereas a below-focus position (−0.2 mm) increases the HAZ’s depth. The optimal parameters for minimizing HAZ dimensions while ensuring efficient cutting were identified as a 1500 mW laser power, a 10 mm/s
traverse speed, and an on-focus position (0 mm). This study provides practical insights into laser parameter optimization for tropical wood species, contributing to improved
precision in laser machining and sustainable wood processing practices. These findings support industries in adopting advanced, high-quality laser cutting techniques tailored to
fast-growing wood resources
Location independent human activity recognition using self-training CSI-based techniques for wireless sensor networks
No full textHuman activity recognition (HAR) using WiFi is applied across various domains ranging from smart environments, the Internet of Things (IoT) and immersive virtual gaming. The environmental effects of WiFi sensing lie in its susceptibility to variations in physical surroundings, which influence signal strength and accuracy in detecting human activity. Innovative solutions are needed to meet these demands, such as activity-adapted learning for seamless feature transfer
and recognition across various locations, reducing the reliance on extensive training datasets. This work proposes a framework incorporating a confidence threshold to filter unreliable samples, a progressive self-training strategy to integrate unlabeled data, and a weighted self-training approach to counter class imbalance. The proposed model explores HAR and its improved performance by integrating self-training techniques. This work enhances HAR by reconciling self-training’s potential with challenges and
offering practical insights for reliable activity recognition within wireless sensor networks. The results of experiments show that the self-training method, which uses channel state information based features to train the model with unlabeled data, is up to 97.5% accurate. Additionally, experiments using HAR datasets validate the proposed method and displays performance improvements over baselines
Autism spectrum disorder screening using DSM-5 fulfillment and machine learning adaptation
Get PDFAutism Spectrum Disorder (ASD) is a complex neurodevelopmental condition characterized by persistent challenges in social communication, restricted interests, and repetitive behaviours. The prevalence of ASD has increased globally, prompting the need for more reliable, objective, and scalable screening and diagnostic methods. Traditional diagnostic tools, such as the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), remain widely used in clinical settings. However, these tools are inherently dependent on subjective human judgment and clinician expertise, which can lead to inconsistencies in diagnosis and delayed interventions, particularly in early developmental stages. To address these limitations, this study explores a data-driven approach by integrating DSM-5 diagnostic criteria with advanced machine learning (ML) and deep learning (DL) models to enhance ASD detection and severity classification. Two datasets were employed in this research: the Autism Screening dataset, consisting of 1054 toddler data, 104 adolescence data, and 704 adult data samples, used for binary classification between ASD and non-ASD individuals; and the DSM-5 Diagnostic Dataset from Hospital Canselor Tuanku Muhriz UKM (HCTM), comprising 177 clinical samples after oversampling, used for multi-class classification of ASD severity (mild, moderate, and severe). Given the imbalance in class distribution, particularly in the severity-level dataset, oversampling techniques were implemented to improve model fairness and performance across all severity categories. The machine learning models evaluated in this study include Support Vector Machine (SVM), Decision Tree, and k-Nearest Neighbour (kNN). A Deep Neural Network (DNN) architecture was also designed and trained for comparative analysis. Model performance was assessed using standard classification metrics such as accuracy, precision, recall, and F1-score. The results demonstrate that the DNN model outperformed traditional ML models in both binary and severity-level classification tasks. Notably, the DNN achieved 100% accuracy in detecting ASD among younger children, reinforcing its potential as a tool for early screening. Furthermore, the severity classification results showed improved granularity and consistency compared to outcomes generated by manual assessments alone. This research highlights the value of integrating clinical standards with artificial intelligence to improve the speed, accuracy, and objectivity of ASD screening processes. The findings suggest that such hybrid approaches could support clinicians in making more informed decisions, reduce diagnostic delays, and enable timely interventions. Future research should explore larger and more diverse populations, refine model generalizability, address ethical considerations such as data privacy and bias, and assess real-world clinical deployment feasibility
Latent heat validation of phase change material using t-history method
Get PDFPhase change material (PCM) is a material that will absorb and release heat over a specified timeframe and it functions as a cooling technique to reduce temperature of the photovoltaic panels. The T-history method is a technique used to measure the thermal diffusivity of a material by subjecting a sample to a sudden temperature change. This project proposed PCM36 as a cooling method for PV temperature reduction and the expected result for this study is to meet the data from manufacturer with a small different percentage. T-history is introduced to validate the PCM36 latent heat capacity and melting point and thus compare it with the manufacturer data. The manufacturer data's latent heat capacity and melting point are 220 J/g and 36 ˚C, respectively. Based on the result obtained, the latent heat capacity from T-history is 217.891 J/g and is 0.9591% different compared with the manufacturer data. On the other hand, the melting point based on the T-history curve is in the range of 36 ˚C-38 ˚C, which is similar to the manufacturer data
Experimental analysis of color influence on optimized FDM parameters for PLA using the Taguchi method
Get PDFThis study examines the influence of filament color on optimizing FDM process parameters for PLA parts using the Taguchi method. Parameters such as layer thickness, print speed, and printing temperature were varied to identify optimal settings for white and black PLA filaments. The results demonstrate that the optimal parameters vary based on color: for white PLA, the best configuration involves a layer thickness of 0.35 mm, print speed of 50 mm/s, and a printing temperature of 210°C. For black PLA, the same layer thickness and print speed are optimal, but the printing temperature is lower at 200°C. Layer thickness was identified as the most significant factor affecting tensile strength across both filament types. However, the ideal printing temperature depended on the color of the filament. Notably, white PLA exhibited higher tensile strength than black PLA, with an increase ranging from 1.33% to 15.54%, attributed to the thermal properties of color pigments. These findings highlight the critical role of filament color in determining mechanical performance during FDM printing. Incorporating filament color into the optimization of FDM parameters can enhance the quality, strength, and reliability of 3D-printed components. This research provides valuable insights for improving additive manufacturing outcomes across a range of applications
Tribological effect of thermal energy on TIG arc surfacing techniques for surface modification of stainless steel
Get PDFTIG (tungsten inert gas) torch surface modification is a
unique process that can produce the surface alloying on a
work surface and effectively improve surface hardness
while altering the tribological behaviors. The significance
of this work is to examine the influence of thermal energy
on the surface characteristic of 2205 duplex stainless steel
samples. The thermal energy varied from 0.48 to 1.440
KJ/mm. The surface hardness and microstructure features
of the tribological properties of the materials were examined. Results indicate that, as thermal energy increases, the hardness value increases, thereby resulting in an increase of tribological properties. However, at higher thermal energy of 1.440 KJ/mm, the modified surface exhibits cracking in the melt layer. The microstructure transformed into different populations of dendritic structures. The best thermal energy obtained was 0.768 KJ/mm that resulted in the lowest wear rate of 3.0 x 10-4 mm3/Nm and friction coefficient of 0.43. High hardness of the surface modification and increased
tribological behavior were linked to higher levels of arc
energy during TIG melting process
Probing defect formation in sulfur-annealed graphene for TMDC integration
Get PDFThe integration of graphene with other 2D materials has been extensively studied over the past decade to realize high-performance devices unattainable with single materials. Graphene-transition metal dichalcogenides (TMDCs) such as MoS2, WS2, MoSe2, and WSe2 vertical heterostructures have demonstrated promise in numerous electronic and optoelectronic applications due to the wide bandgap range and strong light–matter interaction in TMDCs, and the ability to form electrostatically tunable junctions with graphene. However, conventional methods for TMDCs growth, including chemical vapor deposition (CVD), electrodeposition, and atomic layer deposition (ALD), require high temperatures, which can degrade graphene's electrical and structural properties. Here, we investigate the impact of sulfur annealing on graphene, revealing significant etching and electrical degradation. Density functional theory (DFT) calculations identify the divacancy defect with two sulfur adatoms (DV-2S) and C–S–C bonds as the dominant defect, differing from the previously reported monovacancy with one sulfur adatom (MV-1S). This defect induces p-doping in graphene, consistent with experimental observations. To address these challenges, we introduce a protective strategy utilizing self-assembled monolayers (SAMs) during annealing, enabling the growth of high-quality WS2 on graphene via electrodeposition. Our findings provide a foundation for integrating TMDCs with graphene while preserving its properties, advancing high-performance electronic and optoelectronic applications
Multi-modal biometric authentication system using score fusion techniques
Get PDFBiometric as an advanced access control method, however, the security can be enhanced through combination of more than one biometric element into one system. This study investigates the enhancement of security in access control systems by implementing a multi-modal biometric authentication system. It explores three biometric combinations: face and fingerprint, face and iris, and fingerprint and iris by using datasets from the CASIA database. The methodology includes biometric image preprocessing, feature extraction using DeepFace (for face), minutiae points (for fingerprints), and Gabor filters (for iris), followed by score-level fusion using weighted average techniques. Experimental analysis reveals that the face-fingerprint combination achieves the highest accuracy of 90.8%, followed by face-iris at 88.8%, outperforming unimodal systems. These results demonstrate the advantage of combining biometric traits for a more reliable and secure authentication system, contributing to the advancement of biometric security technologies