Journal of Applied Science and Technology

STATISTICAL MODELING AND OPTIMIZATION OF SURFACE ROUGHNESS IN WIRE EDM OF NONCIRCULAR GEARS MADE OF SKD11 STEEL

2026-02-12T11:30:41+07:00

This study presents the statistical modeling and optimization of surface roughness in wire EDM of SKD11 noncircular gears. A Taguchi L9 design was employed to evaluate the effects of discharge current (I), feed rate (F), and wire winding speed (V). ANOVA results indicated that discharge current had the greatest influence (58.68%), followed by feed rate (31.95%) and wire winding speed (9.37%). A quadratic model developed using Response Surface Methodology (RSM) showed excellent accuracy (R² > 0.99). Surface roughness increased monotonically with higher parameter levels due to thermal effects. The optimized parameter combination significantly reduced surface roughness, confirming the effectiveness of the proposed statistical approach for precision gear manufacturing by WEDM.

VELOCITY-DEPENDENT DEFORMATION MECHANISM OF FeNiCrCoTi HIGH-ENTROPY ALLOY UNDER VIBRATION-ASSISTED MACHINING

2026-02-12T11:30:47+07:00

Vibration-assisted machining has shown great potential for improving the processing of hard and tough materials, but its mechanisms in high-entropy alloys (HEAs) remain unclear. In this study, molecular dynamics simulations are employed to investigate the velocity-dependent behavior of FeNiCrCoTi monocrystal during vibration-assisted nano-machining. The study examines machining forces, shear strain, von Mises stress, dislocation evolution, temperature distribution, and surface morphology. The results reveal that higher velocities increase normal and tangential forces, stress concentration, and local temperature, while reducing shear strain, dislocation accumulation, groove width, and worn atom number. These findings provide new atomistic insights into the deformation and material removal mechanisms of HEAs under vibration-assisted machining.

CUTTER CORRECTION METHOD FOR IMPROVING THE ACCURACY OF MANUFACTURED SCREW ROTOR BY END MILLING CUTTER

2026-02-12T11:30:49+07:00

The accuracy of the rotor profile is a significant factor in the vacuum pump performance. However, the manufactured rotor profile deviates from this theoretical form because of cumulative machine-tool errors—such as temperature changes, humidity, and nearby vibrations. Therefore, a cutter-correction method is proposed to improve the accuracy of the manufactured profile by compensating for profile errors during cutter design. A mathematical model of the design end-milling cutter profile, accounting for compensated cutter parameters, is first proposed. The influence of the compensated cutter parameters on the rotor profile error is analyzed and summarized in a sensitivity matrix. The correction cutter profile is established using the compensated cutter parameters, which are solved by applying the sensitivity matrix and Levenberg-Marquardt algorithm. The numerical example is used to verify the effectiveness of the proposed method for reducing rotor profile error.

ATOMISTIC INSIGHTS INTO Al/Al EXPLOSIVE WELDING: A MOLECULAR DYNAMICS STUDY OF INTERFACIAL BONDING AND DIFFUSION MECHANISMS

2026-02-12T11:30:51+07:00

Explosive welding (EXW) is widely used to join similar and dissimilar metals, yet its atomistic mechanisms remain unclear. In this study, molecular dynamics simulations were performed to investigate aluminum–aluminum EXW using an embedded atom method potential. Results show that flyer impact causes temperature to exceed 2500 K and pressure to peak at ~27 GPa, inducing severe plastic deformation without bulk melting. Mean square displacement and concentration analyses reveal diffusion during loading, while unloading enables interfacial mixing, resulting in a thin diffusion layer of ~12 Å. Radial distribution function results confirm solid-state bonding dominated by plastic deformation and localized atomic rearrangement. These findings provide atomistic insights into Al–Al explosive welding, supporting process optimization for reliable similar-metal joints.

INDENTATION SIZE EFFECT ON THE DEFORMATION BEHAVIOR OF Ta-Cu AMORPHOUS THIN FILMS

2026-02-12T11:30:53+07:00

Molecular dynamics simulations were conducted to investigate the deformation characteristics of Ta₄₀Cu₆₀ amorphous coatings under nanoindentation. This study provides insights into the influence of indenter size on the deformation mechanisms at the Ta₄₀Cu₆₀ amorphous coating/Cu substrate interface. The results reveal that both the maximum indentation force and the extent of the plastic deformation region increase with increasing indenter diameter.

SIMULATION STUDY OF A THREE-PHASE GRID-CONNECTED PHOTOVOLTAIC SYSTEM USING SVPWM TECHNIQUE FOR THE INVERTER

2026-02-12T11:30:54+07:00

This paper presents the design and simulation of a two-stage grid-connected three-phase photovoltaic (PV) system using the MATLAB/Simulink platform. The research focuses on evaluating the performance of the Perturb and Observe (P&O) MPPT algorithm integrated with Space Vector Pulse Width Modulation (SVPWM). Simulation results under varying irradiance conditions demonstrate that the P&O algorithm provides a fast dynamic response while maintaining a stable DC-link voltage at 800 V. Furthermore, the proposed control scheme optimizes the quality of the grid-injected current, achieving a Total Harmonic Distortion (THD) range of 1.89% to 4.17%, which strictly complies with the IEEE Std. 519-2022 requirements. The study confirms that the synergy between P&O and SVPWM effectively enhances the efficiency and reliability of high-power PV systems.

EVALUATION OF SOME MECHANICAL PROPERTIES OF WARP-KNITTED FABRICS USING RECYCLED POLYESTER YARN

2026-02-12T11:30:57+07:00

To meet the trend of sustainable development, the textile industry has promoted the use of recycled materials in the production of garments. Recycled Polyester fiber (r-PET) is a sustainable manufacturing material that benefits the environment by reducing plastic waste and carbon emissions in textile production. This paper presents the results of a study evaluating several indicators such as tensile strength, elongation, dimensional change after washing, and water capillary action of three warp-knitted fabric samples with different weave patterns, course/wale densities, weights g/m² and thicknesses, but using the same type of 100% recycled Polyester 40D combined with Spandex 30D yarn, for use in sportswear. The research results provide a scientific basis for designing warp-knitted fabrics using recycled r-PET yarn, aiming to create fabrics suitable for sportswear products according to specific usage requirements.

SMOTE‑ENSEMBLE: A REVIEW OF DATA‑BALANCING TECHNIQUES AND HYBRID MACHINE LEARNING MODELS FOR EARLY PREDICTION OF DIABETIC RETINOPATHY COMPLICATIONS

2026-02-12T11:31:00+07:00

Diabetic retinopathy (DR) is one of the most common microvascular complications of diabetes and the leading cause of vision loss worldwide, yet its early stages often produce no noticeable symptoms. Consequently, large‑scale fundus image screening programs face challenges due to the inherent class imbalance in DR datasets: severe cases requiring urgent intervention are scarce compared to mild or non‑DR images. To address this, we provide a comprehensive overview of SMOTE ensemble strategies to enhance detection sensitivity for underrepresented classes. First, we analyze DR’s pathophysiological progression and dataset characteristics, demonstrating how imbalance reduces model recall. We then detail SMOTE and its variants—including Borderline‑SMOTE, ADASYN, and Geometric SMOTE—highlighting controlled synthetic minority over‑sampling [1]. Next, we review ensemble learning frameworks (Bagging, Boosting, Voting, Stacking) and their integration with SMOTE, with emphasis on the SMOTEBoost algorithm and recent refinements. Synthesizing results from over thirty studies, we show that SMOTE ensemble methods yield 5–18 % improvements in AUC, F1 score, and recall for severe DR detection [2].

Finally, we discuss current limitations and propose future research directions—such as GAN‑based augmentation, multitask learning, and interpretable model design—to accelerate clinical deployment.

AN EFFECTIVE APPROACH FOR LOW-ALTITUDE AERIAL OBJECT DETECTION

2026-02-12T11:31:04+07:00

Over the past decade, advances in aerial surveillance systems have driven increasing interest in low-altitude flying object detection within the computer vision community. However, the effectiveness of detecting and classifying flying objects, such as unmanned aerial vehicles (UAVs) and kites, is strongly influenced by the availability of specialized datasets, which remain limited. To overcome this challenge, we introduce a new dataset named UKD (UAV and Kite Dataset). The UKD dataset includes 2,121 real-world images of UAVs and kites with bounding box annotations and is employed to train and assess four representative modern detection models, such as YOLOv7, YOLOv10, RT-DETRv3, and Salience-DERT. Experimental results demonstrate that leveraging the UKD dataset leads to notable performance improvements, with the detection results achieving up to 80.3%.

THE IMPACT OF CERTAIN WORKING PARAMETERS ON THE DRILLING PROCESS OF PERCUSSIVE-ROTARY DRILLING

2026-02-12T11:31:08+07:00

This study examines the impact of critical operational factors on the efficacy of rotary-percussive drilling through a multibody dynamic model developed using Lagrange's equations, whereby the rock is represented as a visco-elastic-plastic material. The characteristics examined encompass impact frequency, impact impulse, and rock rigidity. Simulation results demonstrate that with a rock stiffness of k = 5×10⁸ N/m, the ideal impact frequency is between 50 and 70 Hz, achieving maximum drilling efficiency at roughly 65 Hz. As the impact impulse escalates from 120 kN to 280 kN, the ideal frequency transitions from 55–65 Hz to 75–85 Hz. Furthermore, as rock rigidity escalates from 8.93×10⁷ to 8.4×10⁸ N/m, the ideal impact frequency accordingly climbs from 32 Hz to 69 Hz. The findings indicate that drilling efficiency is significantly influenced by the interplay of impact parameters and rock mechanical properties, establishing a solid foundation for the selection of optimal operating settings to improve drilling performance and minimize energy losses.

USING ROLE-PLAYING ACTIVITIES TO ENHANCE SPEAKING SKILLS OF SECOND-YEAR NON-ENGLISH MAJOR STUDENTS: AN ACTION RESEARCH AT A UNIVERSITY IN HUNG YEN

2026-02-12T11:31:11+07:00

Speaking English is widely regarded as one of the most challenging skills for non-English major students because it requires linguistic knowledge, confidence, and fluency. This study was conducted as a classroom-based action research project to investigate the effects of role-playing activities on students’ participation and speaking performance in English-speaking lessons. The research was carried out with 41 non-English major second-year students at a university in Hung Yen. A mixed-methods approach was employed, using observation checklists, pre-speaking tests, post-speaking tests, and student questionnaires, with data triangulated across two research cycles. The results indicated noticeable improvements in students’ speaking performance, particularly in fluency, confidence, and active participation. It is concluded that role-playing activities are an effective instructional technique for promoting spontaneous communication and interaction in real-life speaking contexts.

RESEARCH AND DEVELOPMENT OF AN AUTONOMOUS MOBILE ROBOT FOR EDUCATIONAL SUPPORT WITH INTELLIGENT RECOGNITION AND DIALOGUE FUNCTIONS

2026-02-12T14:50:22+07:00

In the context of educational digital transformation, the application of autonomous mobile robots for training support has been increasingly investigated. However, most existing studies have mainly focused on navigation, while user identification and dialogue interaction for educational purposes have not been sufficiently addressed. In this study, an autonomous mobile robot integrated with Light Detection and Ranging sensors and artificial intelligence was designed and implemented to support training activities in educational institutions.

The proposed system was developed by integrating a Delta-2A Light Detection and Ranging sensor with a simultaneous localization and mapping algorithm for autonomous navigation, an image processing module for face recognition, and a natural language model for user dialogue interaction. Experimental evaluations were conducted in a school corridor environment. The results showed that a 94% success rate in obstacle avoidance was achieved, face recognition accuracy reached 85% over 50 experimental trials, and correct dialogue response accuracy was 88%.

These results indicate that intelligent autonomous mobile robots can be effectively applied to training support and learner consultation, contributing to improved management efficiency and enhanced learning experiences in educational environments.