This study evaluates the accuracy and reliability of brain hemorrhage prediction using the EfficientNet B7 model. The model achieved an accuracy of 94.2% and a recall of 94.0%, demonstrating high sensitivity that enhances its clinical applicability. The model achieved a loss of 0.40 during training and validation, showing stable convergence. These results expand the potential for AI in medical image analysis, ultimately contributing to improved diagnostic accuracy for healthcare professionals. Future research will verify the model's versatility using diverse datasets and increase interpretability for better clinical integration.
The seriousness of environmental pollution and Particulate matter is becoming a hot topic around the world, and interest in air pollution problems caused by exhaust gases generated from industrial sites, automobiles, and ships is also increasing. Korea's air environment has a significant external impact due to its regional characteristics, so there is a limit to establishing an air environment management plan according to regional emission characteristics. In order to reduce Particulate matter emissions, various industrial fields use technology to remove air pollutants by using an electric precipitator to reduce fine pollutants. In this study, we intend to optimally design the dust collecting plate and electrode plate of the low-pressure electric precipitator to confirm the changes in physical properties and properties of SUS316L materials before and after exposure to diesel engine exhaust gas.
Various vibrations occur while a ship is in operation. In the propulsion shaft systems of medium and small-sized vessels, it is common to install elastic couplings to prevent fluctuating torque from being directly transmitted to the reduction gear system and to avoid damage caused by torsional vibrations. However, the majority of elastic couplings are currently imported. The engine speed was gradually increased during the coupling test, while the vibrations were measured. At the point of maximum output, the engine speed was maintained at a constant level to observe the structural vibrations of the engine. This study aims to localize the production of elastic couplings by addressing the issue from the perspectives of torsional and structural vibrations to verify their operational performance.
In this study, high-efficiency InZnP/ZnSe/ZnS quantum dots (QDs) were synthesized and applied to the development of self-luminous traffic signs. The synthesized QDs exhibited a photoluminescence (PL) peak wavelength of 613 nm and a uniform nanoparticle size of 6 nm in diameter. Quantum dot films were coated using a roll-to-roll process with varying thicknesses (110 μm, 85 μm, and 75 μm), and color coordinate analysis confirmed that all films fell within the red region. A 4CL resin and PCB with high thermal dissipation properties were employed to fabricate the self-luminous traffic signs. Performance evaluations showed a white luminance of 625.4 cd/㎡, a red luminance of 220.2 cd/㎡, an insulation resistance of 10,000 MΩ, and an insulation breakdown voltage of 500V, indicating excellent performance. This study confirms the potential of InZnP-based QDs as an environmentally friendly alternative to conventional Cd-based QDs. Furthermore, the integration of high-efficiency QD films with advanced heat-dissipating PCB technology is expected to improve traffic safety, particularly in low-visibility and nighttime environments
Conventional bipolar electrodes (typically with round or flat tips) deliver radiofrequency energy in a broad, continuous manner. Their larger tip size and simple shape cause the applied energy to disperse over a wide area, making precise lesion control difficult and often leading to collateral tissue damage. As a result of these design limitations, traditional electrodes exhibit lower energy efficiency and tend to create lesions that unintentionally extend beyond the target area, with excessive thermal spread to surrounding tissues. In contrast, the five newly developed bipolar electrode designs concentrate energy delivery more effectively and provide improved control over lesion size and shape. These novel electrodes demonstrated higher energy efficiency, produced well-confined lesions, and minimized thermal injury to adjacent tissues, thereby overcoming the major drawbacks of conventional designs.
This experiment was conducted in a 1-pyeong cold room. Looking at the experimental results, it can be seen that the power consumption is twice as high when the thermoelectric module cooler is used compared to when the steam compression type cooler is operated. As is commonly known, steam compression coolers have a COP of about 3.0 in commercial coolers such as air conditioning systems and refrigerators, and the COP of coolers using thermoelectric modules is about 1.0, showing a performance difference of about 3 times. The reason for this difference is that the heat conduction fin material and shape were optimized, and the defrosting power consumption was relatively small in the thermoelectric method. The performance of thermoelectric devices shows that although there are still many improvements to be made over existing methods, they can exhibit sufficiently different advantages depending on the system.
The purpose of this study is to examine the application and effectiveness of tuned mass dampers for reducing cabinet vibration in plants. Cabinet with lower structural rigidity than plant subject to seismic design standards is susceptible to resonance. SolidWorks was used for 3D modeling of the cabinet, and ANSYS Workbench was used to create a mesh. The vibration characteristics of the cabinet were investigated through modal analysis, and the possibility of resonance and vibration reduction performance of the cabinet were evaluated. The number of modes in the cabinet was set to 100, and the frequency and modal participation mass ratio of each mode were calculated. In order to examine the possibility of vibration reduction by tuned mass dampers, the vibration response characteristics of cabinets with and without tuned mass dampers were compared. The analysis results showed that the third mode had a significant effect on the dynamic behavior of the cabinet and that the modal participation effective mass ratio was larger than that of other vibration modes. And as the mass of the tuned mass damper increased, the vibration response of the cabinet decreased significantly, and the peak value of the cabinet decreased by up to 52%.
This study aims to enhance the efficiency of the after-sales service (A/S) process for commercial trucks by implementing a data-driven approach. Traditional A/S methods result in long repair wait times, especially for intermittent faults requiring symptom reproduction. To address this, a system that records Diagnostic Trouble Code (DTC) and Vehicle Running Mode (VRM) data at failure moments is proposed. By storing data from 10 seconds before and after an event, fault diagnosis can be performed without symptom reproduction. Additionally, for exported vehicles, stored data enables remote analysis, overcoming real-time data limitations due to varying environmental factors. This approach improves maintenance reliability, optimizes repair accuracy, and supports proactive quality improvements for newly developed vehicles.
This paper presents an 2 kW commercial Li-ion(lithium-ion) battery model using PSIM simulation. An in-house designed transformer was employed to measure the leakage and magnetic inductance and determine the resonant frequency. The LLC resonant converter’s transfer function was derived using the FHA method, and voltage gain characteristics were analyzed concerning frequency, Q factor, and inductance variations. The CC-CV charge mode was evaluated at different C-rates, targeting the charging voltage for e-bikes. The results demonstrated that battery voltage increased with SOC until reaching the final charging voltage.
In this study, a compact 20X magnification optical system was designed for application not only in hospitals and research institutions but also in hair salons and personal hair diagnostic devices. The system was developed with a measurement field diameter of 0.35 mm and a total optical system length of 125 mm to ensure practical usability. Through optical system analysis, it was confirmed that the designed system achieved an MTF of over 0.4 at 20 lp/mm, and the CCD optical axis tolerance exceeded 2 mm, providing sufficient alignment stability. A review of the microscope illumination structure confirmed that coaxial illumination alone was sufficient to observe the cuticle at the central region of the hair. Furthermore, experiments using the fabricated prototype clearly demonstrated structural changes in the cuticle depending on the degree of hair damage. In future research, by acquiring a larger dataset of hair samples with varying degrees of damage and thickness, AI-based analytical techniques could be utilized to quantitatively evaluate hair damage. This system holds significant potential for practical applications in precise, high-resolution hair diagnostics across various fields.
Ulsan Airport cannot operate precision instrument approach procedures from the south direction (Runway 18) due to obstacles. Even non-precision instrument approach procedures have higher approach angles and minimum descent altitudes (MDA) compared to other airports, which can pose safety risks for pilots following the flight procedures. Recently, since the introduction of SBAS-based satellite navigation flight procedures in Korea, Ulsan Airport is expected to experience improvements, including reduced offsets and lower minimum descent altitudes in its existing flight procedures. During the design process of new flight procedure routes, a comprehensive analysis of noise differences from existing routes and the noise impact on new areas is necessary. Therefore, this study aims to present the changes in aircraft noise resulting from the implementation of new flight procedures using the Aviation Environmental Design Tool (AEDT)
This study examines the innovative applications and future prospects of Convolutional Neural Networks (CNN) in the field of medical image analysis. CNNs significantly enhance the accuracy and efficiency of medical image diagnostics through their powerful data processing and feature extraction capabilities. This review analyzes various CNN architectures and recent technological advancements, highlighting the importance of transfer learning and data augmentation techniques. It also discusses the potential for integrated multi-modality data analysis and real-time clinical applications, while emphasizing the need for ethical considerations and data security. This research underscores the potential of CNN technology to improve healthcare quality and contribute to patient health management.
This study explores structural dynamics using experimental modal analysis with tri-axial accelerometers and advanced signal processing. By improving the accuracy of modal parameters such as natural frequencies and damping ratios, the research enhances vibration analysis techniques. The findings have applications in structural health monitoring, predictive maintenance, and mechanical system optimization.
When manufacturing a square plate with a pinhole, the following conclusions were obtained as a result of analyzing the effect of changes in the size of the grid filling the inside of the plate and the presence or absence of a pinhole on the stress concentration factors. 1. It can be seen that in the case where there is a pinhole, the overall stress concentration factors is twice as high as in the case where there is no pinhole. In addition, in the case where there is no pinhole, it can be seen that the stress concentration factors fluctuates at a certain standard depending on the position of the pinhole and the internal grid, and in the case where there is a pinhole, it decreases at a certain point and then increases again. 2. In the case where there is no pinhole, when the L/H ratio increases from 0.1 to 0.3, the Kt value increases from 1.8 to 2.7, an increase rate of approximately 50%. Similarly, in the case of a pinhole, when the L/H ratio increases from 0.1 to 0.4, the Kt value increases from 2.0 to 3.0, an increase rate of about 50%.
In this study, the effect of underwater hydrofoils was analyzed through model tests on a motor boat with a total length of approximately 8.5 m and a maximum speed of 25 knots(approximately 12.9 m/s). The degree of hull sinkage and resistance performance according to the size and angle of attack of the foils with the same cross sectional shape were analyzed and the required horsepower was estimated. Through this study, it was confirmed that the area and angle of attack of the foils had a great effect on hull sinkage and that a rapid decrease in resistance and change in required horsepower occurred depending on the degree of hull sinkage.
In this study, the molten steel flow inside the mold according to the shape change of the submerged entry nozzle (SEN) was simulated in the continuous casting process, and the effects of the immersion depth and long side length of the mold on the molten steel flow were compared and analyzed. The heat flow analysis was performed using the ANSYS Fluent software, and the turbulence model used the standard k-ε model. Based on the analysis results, the meniscus velocity decreased with deeper immersion depth, but it was slightly faster with lower immersion depth. In addition, the flow velocity of the molten steel of the outlet was increased because the molten steel exiting the upper outlet spread directly without colliding with the short side of the mold when the long side of the mold length was increased.
In this study, a response model of a beam structure was established through finite element analysis by analyzing the vibration response to external excitation. The vibration control performance of the beam was then evaluated by applying the narrow-band Fx-LMS algorithm for active structural control. The transfer function was obtained at the error sensor location when the structure was excited and the three-axis actuator was operated. The performance of the active control was investigated with 18 channels for error input and actuator output. When the equipment is exciting, the response of the error sensor is the primary path, and when the inertial 3-axis actuator operates, the response of the error sensor position is the secondary path, and the Fx-LMS algorithm is applied. The simulation was performed by changing the control parameters so that the response of the error sensor can satisfy the target performance. From the results of this study, the acceleration results over time showed about 70% vibration reduction after active control, and the average error value of the error sensor also decreased by about 68%. In addition, it was confirmed that real-time control of a system with 18 sensors and 18 actuators is possible even if the secondary path is configured in two orders.
This study analyzed the methods and characteristics of hydrogen production, storage, transportation, charging, and use of hydrogen presented as an energy supply and demand system for hydrogen. Hydrogen produced by reforming hydrogen, which exists in the form of compounds, is essential to use metal materials exposed to the hydrogen atmosphere in storage and transportation. The mechanism of hydrogen embrittlement and damage cases, which are phenomena in which hydrogen atoms penetrate into the crystal lattice of metal and cause crack failure, were investigated. In addition, it is intended to present a research direction related to the evaluation of physical properties such as thermal conductivity, thermal expansion coefficient, and heat capacity, which are the criteria for selecting materials for hydrogen in a cryogenic environment.
In four-wheel-drive vehicle, improving traction with the road surface enhances the vehicle's ability to respond to various driving conditions, increasing its overall versatility. Consequently, various studies have been conducted on four-wheel-drive vehicles that support torque distribution through electronic control. The driving unit that operates the transfer case assists in smooth torque distribution by providing high torque. Therefore, this study developed a reduction mechanism by vertically arranging a planetary gear set in the driving unit to increase the reduction ratio. To achieve this, a common ring gear with 52 teeth was used, and the design included a first-stage planetary gear with a sun gear having 18 teeth, a planet gear with 17 teeth, a second-stage sun gear with 12 teeth, and a planet gear with 20 teeth. The corresponding tooth profiles and structures were also designed. Based on this, a transfer case drive reduction module was developed, which improved torque performance: the first-stage planetary gear system provides 4.23 kgf·m of torque, and the second-stage planetary gear system achieves a final torque of 5.98 kgf·m
This research explores the mechanical behavior of whole blood as a shear-thinning fluid and examines yield stress as a crucial parameter for evaluating flow properties in the microvasculature. While previous studies have primarily focused on whole blood viscosity (WBV), this study presents an engineering-based approach to quantitatively evaluate the yield stress of non-Newtonian whole blood using a U-shaped capillary tube. A custom-designed scanning capillary tube, incorporating a biocompatible U-shaped capillary channel, was employed to determine whole blood viscosity (WBV) across shear rates from 1 to 1,000 s⁻1. Yield stress was estimated by extrapolating the shear stress-shear rate relationship using a non-Newtonian shear-thinning flow model. The analysis revealed a strong cubic correlation between hematocrit and yield stress, suggesting an increased risk of flow impairment in the microvasculature with elevated hematocrit levels. This work emphasizes the critical influence of yield stress in regulating microvascular blood flow, particularly under low-shear conditions.