To reduce greenhouse gas emissions in maritime transport, the demand for LNG-fueled ships has increased. This study introduces a newly designed combined heat exchanger capable of simultaneously vaporizing LNG and heating BOG, reducing weight, installation area, and volume compared to conventional systems using separate devices. Performance evaluation was conducted through both experiments and CFD simulations using the standard k-ε turbulence model. The results showed good agreement in pressure drop but some deviation in temperature due to complex internal geometry. Additional simulations under operational conditions demonstrated the exchanger's effectiveness in achieving the required temperature difference for LNG vaporization. The flow field analysis confirmed uniform heat transfer around both LNG and BOG tubes, indicating minimal flow stagnation. Future work will explore improved turbulence models and alternative tube arrangements to further enhance thermal performance.

Droplet formation via microfluidics is widely used in bioanalysis, drug delivery, and microparticle synthesis, but stable generation under high-viscosity conditions (μd ≳100mPa·s) remains challenging due to delayed neck thinning and unstable breakup. To address this, a hybrid microchannel combining a T-junction with step expansion and horizontal shear enhancement was proposed. CFD simulations using the volume-of-fluid (VOF) method were conducted for viscosity ratios (λ) from 5 to 100. The study analyzed pressure, strain rate, and droplet size to characterize breakup behavior. In the dripping regime, concentrated strain rate at the neck enabled fast breakup and smaller droplets, while the jetting regime showed higher pressure and delayed breakup, producing larger droplets. Despite the challenges at high viscosity, the hybrid design maintained stable droplet generation up to λ = 100, where conventional T-junctions fail. These results demonstrate the hybrid geometry's effectiveness for robust droplet formation in high-viscosity regimes.

In this study, an extended uncertainty evaluation methodology was developed by introducing explicit correction terms for these loss mechanisms and by incorporating a ΔT-dependent sensitivity model to address nonlinear error propagation under small temperature differences. The proposed procedure was validated using experimental data obtained from a counter-flow tubular heat exchanger operating with a liquid-metal and air pair. The expanded uncertainty of the net heat transfer rate was -5.69 % to +3.90 % on the hot side and ±10.58 % on the cold side, demonstrating good agreement with physically observed heat-loss behavior. The results confirm that the present methodology provides a physically consistent framework for quantifying experimental uncertainties in low-Prandtl-number thermal systems, and can serve as a standardized approach for future high-temperature liquid-metal heat transfer experiments.

This study evaluates the performance of three AI instance segmentation models YOLOv8l-seg, Detectron2, and StarDist for microbubble diameter analysis. A ground truth dataset of 100 test images was prepared using Roboflow, and model performance was assessed using eight metrics: FPS, AP, Precision, Recall, PQ, RQ, SQ, and Boundary F1 score. YOLOv8l-seg achieved the highest FPS and Recall but tended to overestimate bubble diameters. Detectron2 showed the most balanced and accurate performance, with the smallest deviation from ground truth, while StarDist exhibited good segmentation accuracy but frequent false detections of small particles, leading to underestimation. Diameter analysis was conducted using an equivalent circle algorithm, with results compared via probability and cumulative density functions, scatter plots, and Bland-Altman analysis. Overall, Detectron2 demonstrated the highest reliability and accuracy, suggesting its strong potential for automated and high-volume micro-bubble diameter measurements.

Flexible fibers in a fluid experience hydrodynamic forces, and some spiders utilize this mechanism through a fiber bundle composed of multiple fibers. In this study, we introduce a model to simulate the dynamics of a flexible fiber bundle in a fluid. The model is based on the bead-rod model with a correction coefficient that represents the collective effects of multiple fibers. The modeling results are presented for two cases: a fiber bundle with one end translating at a constant velocity and a bundle freely falling with one end fixed. The predicted deformations are consistent with experimental observations, indicating that the model successfully captures the dynamics of the fiber bundle in a fluid.

Power-law fluids exhibiting shear-thinning behavior frequently appear in food colloidal and emulsion systems. Because such fluids undergo both shear and extensional deformation, quantifying their extensional rheology is essential for understanding food flowability and sensory properties. In this study, the extensional rheological properties of tomato ketchup, a representative power-law fluid, were characterized using a Capillary Breakup Extensional Rheometer (CaBER). Experiments were performed under an acceleration a = 3 m/s² and a final aspect ratio Λ_f = 1.83, varying the initial stretching velocity. A stable passive stretching regime, allowing rheological analysis, appeared at speeds above 60 mm/s. The dimensionless filament diameter evolution of ketchup followed a power-law model whose index matched that obtained from steady-shear viscosity. The extensional viscosity decreased with increasing extensional rate following a power-law trend, consistent with the shear viscosity-shear rate relationship.

In the present study, leading-edge tubercles were applied to a paraglider canopy to experimentally investigate changes in aerodynamic performance and flow characteristics at a Reynolds number of 3.0×10⁵. The tubercle model exhibited, on average, a 15% improvement in the lift-to-drag ratio (L/D) in the post-stall region. In the trough regions of the tubercles, a bistable flow behavior was observed, characterized by alternating flow separation and reattachment over time. As a result, the tubercle model showed a smaller wake size in the trough regions compared to the base model. These results contrast with the trends typically reported for conventional leading-edge tubercle geometries. The observed bistable flow appears to contribute to the recovery of the lift coefficient of the tubercle model at high angles of attack.

Efficient intracellular delivery of biomolecules is vital for cell therapy and genetic engineering. Microfluidic mechanoporation enables physical membrane permeabilization but often causes cell damage under excessive shear stress. Here, we investigated the relationship between viscoelastic flow and intracellular delivery using hyaluronic acid (HA)-based solutions in a constricted microchannel. Jurkat E6.1 cells were processed under varying flow rates (0.5-4 mL/min) and HA concentrations (0-2.0 mg/mL). Increasing flow rate and HA concentration enhanced cell deformation and molecular uptake, while excessive stress reduced recovery. To quantify these effects, elastic and drag stresses acting on cells were estimated using viscoelastic approximations. Delivery efficiency exhibited a strong correlation with elastic stress (R² = 0.925, Boltzmann model), whereas recovery correlated inversely with drag stress (R² = 0.910, logistic model). These results indicate that membrane permeabilization is governed by the balance of elastic and viscous forces, providing a quantitative framework for optimizing viscoelastic mechanoporation systems.

This study presents an eccentric-motor-driven microdroplet dispensing system for localized wet cleaning of semiconductor wafers. The device generates single microdroplets via negative-pressure suction and periodic vibration in a micro-nozzle, allowing independent control of pressure, vibration, and nozzle temperature. High-speed imaging is used to characterize droplet formation and derive a scaling relation between normalized droplet size and key dimensionless groups, enabling prediction and control of operating conditions. Cleaning performance is evaluated on flat and line-patterned wafers contaminated with micron-scale particles. Compared with millimetric droplets under similar impact conditions, microdroplet impacts produce thinner liquid films and stronger near-wall shear, leading to more efficient particle removal, particularly inside line pattern structures. The proposed system offers a compact, mechanically simple alternative to piezoelectric and conventional spray cleaning methods, and highlights the potential of microdroplet-based impact cleaning for advanced semiconductor and MEMS applications.

Previous Tesla turbine studies, limited to 2D analyses, have overlooked 3D flow effects, particularly how disc thickness can cause flow separation and reduce efficiency. This study investigates how 3D disc hole geometries influence turbine performance. Four disc prototypes were fabricated and tested under controlled hydraulic conditions. Performance was evaluated via efficiency measurements, and Particle Image Velocimetry (PIV) was used to analyze flow characteristics. The optimized D₃₀ geometry achieved the highest efficiency Best Efficiency Point (BEP) of 19.68%, a significant improvement over the standard disc (15.19%). The D_20+d model also showed high efficiency at 18.52%. PIV analysis confirmed that these 3D geometries enhanced fluid-disc energy transfer, indicated by a greater rotational speed difference between the inlet and outlet. These results demonstrate that 3D hole geometry is a critical factor for enhancing bladeless turbine performance.