• Journal of Advanced Marine Engineering and Technology
• /
• v.33 no.5
• /
• pp.679-685
• /
• 2009

In order to simulate a smoke or poisonous gas emergency in a rectangular tunnel and to investigate a better way to exhaust the smoke, the characteristics of smoke flow have been analyzed using flow field data acquired by Particle Image Velocimetry(PIV). Olive oil has been used as tracer particles with the kinematic viscosity of air, $1.51{\times}10^{-5}\;m^2/s$. The investigation has done in the range of Reynolds number of 1600 to 5333 due to the inlet velocities of 0.3 m/s to 1 m/s respectively. The average velocity vector and instantaneous kinematic energy fields with respect to the three different Reynolds numbers are comparatively discussed by the Flow Manager. In general, the smoke flow becomes more disorderly and turbulent with the increase of Reynolds number. Kinematic energy in the measured region increases with the increase of Reynolds number while decreasing at the leeward direction about the outlet region.

• 유체기계공업학회:학술대회논문집
• /
• 2005.12a
• /
• pp.670-677
• /
• 2005

In this study, we analyzed the three dimensional unsteady flow field around the motor cooling fan using the unsteady lifting surface theory. We obtained the flow rate for various geometries of fan from the calculated results of velocity field. For the data of design parameter and rotating speed(rpm) of the fan, we can predict the flow rate of the motor cooling fan with thin thickness through numerical analysis without the experimental data of the free stream velocity which is a boundary condition of flow field. the numerical results showed the flow rate within 10% of error in comparison with experimental results. The radial fans, which are often used as internal motor fan were also investigated with the same procedure.

• The KSFM Journal of Fluid Machinery
• /
• v.13 no.1
• /
• pp.35-41
• /
• 2010

Experimental study on the flow field inside the nozzle for radial turbine was performed. At design point, the pressure is high and the Mach number is low at the pressure side of the nozzle inlet semi-vaneless space as the flow turns through the nozzle vanes. As the flow accelerates through the nozzle passage to the throat the pressure level at the pressure and suction sides becomes similar. The flow continued accelerating from the throat to the inlet of turbine wheel and the pressure field became uniform in the circumferential direction in the vaneless space. In high expansion ratio condition, strong favorable pressure gradient band region occurred just after the throat in the semi-vaneless space in the circumferential direction and the pressure became uniform in the circumferential direction after this band. In low expansion ratio condition, core flow acceleration is dominant after the throat and this non-uniform pressure field reached to the inlet of turbine wheel.

• Journal of computational fluids engineering
• /
• v.15 no.2
• /
• pp.21-27
• /
• 2010

Immersed boundary lattice Boltzmann method (IBLBM) has been applied to simulate a turbulent flow over circular cylinder in a flow field effectively. Although IBLBM is very effective method to simulate the flow over a complex shape of obstacle in the flow field regardless of the constructed grids in the calculation domain, the results, however, become numerically unstable in high reynolds number flow. The most effective suggestion to archive the numerical stability in high Reynolds number flow is applying the multiple relaxation time (MRT) model instead of single relaxation time(SRT) model in the collision term of lattice Boltzmann equation. In the research MRT model for IBLBM was introduced and comparing the numerical results obtained by applying SRT and MRT. The hydraulic characteristic of cylinder in a flow field between two parallel plate at the range of $Re{\leqq}2000$represented and it is also compared the drag and lifting coefficients of the cylinder calculated by IBLBM with SRT and MRT model.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.24 no.2
• /
• pp.251-259
• /
• 2000

The ionic wind formed in a nonuniform electric field has been recognized to have a significant effect on particle collection in an electrostatic precipitator(ESP). Under normal operating conditions the effect of ionic wind is not pronounced. However, as the flow velocity becomes smaller, the ionic wind becomes pronounced and induces secondary flow, which has a significant influence on the flow field and the particle collecting efficiency. In this paper, experiments for investigating the effect of secondary flow on collection efficiencies were carried out by changing the flow velocities in 0.2-0.7m/s and the applied voltages in 9-11kV/cm. The particle size distributions and concentrations are measured by DMA and CNC. To analyze the experimental results, numerical analysis of electric filed in ESP was carried out. It shows that particle collection is influenced by two independent dimensionless numbers, $Re_{ehd}\;and\;Re_{flow}$ not by $N_{ehd}$ alone. When $Re_{flow}$, decreases for constant $Re_{ehd}$, the secondary flow prohibits the particle collection. But when $Re_{ehd}$ increases for constant $Re_{flow}$, it enhances the particle collection by driving the particles into the collection region.

• Proceedings of the KSME Conference
• /
• 2001.11b
• /
• pp.211-216
• /
• 2001

Flow motion and variation of thermal field around a bubble which attached at the upper cooled solid wall in a $B\dot{e}nard$ convection flow is studied experimentally using thermo-sensitive liquid-crystal tracers and image processing for flow visualization and analysis. The air is injected gradually by $0.1m\ell$ to make the bubble. As the growing of the bubble in a $B\dot{e}nard$ convection flow, the variation of temperature field and surface tension along the bubble, which in turn cause to change the thermal field patterns and the flow direction and patterns. 6 cells flow pattern is transformed into diverse flow pattern. At the large size of a bubble, it's only conduction mechanism under the region of the bubble because of low Ra number 1137, but the convection flow both sides of the bubble leads to another convection flow in the bubble influence area which has been remained stable stagnation.

• Proceedings of the Korean Society of Propulsion Engineers Conference
• /
• 2004.03a
• /
• pp.85-92
• /
• 2004

As a research to develop a SCRAM jet engine is actively conducted, a necessity to produce a high-enthalpy flow in a laboratory is increasing. In order to develop the SCRAM-jet engine, stabilized combustion in a supersonic flow-field should be attained, in which a duration time of flow is extremely short. Therefore, a mixing process of breathed air and fuel, which is injected into supersonic flow-fields is one of the most important problem. Since, the flow inside SCRAM jet engine has high-enthalpy, an experimental facility is required to produce such high-enthalpy flow-field. In this study, a detonation-driven shock tunnel was built and was used to produce high-enthalpy flow. Further-more, SCRAM jet engine model equipped backward-facing step was installed at test section and flow-fields were visualized using color-schlieren technique and high speed video camera. The fuel was injected perpendicular to the flow of Mach number three behind backward-facing step. The height of the step, distance of injection and injection pressure were changed to investigate the effects of step on a mixing characteristic between air and fuel. The schlieren photograph and pressure histories show that the fuel was ignited behind the step.

• Nuclear Engineering and Technology
• /
• v.54 no.1
• /
• pp.36-48
• /
• 2022

As a valid numerical method to obtain a high-resolution result of a flow field, computational fluid dynamics (CFD) have been widely used to study coolant flow and heat transfer characteristics in fuel rod bundles. However, the time-consuming, iterative calculation of Navier-Stokes equations makes CFD unsuitable for the scenarios that require efficient simulation such as sensitivity analysis and uncertainty quantification. To solve this problem, a reduced-order model (ROM) based on proper orthogonal decomposition (POD) and machine learning (ML) is proposed to simulate the flow field efficiently. Firstly, a validated CFD model to output the flow field data set of the rod bundle is established. Secondly, based on the POD method, the modes and corresponding coefficients of the flow field were extracted. Then, an deep feed-forward neural network, due to its efficiency in approximating arbitrary functions and its ability to handle high-dimensional and strong nonlinear problems, is selected to build a model that maps the non-linear relationship between the mode coefficients and the boundary conditions. A trained surrogate model for modes coefficients prediction is obtained after a certain number of training iterations. Finally, the flow field is reconstructed by combining the product of the POD basis and coefficients. Based on the test dataset, an evaluation of the ROM is carried out. The evaluation results show that the proposed POD-ROM accurately describe the flow status of the fluid field in rod bundles with high resolution in only a few milliseconds.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2002.11a
• /
• pp.365.2-365
• /
• 2002

The flow field around a symmetrical airfoil in a uniform flow under the generation of noise was studied by experiments and numerical simulation. The experiments are conducted by visualizing the surface flow over the airfoil with a shear-sensitive liquid-crystal coating and by measuring the instantaneous velocity field around the trailing edge of the airfoil. The numerical simulation was carried out by LES. (omitted)

• Journal of Broadcast Engineering
• /
• v.24 no.5
• /
• pp.755-764
• /
• 2019

Currently commercially available light field cameras are difficult to acquire 5D light field video since it can only acquire the still images or high price of the device. In order to solve these problems, we propose a deep learning based method for synthesizing the light field video from monocular video. To solve the problem of obtaining the light field video training data, we use UnrealCV to acquire synthetic light field data by realistic rendering of 3D graphic scene and use it for training. The proposed deep running framework synthesizes the light field video with each sub-aperture image (SAI) of $9{\times}9$ from the input monocular video. The proposed network consists of a network for predicting the appearance flow from the input image converted to the luminance image, and a network for predicting the optical flow between the adjacent light field video frames obtained from the appearance flow.