• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.32 no.6
• /
• pp.429-445
• /
• 2008

Two-phase compressible flow fields of air-water are investigated numerically in the fixed Eulerian grid framework. The phase interface is captured via volume fractions of each phase. A way to model two phase compressible flows as a single phase one is found based on an equivalent equation of states of Tait's type for a multiphase cell. The equivalent single phase field is discretized using the Roe‘s approximate Riemann solver. Two approaches are tried to suppress the pressure oscillation phenomena at the phase interface, a passive advection of volume fraction and a direct pressure relaxation with the compressible form of volume fraction equation. The direct pressure equalizing method suppresses pressure oscillation successfully and generates sharp discontinuities, transmitting and reflecting acoustic waves naturally at the phase interface. In discretizing the compressible form of volume fraction equation, phase interfaces are geometrically reconstructed to minimize the numerical diffusion of volume fraction and relevant variables. The motion of a projectile in a water-filled tube which is fired by the release of highly pressurized air is simulated presuming the flow field as a two dimensional one, and several design factors affecting the projectile movement are investigated.

• Nuclear Engineering and Technology
• /
• v.54 no.11
• /
• pp.4322-4328
• /
• 2022

The two-fluid model is widely used to describe two-phase flows in complex systems such as nuclear reactors. Although the two-phase flow was successfully simulated, the standard two-fluid model suffers from an ill-posed nature. There are several remedies for the ill-posedness of the one-dimensional (1D) two-fluid model; among those, artificial viscosity is the focus of this study. Some previous works added artificial diffusion terms to both mass and momentum equations to render the two-fluid model well-posed and demonstrated that this method provided a numerically converging model. However, they did not consider mass conservation, which is crucial for analyzing a closed reactor system. In fact, the total mass is not conserved in the previous models. This study improves the artificial viscosity model such that the 1D incompressible two-fluid model is well-posed, and the total mass is conserved. The water faucet and Kelvin-Helmholtz instability flows were simulated to test the effect of the proposed artificial viscosity model. The results indicate that the proposed artificial viscosity model effectively remedies the ill-posedness of the two-fluid model while maintaining a negligible total mass error.

• 한국전산유체공학회:학술대회논문집
• /
• 2007.10a
• /
• pp.214-219
• /
• 2007

Both the bubble rising in a fully filled container and the droplet splash are simulated by a solution code(PowerCFD). This code employs an unstructured cell-centered method based on a conservative pressure-based finite-volume method with interface capturing method (CICSAM) in a volume of fluid(VOF) scheme for phase interface capturing. The present results are compared with other numerical solutions found in the literature. It is found that the present code simulate complex free surface flows such as multi phase flows due to large density difference efficiently and accurately.

• 한국전산유체공학회:학술대회논문집
• /
• 2009.11a
• /
• pp.115-119
• /
• 2009

We numerically investigated propagation of various waves in the two-phase flows such as sound wave, shock wave, rarefaction wave, and contact discontinuity in terms of pressure, void fraction, velocity and density of the two phases. The waves have been generated by a hydrodynamic shock tube, a pair of symmetric impulsive expansion, impulsive pressure and impulsive void waves. The six compressible two-fluid two-phase conservation laws with interfacial friction terms have been solved in two fractional steps. The first PDE Operator is solved by the HLL scheme and the second Source Operator by the semi-implicit stiff ODE solver. In the HLL scheme, the fastest wave speeds were estimated by the analytic eigenvalues of an approximate Jacobian matrix. We have discussed how the interfacial friction terms affect the wave structures in the numerical solution.

• Journal of computational fluids engineering
• /
• v.15 no.3
• /
• pp.73-80
• /
• 2010

In this paper, we present the exact Riemann solver for the compressible liquid-gas two-phase shock tube problems. We hereby consider both isentropic and non-isentropic two-phase flows. The shock tube has a diaphragm in the mid-section which separates the liquid medium on the left and the gas medium on the right. By rupturing the diaphragm, various waves are observed on the phasic field variables such as pressure, density, temperature and void fraction in the form of rarefaction wave, shock wave and material interface (contact discontinuity). Both phases are treated as compressible fluids using the linearized equation of state or the stiffened-gas equation of state. We solve several shock tube problems made of a high/low pressure in the liquid and a low/high pressure in the gas. The wave propagations are well resolved by the exact Riemann solutions.

• Journal of the Korean Society of Visualization
• /
• v.15 no.3
• /
• pp.20-26
• /
• 2017

The two-fluid equations are widely used to simulate two-phase flows in a nuclear reactor. For the two-fluid momentum equation, the wall and interfacial drag terms play an important role in predicting a two-phase flow behavior. Since the bubble density is much smaller than the water density, the bubble accelerates faster than the liquid in a nozzle. As a result, the bubble phase becomes faster than the liquid phase in the nozzle. In contrast, the opposite phenomena occur in the diffuser. The purpose of our study is to experimentally show these behaviors in an area-varying channel such as nozzle and diffuser. Experiments were made of turbulent bubbly flows in an area-varying horizontal channel. The velocities of the bubble and liquid phases were measured by the PIV technique. It was shown that the two-phase velocities were no longer close to each other in the area-varying regions. The bubble was faster than the liquid in the nozzle; in contrast, the bubble was slower than the liquid in the diffuser. Code simulations were also performed using the MARS code. By replacing the original wall drag model in the MARS code with Kim (1)'s wall drag partition model, we obtained the simulation results being consistent with experimental observations.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.4 no.1
• /
• pp.187-197
• /
• 1996

Many investigations have been made to determine the pressure drop and heat transfer characteristics for single phase flow in tape generated swirl flow. But few studies have been carried out to investigate the heat transfer in two component, two phase swirl flow with non-boiling. An experimental study has been conducted to determine the effects of tape twist ratios on two phase convective heat transfer coefficients, pressure drop, and void fraction distribution in a non-boiling, air-water, two phase flow. The flow conditions were both swirl and non swirl flows. The internal diameter of the test section is 42.5mm. The tape twist ratios of pitch to diameter ratio varied from 4.0 to 10.6. The heating conditions were isothermal and nonisothermal. The flow patterns identified with experiments were bubbly, bubbly-slug, slug, and slug-annular flow in up-flow. This study has concluded that no significant difference in void fraction distribution were observed both isothermal and nonisothermal conditions, the pressure drop for two phase flow with twisted tape swirler increase as the tape twist ratio decrease, and that values of two phase heat transfer coefficient increase when the tape twist ratio decreases.

• Journal of the Korean Society of Safety
• /
• v.32 no.1
• /
• pp.75-81
• /
• 2017

A series of laboratory experiments has been performed in order to investigate the behavior of water-air two-phase flow in a horizontal pipe. A conductivity meter has been applied to detect the irregular alternation of air at the specific points in flows. The experimental condition has been established according to the water and air flowrates. Passing time, which is the time length for a measuring probe to pass through the entire length of a specific bubble, has been defined to evaluate the size of bubbles in the flow. Passing length, which can be considered as the equivalent value to bubble size and determined from the product of passing time and cross-sectional averaged velocity, and its corresponding occurrence frequency have been analyzed to classify the air flow patterns according to the condition of air and water fluxes. From the result, the dependancy of flow patterns on the variation of air-water flux ratio has been investigated and the existence of thresholds also checked for classifying the behavior of air in the flow.

• Journal of the Korean Society of Visualization
• /
• v.10 no.2
• /
• pp.14-19
• /
• 2012

An optical fiber probe system for measuring the local void fraction in the air-water two-phase flow was developed with a 1550 nm light source. Air was injected through a nozzle placed in the center of the bottom wall of a water-filled cylindrical tank. The optical fiber probe having a diameter of $125{\mu}m$ was sufficiently thin to resolve the air-water interface of the bubbly flows. To verify the performance of the optical fiber probe, the synchronized high speed visualization study using a high speed camera was carried out. Comparison between the optical signals and the instantaneous bubble diffraction images confirms that the optical fiber probe is very accurate to measure the void fraction in two-phase flows. The estimated bubble diameter and the rising velocity by the optical fiber probe have 1% and 5% of accuracy, respectively.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.31 no.11
• /
• pp.904-911
• /
• 2007

A measurement technique fur void fraction has been proposed using a time-resolved two-phase PIV system and the bubble dynamics has been investigated in gas-liquid two-phase flows. For the three-dimensional evaluation of the bubble information, both the images from the front and side views are simultaneously recorded into a high speed CCD camera by reflecting the side view image on a $45^{\circ}$ oriented mirror to be juxtaposed with the front view image. Then, a stereo-matching technique is applied to calculate the void fraction, bubble size and shape. To obtain the rising bubble velocities, the 2-frame PTV method was adopted. The present technique is applied to freely rising bubby flows in stagnant liquid. The results show that the increase of bubble flow rate gives rise to the increase of bubble size and rising velocity at first. If it goes over a certain level, the rising velocity becomes constant and the horizontal velocity grows bigger instead due to the obstruction of other bubbles.