In recent years, we can easily find the gas-liquid two-phase flow in narrow channel which has straight section and curved section in many industrial products. In order to improve the performance of these industrial products, it is important to clarify the effects of curved section on gas-liquid two-phase flow behavior in narrow channel. In this study, we have measured the pressure loss precisely on straight section and curved section in milli-channel respectively. From the measured pressure loss, we evaluated the mean pressure loss and its intensity. Flow visualization by using high-speed video camera was additionally performed to make clear the relation between modification of pressure loss and flow pattern in curved section. As a result, effects of curved section on gas-liquid two-phase flow in narrow channel were evaluated.

A new code coupled between a sub-channel analysis code and a computational multi-fluid dynamics (CMFD) code was applied to a PWR rod bundle with mixing vane grid. The code was developed to predict the departure from nucleate boiling (DNB). This is a new technology of CMFD based on abundant experience and models developed for sub-channel analysis: CMFD computed void distribution to fit the average value calculated by two-phase models in sub-channel analysis code. A new source term represented centripetal motion of small bubbles in the wake behind a rising vapor slug. In order to apply the code to a PWR rod bundle, effects of local mass flux, pressure and mixing vane grid was modeled. The present method was applied to the analysis for DNB tests of simulated PWR fuel assembly with mixing vane grids. The result showed that, using only a critical void fraction, reasonable prediction was achieved in a wide range of flow condition and a variety of flow regimes.

The phenomena of frosting and defrosting are quite important in evaluation of efficiency of various heat equipments such as heat pump etc. In this paper, hydrodynamic phenomena in frosting and defrosting were studied experimentally. Cooling surfaces were flat or artificially grooved with the size of several tens of microns in pitch and depth. In frosting process, super cooled small water droplets were formed on the cooling surface. In the flat surface, droplets are formed around the cavities which are small holes or scars on the surface. In the artificially grooved surface, droplets are formed in the groove or flat part between grooves. Maximum size of droplets on the artificially grooved surface was controlled by pitch and depth of the grooves. In the defrosting, larger water droplets or water films were formed on the cooling surface. On the flat surface, water droplets or films remained on the surface for long time. On the other hand, on the grooved surface, water droplets or films were drained rapidly through microgroove due to the capillary driven flow.

Accurate analyses of turbulence structure and void fraction distribution are quite important in designing and safety evaluation of various industrial equipments using gas-liquid two-phase flow such as nuclear reactor, etc. Using turbulence model of two-phase flow and models of bubble behaviors in bubble flow and slug flow, systematic analyses of distributions of void fraction, averaged velocity and turbulent velocity were carried out and compared with experimental data. In bubbly flow, diffusion of bubble and lift force are dominant in determining void fraction distribution. On the other hand, in slug flow, large scale turbulence eddies which convey bubbles into the center of flow passage are important in determining void fraction distribution. In turbulence model, one equation turbulence model is used with turbulence generation and turbulence dissipation due to bubbles. Mixing length due to bubble is also modeled. Using these bubble behavior models and turbulence models, systematic predictions were carried out for void distributions and turbulence distributions for wide range of flow conditions of two phase flow including bubbly and slug flow. The results of predictions were compared with experimental data in round straight tube with successful agreement. In particular, concave void distributions in bubbly flow and convex distribution in slug flow were well predicted based on the present model.

Unsteady numerical analyses were carried out for the PCCI combustion realization on the next generation diesel engine. The effects of intentional initial fuel concentration distribution on the ignition timing and the expansion of the operable range were focused. The numerical code was originally developed to analyze the unsteady combustion behavior of premixed gas with intentional initial fuel concentration distribution in the combustion chamber of PCCI engine. The fundamental equations for the numerical analysis are the Euler’s equations for compressible fluid, that consist of the conservation equation of mass, momentum, energy and chemical species. The equations are expressed in axisymmetric cylindrical (r-z) coordinate system. The time variation of the internal cylinder volume of the reciprocating crank-piston movement is expressed by the volume change rate, which is substituting the convection term in axial direction of the Euler’s equations. By using this technique, we can reduce the spatial dimension of the equations with keeping the consideration of volume change of the combustion chamber of reciprocating engine. The radial direction corresponds to the bore of the engine cylinder. By solving this equation system, we can consider the radial distribution in combustion chamber during the PCCI combustion such as temperature, mass fraction of chemical species, pressure, and so on. Detailed chemical kinetics with elementary reactions and multi-component diffusion for n-heptane system as fuel, ERC-mechanism, were considered. The NOx emission can be also considered by using the part of GRI-Mech3.0 for generation of thermal NOx. Totally, the 34 chemical species and 61 elementary reactions were considered. This code has a high resolution for time and space to capture the dynamic behavior in PCCI combustion such as a generation and propagation of shockwave causing detonation. A series of unsteady events on PCCI combustion can be simulated by considering the time variation of volume of combustion chamber varied with crank angle; such as the charged compression and auto-ignition of premixed gas, the flame propagation and the detonation with shockwave.

A computational multi-fluid dynamics (CMFD) code that predicted the void distribution in sub-channel was coupled with sub-channel analysis code to predict departure from nucleate boiling (DNB). The main assumption was that the void fraction near heated wall was the dominant parameter in DNB. A sub-channel analysis code was used to calculate three dimensional distribution of sub-channel averaged values of mass flux, void fraction, density and quality. These were used as a boundary condition in the CMFD code to predict local void fraction in a subchannel. A bubble diffusion equation was used assuming the wall peak void distribution caused by turbulence. The present method was applied to the analysis of DNB tests. The coupled codes showed a reasonable profile of void fraction in a rod bundle and reproduced DNB heat flux at low void fraction. To investigate this analysis result, the local condition were compared with a DNB flow regime map. This suggested an approach to improve the predictability: the critical void fraction should be modified at low void fraction condition; the bubble diffusion model should be modified to handle the flow regime transition from the isolated nucleation type to churn turbulent flow type.

In order to predict the critical power or void fraction in BWR fuel bundles and the DNB heat flux of PWR fuel assemblies, the boiling transition analysis code called “CAPE” with mechanistic models has been developed in the IMPACT project by NUPEC. The objective of the CAPE code development is to perform with good accuracy the safety evaluation for a new type or improved fuel bundle design of BWR and PWR without full-scale experiments or any tuning parameters in the analysis code. In the present study dryout heat fluxes of BWR fuel assembly were analyzed by the CAPE code and compared with experimental data of BFBT benchmark test carried out by NUPEC. In the CAPE code, mechanistic model of liquid film dryout in annular flow is used considering entrainment and deposition of droplet. In such mechanistic prediction of dryout, the correlations of entrainment and deposition rates play quite important roles and many correlations have been developed. In addition to the original correlations in the CAPE code, several typical correlations, which are widely used in the analysis of annular dispersed flow, were tested for the prediction of dryout. The results indicated that the CAPE code satisfactorily predicted dryout heat fluxes of fuel assembly for wide range of pressure, mass flux, subcooling and bundle geometries obtained in BFBT benchmark test. The accuracy of prediction depends upon the combination of correlations of entrainment and deposition rates. The evaluation of correlations of entrainment and deposition rates was carried out.

In order to predict the critical power or void fraction in BWR fuel bundles and the DNB heat flux of PWR fuel assemblies, the boiling transition analysis code called “CAPE” with mechanistic models has been developed in the IMPACT project by NUPEC. The objective of the CAPE code development is to perform with good accuracy the safety evaluation for a new type or improved fuel bundle design of BWR and PWR without full-scale experiments or any tuning parameters in the analysis code. In the present study void fraction distribution of BWR fuel assembly were analyzed by the CAPE code and compared with experimental data of BFBT benchmark test carried out by NUPEC. The analyses were carried out by changing the operational parameter such as the inlet subcooling, mass flow rate and the power output of the fuel bundles. Resultantly, the thermal equilibrium quality at the outlet ranges 2% ∼25%. Averaged and local void fractions were compared between experiment and analysis. The results indicated that the CAPE code satisfactorily predicted void fraction distribution of fuel assembly for wide range of pressure, mass flux, subcooling and bundle geometries obtained in BFBT benchmark test. However, the detailed analysis showed that in some subchannels, which was surrounded by the heated fuel rods and partially unheated wall such as an unheated rod, a water rod and a separation wall of the channel box, certain difference between experiment and prediction appeared. In the CAPE code, the drift flux model is used for predicting void behavior in fuel assembly. In drift flux model, correlations of drift velocity and distribution parameters are quite important in predictive accuracy of void fraction. In particular, it is considered that distribution parameter plays important role in void distributions in subchannel surrounded by unheated rods. In addition to the original correlations in the CAPE, some typical correlations of drift velocity and distribution parameter were tested for the prediction of void fraction distribution. The evaluations of correlations of drift velocity and distribution parameter were carried out, and some recommendations for improvement of predictive capability of the CAPE code were made.

Numerical simulation code for predicting void distribution in two-phase turbulent flow in a sub-channel was developed. The purpose is to obtain a profile of void distribution in the sub-channel. The result will be used for predicting a heat flux at departure from nucleate boiling (DNB) in a rod bundle for the pressurized water reactor (PWR). The fundamental equations were represented by a generalized transport equation, and the transport equation was transformed onto the generalized coordinate system fitted to the rod surface and the symmetric lines in the sub-channel. Using the finite-volume method the transport equation was discretized for the SIMPLE algorism. The flow field and void fraction at the steady state were calculated for different average void fractions. The computational result for atmospheric pressure condition was successfully compared with experimental data. Sensitivity analysis for the PWR condition was performed, and the result showed that the secondary flow slightly contributed to homogenizing the void distribution.

Freon thermal hydraulic test is expected to be one of the workable methods to develop high thermal hydraulic performance PWR fuel. That is, high pressure water and high heat flux condition in PWR core can be substituted with lower pressure Freon and lower heat flux by applying appropriate fluid-to-fluid similarity and modeling parameters. Freon DNB tests and mixing tests were carried out against a 4×4 rod bundle configuration where R-134A flowed vertically upwardly. The tests were carried out at Freon thermal hydraulic test loop in Korea Atomic Energy Research Institute (KAERI). The spacer grid used in these tests was modeled on that of conventional PWR fuel, that is, square lattice grid with split type mixing vanes. Diameter of heater rod simulating PWR fuel rod is about 10.7mm and heating length is about 2000 mm. Freon mixing tests were carried out to estimate Turbulence Diffusivity Coefficient (TDC), which was normally used in conventional thermal hydraulic design of nuclear reactor. Freon CHF test results showed that parametric trends agreed with those of existing CHF data. To predict CHF of 4×4 rod bundle, subchannel analysis code Modified COBRA-3C and NFI-1 DNB correlation were applied. TDC value used in subchannel analysis was determined by fitting Freon mixing test data. NFI-1 DNB correlation was developed for predicting DNB heat flux in rod bundle configuration by using water CHF test results at HTRF test loop at Columbia University. The design of spacer grids used in KAERI Freon DNB test was similar to that used in water CHF test at HTRF. Water equivalent flow condition of this R-134A test was estimated using fluid-to-fluid similarities. NFI-1 DNB correlation was applied to this water equivalent condition to estimate water equivalent DNB heat flux. Then R-134A equivalent DNB heat flux was estimated reversely, and compared to Freon DNB test result. The test results were predicted well and applicability of NFI-1 DNB correlation and fluid-to-fluid similarities in 4×4 rod bundle is discussed.

In the present investigation, detailed observations on the flow patterns and measurements for pressure drop characteristics affected by the curvature effect in a micro-channel were carried out. The test channel has a square cross section that has 2.0mm on a side, and the radii of the curvature were 1.0mm and 3.0mm. It was cleared that the pressure fluctuation intensity became larger in the case of the curvature 1.0mm compared with the case of the curvature 3.0mm. Furthermore, in order to investigate the influence of the wall wettabirity, we applied the water repellent channel. In the case of the water repellent micro-channel, the flow pattern tended to easily transit to Annular flow side compared with the normal micro-channel without the water repellent, and we can see a new flow regime like Rivulet flow due to applying water repellent.

The process of oblique impact of single water drops onto a water film falling down on an inclined wall was observed using a high-speed camera. The angle between the velocity vector of the impacting drop and the normal vector to the inclined wall was set to the values greater than 70°. In our previous experiments for the single water drop impact onto a quiescent water surface, no secondary drop was observed in this large impingement angle range. In the present experiments, however, secondary drops were produced occasionally during the collision process. The production of secondary drops was observed more frequently when the surface of falling liquid film was wavy. The process of the production of secondary drops from the falling water film was investigated using the video images to understand the difference between the drop impacts onto the quiescent film and onto the falling film. It was shown that the secondary drops could be produced when the primary drop collided against the bottom or the top of the wave. In the former case, it was considered that the impingement angle was overestimated in the present work since the effect of surface waviness was not taken into account. This would imply that the occurrence of splashing was predicted more precisely if the local inclination of film surface could be included in evaluating the impingement angle. In the latter case, it was inferred that the liquid flow formed below the film surface by the drop impact penetrated the film surface to produce secondary drops.

Water jet is utilized in various industrial application such as cutting carious materials and soil. In particular, jet grouting for soil improvement is one of the most important application of water jet under high pressure and high liquid flow rate. Such technology is already in practical use in civil engineering. In order to improve the efficiency and performance of jet grouting, it is quite important to clarify the hydrodynamic structure of water jet under high pressure and high flow rate. However, basic researches on this subject are quite insufficient both experimentally and analytically. Water jet utilized in jet grouting is usually very high speed ( up to 500 m/s) two-phase dispersed flow. Therefore, it is quite difficult to make visual observation of such water jet using conventional optical methods. In the present work, the authors utilized high sensitive CCD camera with image intensifier of which gate speed is 10 ns. Using this optical device, the authors obtained still image of high speed water jet for the first time. Visual observation revealed that high speed water jet is composed of very fine droplets and has complicated structures such as swirl and cluster of droplets. Velocity of water jet was also measured using two consecutive images taken between 10 microsecond. Such direct measurement of jet velocity of high speed water jet has not been carried out so far. The result indicated that jet speed (droplet speed) is even very high even at the boundary of water jet. Measure water jet velocity was reasonably correlated with jet velocity at the exit of nozzle.

Experimental studies were made on the multi-dimensional behavior of upward gas-liquid two-phase flow through the vertical round tube with an axisymmetric sudden expansion, which is one of the typical multi-dimensional channel geometry. The aims of this study are to clarify the multi-dimensional behavior of bubbly or slug flow affected by sudden expansion channel geometry, and to accumulate the experimental data for two-phase flow analysis, which is applicable to predict with appropriate accuracy the multi-dimensional its behavior. The direct observation using high-speed video camera was performed and revealed the multi-dimensional dynamic flow behavior with bubbles and gas-slug affected by the sudden expansion part (20 mm-tube to 50 mm-tube). The characteristic phenomena were observed such as bubble break-up, deformation due to the strong shear of liquid flow, or liquid micro jet penetration through the gas-slug, and so on. From these results, the flow regime map and the flow patterns at the below and above of the sudden expansion part were classified in relation to the bubble diameter. Additionally, the measurement of the void fraction profiles in the tube cross sections of sudden expansion were conducted at the different axial positions. The void fraction was measured using a point-electrode resistivity probe. The void fraction measurements in this study showed in detail that how the two phase flow develops along the direction of the downstream of the sudden expansion. In just above the sudden expansion, the void fraction distribution appeared the wall peak or the core peak in void fraction in the upstream of the sudden expansion. In the downstream of the sudden expansion, the void fraction distribution changes from a saddle shape or power-low shape into finally the saddle shape as the bubbly flow develops along the downstream of the sudden expansion. These experimental data in the sudden expansion of a vertical upward bubbly flow would become benchmark with respect to the multi-dimensional behavior of the two-phase flow analysis.

Incorporating the recently developed correlations for the rates of droplet deposition and entrainment with the film flow model, critical heat flux due to liquid film dryout in steam-water annular flow was predicted. In the present calculations, the predicted critical heat flux is affected not only by the deposition and entrainment rates but also by the entrainment fraction at the transition to annular flow (Transition quality was estimated by Wallis’s correlation). For this reason, the entrainment fraction at the transition was correlated in terms of dimensionless transition length to annular flow and density ratio. A simple correlation for the occurrence of dryout was also proposed. It was demonstrated by numerical simulations that the present method is to predict the available 1,340 data of dryout heat flux with the mean square relative error of 0.068.

Experiments were conducted to clarify the single bubble rise characteristics in turbulent flows in vertical flow channels. It was revealed that the rise velocity of a bubble relative to the time-averaged local liquid velocity could be much smaller in turbulent upflows than in stagnant liquids. The reduction of relative velocity was estimated to be caused by the two factors: turbulence in continuous phase and steep velocity gradient near wall; new correlations for describing these two effects were proposed. The relative velocity between the phases significantly affects the lateral phase distribution in multidimensional simulation of bubbly two-phase flow and the present correlations can give reasonable predictions for the relative velocity in turbulent flow. It is hence expected that the new correlations can contribute to the further improvement of the simulation models of bubbly two-phase flows.

Single bubble rise characteristics in turbulent flow in a vertical circular pipe were experimentally observed. As a result, it was found that bubble rise velocity relative to the time averaged local liquid velocity decreases with the increase of liquid flowrate and that it also decreases with the decrease of the distance between a bubble and wall. Investigating the mechanisms of the reductions of relative velocity, new correlations were developed for accurately evaluating the bubble rise velocities in turbulent flow. Since the predicted bubble rise velocities reasonably agreed with the measured data, the new correlations could contribute to the improvement of bubbly flow calculations in nuclear reactor safety analysis.

The effects of liquid temperature and injection method on single bubble rise characteristics in clean still water were experimentally investigated. It was confirmed that the shape and rise velocity of a bubble strongly depend on the method of bubble formation. For the rise velocity in high temperature water, the correlation for fully contaminated liquid might be appropriate even in the clean water. Because of the importance in the numerical simulation of bubbly two-phase flow, the experimental information on rise path oscillation such as amplitude and frequency was also reported.

Experimental and analytical studies have been carried out on the hydrodynamic behavior of swirling liquid film flow on a rotating disc. Film flow formation and swirling waves on the liquid film were analyzed through observation using high speed video. Liquid film thickness was measured using the Laser refraction method and compared with prediction. The rotating disc is 200 mm in diameter and was made of Silicon (Silicon wafer in industrial use). The rotating speed is up to 100 rad/sec (2000 rotations per min.) Water is supplied to the center of the disc at a flow rate of 8.3 × 10 −6 m 3 /s (500 cc/min). The film flow is divided into three regimes depending upon rotating speed. For the lower rotating speed (up to 10 rad/sec), formation of liquid film flow is incomplete and some part of the peripheral region of the disc is not completely covered by liquid film. For the intermediate rotating speed (15–25 rad/sec), laminar film flow covered the whole disc. Furthermore, there are swirling waves on the liquid film. This wave is considered to be a continuity wave arising at the center portion of disc due to the water flow rate variation form the nozzle. Wave propagation speed and behavior of these swirling waves were well explained by the theory of continuity wave. For the high rotating speed (more than 30 rad/sec), the liquid film flow changed its flow regime from laminar flow to turbulent flow. The estimated film Reynolds number at transition is about 1200 which is consistent with turbulent flow transition for pipe flow and film flow on non-rotating surface. Three dimensional turbulent waves were observed on this turbulent liquid film. The behavior of such three dimensional turbulent waves were quite random in time and space. Measured film thicknesses ranged from 50 to 300 micron. Film thickness and its fluctuation decreased as the rotation speed of disc increased and distance from disc center increased. The analysis was made on the film thickness based on the force balance between shear stress and centrifugal force acting on the film. The predicted film thickness agreed well with the measured value.