Initial Liquid Film Thickness Research Articles

EXPERIMENTAL STUDY OF INITIAL LIQUID FILM THICKNESS IN SQUARE MICROCHANNEL TWO-PHASE FLOW

EXPERIMENTAL STUDY OF INITIAL LIQUID FILM THICKNESS IN SQUARE MICROCHANNEL TWO-PHASE FLOW

Simulation and experimental validation of pulsating heat pipes

• Numerical model improvements for the thermal performance prediction of pulsating heat pipes. • Model validation with new experimental data for vertical and horizontal orientation. • Flow behavior analysis on the simulated results. The one-dimensional transient numerical code of JJ Cooling Innovation has been significantly improved to simulate the thermal/hydraulic performance of pulsating heat pipes (PHPs). Using a mechanistic approach, the liquid slug and vapor plug flows and their interaction with the surrounding wall are modeled. The heat transfer is predicted using the Three-Zone Model of flow boiling in microchannels, which includes liquid film evaporation and liquid film condensation of vapor plugs and convection of liquid slugs. Discretizing the entire PHP serpentine into one continuous grid, the code solves all variables based on local conditions. The code allows the user to define the working fluid, the channel’s shape, material and wall roughness, the serpentine’s pitch, the geometrical parameters of evaporator, adiabatic region and condenser zones, and the operating parameters. The boundary conditions in the evaporator zone can either be constant heat flux or constant temperature, while on the condenser side, the coolant temperature is constant as well as the heat transfer coefficient. The code works without any fitting factor and predicts all local variables in the 1D grid as a function of time, including the liquid film thickness and the onset of boiling for the formation of new vapor plugs within the evaporating and adiabatic zones, if local nucleation conditions are reached. Growth, collapse and coalescence of generated bubbles are taken into account. The film thermal resistance is corrected to a radial conduction model. The implementation of the initial liquid film thickness laid behind a liquid slug is updated to include the effect of the bubble velocity and acceleration as well as the length of the leading slug. The nucleation of new vapor plug has also been updated. The onset of nucleation is only dependent on the local thermodynamic state. If the threshold is reached, a new method is used to simulate micro-bubbles formation and coalescence to a channel-size vapor plug. These modifications greatly improve the code accuracy and stability allowing simulations with high pressure fluid rather than ethanol only. Furthermore, in the on-going EU project Pulsating Heat Pipes for Hybrid Propulsion Systems (PHP2), a new PHP test bench has been designed and built at Provides Metalmeccanica. A copper tube PHP with a 19-turn of 2.0 mm internal diameter and 12 mm tube pitch was tested over a wide range of imposed heat loads. A microchannel water-cooling plate was used at the condenser side with an inlet temperature of 20 °C. Then, tests were done for vertical and horizontal orientations with R1233zd(E) as the working fluid. The code, which was previously validated for ethanol PHP test data from KAIST, is now successfully validated for the present experimental data points. Accurate results were obtained from low heat loads up to 250 W for a set of 49 data points, predicting 97% within an error bandwidth of 30%.

Accelerating Taylor bubbles within circular capillary channels: Break-up mechanisms and regimes

In the present paper, an enhanced Volume of Fluid model is applied for the conduction of parametric numerical simulations, to investigate break-up phenomena of accelerating, elongated, vapour bubbles, within circular mini-channels. The effect of fundamental controlling parameters in the resulting break-up characteristics is investigated. Four different series of parametric numerical simulations of isolated vapour bubbles within mini-channels are performed, examining the effects of the imposed pressure difference between the inlet and the outlet of the channel, the surface tension, the effect of the applied heat flux as well as the initial liquid film thickness between the bubbles and the channel, on the developed vapour/liquid interface dynamics. The overall dimensionless number ranges examined are 6.76 < We < 1474.7, 0.007 < Ca < 0.14, 694 < Re < 12541 and by introducing a modified Froude number in order to account for the flow acceleration, 1 < Fr* < 21.86. These dimensionless number ranges are selected in order to overlap with experimental observations in zero-gravity Pulsating Heat Pipe experiments that constitute the motivation for the present numerical investigation. The proposed simulation results identify three prevailing regimes. A “full break-up” regime, a “partial break-up” regime and a “no break-up” regime. The entrainment of liquid droplets at the trailing edge of the vapour slugs is in most cases responsible for their subsequent “full break-up”, into a leading and a trailing bubble, as it is identified from the numerical simulations. Moreover, the applied heat flux does not influence the resulting break-up regimes. Finally, these identified break-up regimes, are grouped together into a well-defined flow map with respect to the We and Fr* numbers.

Theoretical and experimental study on liquid film thicknesses of unsteady slug flows in a capillary tube

Liquid film thickness is an important parameter for predicting evaporation and condensation heat transfer performance of two-phase slug flow in a capillary tube. Under flow boiling condition, the bubble of two-phase slug flow is elongated and accelerated. In the present study, the effect of acceleration on the liquid film thickness of unsteady two-phase slug flow is theoretically and experimentally investigated. It is first theoretically found that the accelerated slug flow is affected by the dimensionless numbers of capillary number (Ca), Weber number (We), and Bond number (Bo). The relation of the dimensionless liquid film thickness with the capillary number, Weber number, and Bond number for unsteady accelerated slug flows was obtained. The liquid film thicknesses of two-phase slug flows under acceleration condition were directly measured using laser focus displacement metering technique. Circular capillary tube with inner diameter of 1 mm was used for the test tube, and water, ethanol and FC-40 were used as working fluids. The novel prediction correlation of liquid film thickness for accelerated slug flows was proposed and it well predicts all the experimental data for all conditions. As the Bond number increases, the liquid film thicknesses are gradually shifted in the direction of decreasing film thickness due to acceleration effect. At a fixed capillary number, initial liquid film thickness becomes thinner with increasing Bond number.

Measurement and theoretical analysis of transient liquid film during micro-channel flow boiling

In this paper, three working fluids of deionized water, anhydrous ethanol and FC-72 were used to study the transient variation of liquid film thickness on the bubble's periphery for the flow boiling in micro-channels. The experiments showed that the liquid film declined rapidly under the coupled action of evaporation and shear force of vapor flow until it was dried out. A transition point existed on the curvature of liquid film thickness, and the liquid film thickness at the transition point (initial liquid film thickness) could be predicted by the Taylor flow, which was similar to the liquid film on the bubble's periphery in the adiabatic two-phase flow. Comparisons of the three working fluids showed that the reduction speed of the liquid film thickness depended on many thermophysical properties, including surface tension, liquid viscosity and latent heat. The heat flux also had evident influences on the liquid film behavior. An empirical model was proposed by considering evaporation of the liquid film and shear action on the vapor-liquid interface. The present correlation predicts the transient variation of film thickness within ±15% error and is applicable for all working fluids.

Liquid film thicknesses of oscillating slug flows in a capillary tube

Liquid film thickness is an important parameter for predicting boiling and condensation heat transfer coefficients in a microscale slug flow. In the present study, liquid film thickness of oscillating slug flow in a capillary tube is experimentally investigated under adiabatic condition. Laser focus displacement meter is used to measure the initial liquid film thickness. Circular tube with inner diameter of 1 mm is used for the test tube, and water and ethanol are used as working fluids. Measurement is carried out using a capillary tube with one open end and the other connected to a stepping motor. Driving frequency is ranged from 1 to 10 Hz at equivalent slug stroke of 31.7 mm and 51.2 mm. As the frequency and equivalent slug stroke are increased, the liquid film thickness deviates from that in the steady condition and becomes thinner or thicker under flow acceleration or deceleration, respectively. The empirical correlations for the initial liquid film thickness under acceleration and deceleration conditions proposed in the previous study (Youn et al., 2015, 2016) well predict the liquid film thicknesses of the oscillating flows within 15% accuracy.

Transient film thickness and microscale heat transfer during flow boiling in microchannels

The underlying mechanism of heat transfer for several sub-processes that control a bubble growth cycle during flow boiling in microchannels is studied in this article by means of high frequency measurements of liquid film thickness and temperature accompanying synchronous visualization. The test section is made up of glass tube having an internal diameter of 0.94mm with the Indium Tin Oxide (ITO) electrically conductive layer as heaters. The initial liquid film thickness formed by a boiling bubble slug agreed well with Taylor’s law. For most cases of the liquid film thickness evolution, the thinning of the liquid film was found to be much quicker than the prediction when considering evaporation effect alone. A new model for predicting the liquid film thickness evolution under flow boiling condition was developed by inclusion of the combination effect of evaporation and shear stress. It was found that the sub-processes of a bubble growth cycle were dependent on heat flux. The cyclical fluctuation of the temperature was due to the different heat transfer capability of the sub-processes. The time ratio and relative importance of the different heat transfer mechanisms were quantitatively estimated by linking the transient temperature fluctuations with the liquid film thickness variations.

The effect of bubble deceleration on the liquid film thickness in microtubes

Liquid film thickness is an important parameter for predicting boiling and condensation heat transfer coefficients in microtube slug flows. In the present study, the effect of bubble deceleration on the liquid film thickness is experimentally investigated under adiabatic condition. The laser focus displacement meter is used to measure the liquid film thickness. Circular tubes with three different inner diameters, D=0.7, 1.0 and 1.3mm, are used. Measurement is carried out using a microtube with one open end and the other connected to an actuator motor. Water, ethanol and FC-40 are used as working fluids. It is found that deceleration makes the liquid film thicker than that in the steady flow, and it deviates from the steady case as the deceleration rate is increased. Liquid film thickness remains nearly unchanged just after the onset of deceleration, and then gradually decreases and eventually converges to the steady thickness as the velocity is further decreased. Finally, an empirical correlation is proposed to predict initial liquid film thickness in decelerated flows.

The effect of initial flow velocity on the liquid film thickness in micro tube accelerated slug flow

Liquid film thickness is an important parameter for predicting boiling and condensation heat transfer in micro tubes. In the present study, the effect of initial flow velocity on the liquid film thickness in accelerated flows under adiabatic condition is experimentally investigated. The laser focus displacement meter is used to measure the initial liquid film thickness. Circular tube with inner diameter of 1mm was used for the test tube, and water, ethanol and FC-40 are used as working fluids. When the flow is accelerated from small initial velocities under small Bond number condition, initial liquid film thickness is identical to that of steady flow at small capillary numbers, and then deviates from the steady condition and eventually follows that of accelerated flow from zero initial velocity as capillary number is increased. When the flow is accelerated from large initial velocities, initial liquid film thickness deviates from the steady condition earlier and starts to follow that of accelerated flow from zero initial velocity as capillary number is increased. It is found that the initial flow velocity cannot be neglected in accelerated flows especially at large initial flow velocities and at large Bond numbers. Finally, an empirical correlation is proposed for the initial liquid film thickness of accelerated flows that accounts for the initial flow velocities.

An experimental investigation on bubble dynamics and boiling crisis in liquid films

Abstract This paper presents an experimental study of boiling and boiling crisis in a liquid film on a heater surface. The critical heat flux (CHF) values obtained in the present experiment mirror that of pool boiling, irrespective of initial liquid film thickness and liquid supply rate in the liquid film boiling case. This observation reinforces to the “scale separation” concept that high-heat-flux boiling and burnout are governed by micro-hydrodynamics in the liquid film on the heater surface. In addition to the CHF data, evolutions of bubbles and dry spots in the boiling liquid film are captured by means of high-speed high-resolution video camera. The dry spots were observed over surface heat flux ranging from 0.3 MW/m 2 to CHF, typically covering an area less than 10% of the heater surface. Three types of dry spot evolution are observed: (1) under the low heat flux, dry spots are rewetted by receding water dam upon rupture of corresponding bubbles; (2) as the heat flux reaches 1.25 MW/m 2 , dry spots rewetting is additionally aided by liquid flow driven by growth of bubbles nucleated in the vicinity; (3) upon approaching the CHF, dry spot(s) cannot be rewetted anymore and expand laterally, leading to boiling crisis (burnout of the heater surface). The richness of observations and characterization of micro-hydrodynamics in the present study further demonstrates that observations and measurements on boiling liquid films provide a paramount window for investigation and understanding of physical mechanisms of boiling and boiling crisis.

Experimental Characterization of Front Propagation by Marangoni Forces in Ultrathin Liquid Films

Italian physicist Carlo Marangoni published a treatise in 1865 which first described the spontaneous flow of a liquid film caused by gradients in surface t ension at a gas/liquid interface. Such gradients are established by variations in surface concentration of molecular species which lower the surface tension in proportion to the interface concentration. While the dynamics and power law behavior of the initial circular advancing front in surfactant coated films has been studied for about two decades, there has been relatively little work conducted on characterizing the fractal spreading patterns which develop behind this front. In this thesis, we examine the dynamics of front propagation and subsequent fractal evolution in ultrathin liquid films of glycerol driven to spread by concentration gradients in Aerosol OT, an insoluble anionic surfactant. We vary only the initial deposition volume and concentration while maintaining the initial liquid film thickness at 2.5 microns. Contrary to theoretical analyses in the literature, our results reveal that the coefficient and exponent characterizing the advance of both the first front as well as the following unstable front is not a constant depending only on geometry (i.e. rectilinear vs axisymmetric spreading) but varies with the initial surfactant concentration of the deposition volume. Furthermore, the overall speed of the fronts is observed to increase with ambient relative humidity, an effect we ascribe to a reduction in the viscosity of the glycerol film clue to absorption of water from the ambient atmosphere. Only in cases of relatively high humidity (higher than approximately 40% relative humidity) do we observe the formation of fractal patterns in the unstable front. Measurements of the fractal dimension yield values in the range 1.45-1.65. The largest values approximate those fractal dimensions observed in systems undergoing diffusion driven aggregation or viscous fingering, neither of which mechanism is pertinent to Marangoni flow.

Thermal boundary layer analysis corresponding to droplet train impingement

Simulations of droplet train impingement on a pre-wetted solid surface heated from below are used to study the thermal boundary layer behavior over a parameter space which includes variations in Reynolds, Peclet, and Weber numbers, as well as variations in inter-droplet spacing and initial liquid film thickness. Computationally, a modified version of the Volume-of-Fluid method is developed and employed in this study. The solver is validated against closed form solutions and additional experimental data from the literature. In combination with the simulations, an analytical representation is also developed and compared to the computations yielding favorable agreement. Results show that the boundary layer thickness is mostly affected by changes in inter-droplet spacing, Reynolds, and Peclet number, and influenced minimally by variations in Weber number and initial film thickness. In fact, it is explicitly demonstrated in the analysis that the impact velocity has the greatest effect in local heat transfer. An analytical expression for the Nusselt number radial profile is also developed. It shows that the Nusselt number scales as ∼ Re1/2, and its radial dependence is \documentclass[12pt]{minimal}\begin{document}$\sim \sqrt{r}$\end{document}∼r, which is the same as the circular jet impingement case. The notable difference in the present Nusselt number relationship is the role of inter-droplet spacing, which plays a significant role in the current configuration.

The effect of liquid film evaporation on flow boiling heat transfer in a micro tube

Flow boiling in micro channels is attracting large attention since it leads to large heat transfer area per unit volume. Generated vapor bubbles in micro channels are elongated due to the restriction of channel wall, and thus slug flow becomes one of the main flow regimes. In slug flow, sequential bubbles are confined by the liquid slugs, and thin liquid film is formed between tube wall and bubble. Liquid film evaporation is one of the main heat transfer mechanisms in micro channels and liquid film thickness is a very important parameter which determines heat transfer coefficient. In the present study, liquid film thickness is measured by laser focus displacement meter under flow boiling condition and compared with the correlation proposed for an adiabatic flow. The relationship between liquid film thickness and heat transfer coefficient is also investigated. Initial liquid film thickness under flow boiling condition can be predicted well by the correlation proposed under adiabatic condition. Under flow boiling condition, liquid film surface fluctuates due to high vapor velocity and shows periodic pattern against time. Frequency of periodic pattern increases with heat flux. At low quality, heat transfer coefficients calculated from measured liquid film thickness show good accordance with heat transfer coefficients obtained directly from wall temperature measurements.

Dynamics of the cavity and the surface film for impingements of single drops on liquid films of various thicknesses

This paper presents experimental and numerical investigations of single drop impacts onto liquid films of finite thickness. The dynamics of the drop impingement on liquid surface films, the shape of the cavity, the surface film dynamics and the residual film thickness are investigated and analysed. The shape of the penetrating cavity within the surface film is observed experimentally using a high-speed video system. Additionally, the thickness of the liquid film between the expanding, receding and retracting cavity and the solid wall is monitored in time using an optical sensor based on chromatic confocal imaging. The effects of various influencing parameters, such as the drop impingement velocity, liquid properties (surface tension and viscosity) and the initial liquid film thickness, on the time evolution of the cavity and film dynamics are investigated. Complementary to the experiments direct numerical simulations of the drop impacts and cavity expansion are performed using a volume-of-fluid free-surface capturing model in the framework of the finite volume numerical method. The numerical predictions of the film thickness dynamics agree well with the experiments for most phases of the impingement process. Finally, a scaling analysis of the residual film thickness between the cavity and the solid wall is performed for various impingement parameters.

The effect of bubble acceleration on the liquid film thickness in micro tubes

Liquid film thickness is an important parameter for predicting boiling heat transfer in micro tubes. In the previous study ( Han and Shikazono, 2009a), liquid film thickness under the steady condition was investigated and an empirical correlation for the initial liquid film thickness based on capillary number, Reynolds number and Weber number was proposed. However, under flow boiling conditions, bubble velocity is not constant but accelerated due to evaporation. It is necessary to consider this bubble acceleration effect on the liquid film thickness, since it affects viscous, surface tension and inertia forces in the momentum equation. In addition, viscous boundary layer develops, and it may also affect the liquid film thickness. In the present study, the effect of bubble acceleration is investigated. Laser focus displacement meter is used to measure the liquid film thickness. Ethanol, water and FC-40 are used as working fluids. Circular tubes with three different inner diameters, D = 0.5, 0.7 and 1.0 mm, are used. The increase of liquid film thickness with capillary number is restricted by the bubble acceleration. Finally, an empirical correlation is proposed for the liquid film thickness of accelerated flows in terms of capillary number and Bond number based on the bubble acceleration.

Measurement of the liquid film thickness in micro tube slug flow

Slug flow is one of the representative flow regimes of two-phase flow in micro tubes. It is well known that the thin liquid film formed between the tube wall and the vapor bubble plays an important role in micro tube heat transfer. In the present study, experiments are carried out to clarify the effects of parameters that affect the formation of the thin liquid film in micro tube two-phase flow. Laser focus displacement meter is used to measure the thickness of the thin liquid film. Air, ethanol, water and FC-40 are used as working fluids. Circular tubes with five different diameters, D = 0.3, 0.5, 0.7, 1.0 and 1.3 mm, are used. It is confirmed that the liquid film thickness is determined only by capillary number and the effect of inertia force is negligible at small capillary numbers. However, the effect of inertia force cannot be neglected as capillary number increases. At relatively high capillary numbers, liquid film thickness takes a minimum value against Reynolds number. The effects of bubble length, liquid slug length and gravity on the liquid film thickness are also investigated. Experimental correlation for the initial liquid film thickness based on capillary number, Reynolds number and Weber number is proposed.

Airblast atomization at conditions of low air velocity

Spray quality is determined by a series of measurements of drop size distribution and mean drop size. The atomizer employed is a two-dimensional design, producing a flat liquid sheet of variable thickness sandwiched between two coflowing nitrogen gas streams. The range of test conditions is chosen to simulate those typically found in a gas turbine combustor during startup, when air velocities are low. Four different liquids—tap water, heating oil, aviation kerosine (Jet-A), and a high-viscosity mixture of glycerine and distilled water—are used to investigate the effects of liquid physical properties on atomization quality. The results generally confirm those obtained in previous studies in regard to the effects of relative velocity, air/liquid mass flow ratio, and liquid physical properties on Sauter mean diameter. However, the significant dependence of mean drop size on initial liquid film thickness reported by other investigators is not observed in this study. Also, the results suggest the existence of a threshold of relative velocity below which atomization of the liquid sheet is not possible. As the relative velocity is reduced, the Sauter mean diameter eventually reaches a maximum value, and any further reduction in relative velocity results in the loss of the spray.

The Influence of Liquid Film Thickness on Airblast Atomization

The influence of initial liquid film thickness on mean drop size and drop-size distribution was examined using two specially designed airblast atomizers. Both were constructed to produce a flat liquid sheet across the centerline of a two-dimensional air duct with the liquid sheet exposed on both sides to high velocity air. In one case a thin film of uniform thickness was produced by injecting the liquid through a porous plate located just upstream of the atomizing edge. The film thickness, t, was then measured by a needle contact device. In the second design the fuel entered the air stream through a thin slot whose height could be adjusted accurately to vary and control the initial film thickness. Drop sizes were measured by the well-established light-scattering technique. From analysis of the processes involved, and from correlation of the experimental data, it was found that high values of liquid viscosity and liquid flow rate result in thicker films. It was also observed that thinner liquid films produce better atomization, according to the relationship, SMD ∝ t0.38.