Measured Heat Transfer Coefficients Research Articles

Flow regimes and heat transfer mechanisms affecting supercritical transition in microchannels

Supercritical carbon dioxide (sCO2) exhibits unique thermophysical and transport properties, which have the potential to enhance a wide range of thermal systems. Significant property variations accompanying the pseudocritical transition preclude accurate and generalized predictions of heat transfer, particularly at the microscale. A novel method for investigating fundamental fluid flow and heat transfer mechanisms through heat transfer coefficient measurements and side-view high-speed (8000 fps) schlieren imaging was developed. Experiments are conducted in a square microchannel (Dh=500μm) at reduced pressure near unity (PR=1.05) over a range of heat and mass fluxes (qc″=1.3−82.6 Wcm−2; G=280−1380 kgm−2s−1). Non-uniform density profiles within the boundary layer and distinct, freely mixed liquid-like and gas-like packets at various stages of pseudocritical transition were observed. Three flow regimes were identified as a function of heat flux with unique convection boundary layer characteristics. Transport of liquid-like and the production of gas-like sCO2 at the wall were found to be the primary mechanisms affecting heat transfer and were quantified using a modified form of the Richardson number. The experimental approach and mechanistic insight developed herein provide a basis for high-fidelity heat transfer models for the design of supercritical fluid systems.

A methodology for measuring heat transfer coefficient and self-similarity of thermal regulation in microvascular material systems

Fluid transport through microvascular networks—a hallmark for homeostasis in living systems—has transcended to engineered materials, primarily made possible because of modern manufacturing advancements. Vascular-enabled multifunctionality, including thermal regulation and self-healing, holds great potential for extending the lifetime of structural materials and expanding the operational envelope. Prior studies on vascular-based active cooling use a “combined” heat transfer coefficient (HTC): a single parameter lumps convection and radiation effects. Although the resulting mathematical models are linear—an attractive feature for computational modeling, the combined coefficient approach may not be accurate or even applicable if the operating temperature is unknown, which is the case with many thermal regulation applications (e.g., space probes). In this paper, we illustrate the remarked limitations of the lumped approach and advocate the need to use a decoupled HTC by splitting convective and radiative heat transfer modes. We show the broad applicability of the proposed method by applying it to three material systems: glass and carbon fiber-reinforced polymer composites and an additive manufactured metal. We show, using numerical simulations, the differences in the predictions from the decoupled approach with that of the combined HTC; these differences are prominent at higher heat fluxes. Also, the decoupling has enabled us to establish a scaling law that allows transferring of solutions fields across material systems, strengthening further the validity and utility of our approach. This work's significance is two-fold. First, the research is fundamental, providing accurate measurement protocols for critical model parameters. Second, this work facilitates the development of mathematical models for vascular-based thermal regulation that are predictive even for hostile environments (which are often difficult to realize in laboratories), such as outer space.

Assessment of Binary Pressure-Sensitive Paint for Temperature and Heat Transfer Coefficient Measurement of Leading-Edge Film Cooling

Abstract Film cooling is a common technique for protecting gas turbine components from the hot combustor exhaust. Highly resolved film cooling effectiveness distributions are often obtained by measuring the mass transfer of a foreign gas coolant in mainstream air using pressure-sensitive paint (PSP). However, PSP is not able to measure the heat transfer coefficient, which is necessary to fully quantify the impact of film cooling. Instead, binary pressure-sensitive paint (BPSP) has an additional luminophore that is sensitive to temperature and can be used to measure the heat transfer coefficient. In this experiment, the film cooling effectiveness and heat transfer coefficient were measured using BPSP on the leading edge of a cylinder. The cylinder had a 7.62-cm diameter with two rows of cooling holes at ±15 deg from the leading edge. Each row contained ten holes with a 0.475-cm diameter, spaced 4 diameters apart in the spanwise direction, and angled 30 deg from the cylinder axis. The mainstream Reynolds number was 100,000 based on cylinder diameter with a turbulence intensity of 7.1%. The coolant-to-mainstream density ratio was 1.0, and the blowing ratio was 0.8. The heat transfer coefficient was measured in a transient heat transfer experiment using the reference signal from the BPSP. Despite the high uncertainty of the measurement, ranging from 24.0% to 71.1%, the results demonstrate the feasibility of the method and identify the best test methodology to minimize conduction errors.

Heat transfer of bubbly flows in microchannels at varied aspect ratios and hydraulic diameters

With the fast progress of new technologies applied to cloud computing, artificial intelligence, and fifth generation communication, the researches on the heat dissipation of high-power electronic devices have concentrated on the two-phase flows in the microchannel, with the phase change used to achieve excellent cooling efficiency. The objective of this paper is to conduct the computational and experimental studies for investigating the thermal-fluid behaviors of subcooled flow boiling in microchannels. In the computational fluid dynamics (CFD) model, the numerical solver is built in ANSYS/Fluent® implementing the multiphase formulation with the full consideration of the effects of turbulence, surface tension, and phase change to appropriately imitate the interfacial movements between liquid water and vapor for probing the subcooled boiling and bubble nucleation processes over the microchannel. In the experimental investigation, a syringe pump and a programmable DC power supply is employed to regulate the heat flux and subcooled temperature for determining the heat transfer and pressure drop outcomes across the microchannel at varied aspect ratios and hydraulic diameters. In addition, this study sets up a high-speed camera with a LED fiber optical light source to acquire the close-up observation images over the complex flow boiling phenomena. The accuracy of CFD predictions is assessed by comparison against both the measured heat transfer coefficients and pressure drops as well as the photo-captured interfacial behaviors. The numerical simulations are then conducted to resolve the temperature gradients on the heating wall surface due to intense disturbances incited by the confined bubbly flow and sweeping flow at high aspect ratios, enhancing the cooling performance over the microchannels. The experimental measurements show that a reduction in inlet subcooling from 65 to 50 °C tends to enlarge the average heat transfer coefficient and pressure drop by 21.5 and 17.1%, respectively. The design impact study also reveals the average heat transfer coefficient and pressure drop at an aspect ratio of 5 greater than those at an aspect ratio of 2 by up to 41.1 and 27.2%, respectively. In contrast, increasing hydraulic diameter from 0.92 to 1.38 mm can strengthen the thermal and frictional outcomes by 17.2 and 20.3%. The microchannel design having an aspect ratio of 5.0 can realize the estimated coefficient of performance (COP) up to 116,277.8, achieving the satisfactory overall thermal and frictional outcomes.

Measurements of heat transfer coefficients from supercritical fluid flowing in vertical mini channels with constant wall temperature

Currently, efficient heat transmission for compact electronic elements is an essential matter. It needs a heat sink with a liquid cooling scheme that meets these demands as much as feasible. The dimensions of 50.8 × 40.6 × 5.5 mm were adopted for features of heat transfer as well as the fluid flow of supercritical CO2 in the heat sink in this study. The adopted pressures, tem-peratures, and mass velocity ranges were 7.5 to 12 MPa, 35 to 50oC, and 100 to 500 Kg/m2s, respectively, wherein the CO2 cooled under these conditions. The factors of heat transfer, the pressure at levels of local as well as medium degree were determined under these conditions. The medium temperature of CO2 in the adjacent significant point area increased, the pressure decreased and the medium temperature movement factor augmented dramatically. It was also noted that the medium temperature movement factor peaked at the pseudo-critical tempera-ture. However, the maximum temperature movement factor declined increased pressure. Fur-thermore, in contrast to the pressure factor, it was revealed that mass velocity and temperature movement factor had a direct relationship. Using the obtained data, a novel correlation mech- anism for limited convection of super-critical CO2 in regular multi-port micro tubes based on chilling conditions was constructed using the obtained coefficients in this study.

Importance of Kv Distribution in Freeze Drying Part I: A Holistic Model to Predict Changes in Kv Bimodal Distribution as a Function of Pressure

Measurement of heat transfer coefficients (Kv) is an important part of freeze-dryers characterization and as well a necessary step for executing any modelling. In most cases only an average value of Kv is calculated, or an average value of center and edge vials is provided. Our aim is to go a step further and to describe the overall Kv distribution various vial/ freeze drier combinations, whatever the pressure. From an experimental point of view, in this article we propose three methods to calculate the Kv value for individual vials based on the ice sublimation gravimetric method. The first method we use is the most usual one, where the Kv value is calculated based on the mass of sublimated ice and the product temperature measured in selected vias. In the second method, the average product temperature is estimated for each vial, based on the mass difference before and after sublimation and the Kv value is calculated accordingly. In the third method, the Kv is estimated by comparison to sublimation results from a simulation. Results from methods 2 and 3 are very similar results and are slightly different from those of method 1. Method 1 was shown to exhibit a systematic bias due to the fact that it is based on the temperature of recording of selected vials only, which are not representative for all positions.Once the individual values of Kv have been calculated, it is possible to establish a distribution for each method. It was observed that an overlay of two normal distributions describing the center and the edge vials provides a good representation of the empirical distribution. Furthermore, we propose a holistic model aiming to calculate the Kv distribution for any specified pressure.

Research on Heat Transfer Coefficient Measurement Based on Heat Exchanger Experiment

Heat transfer occurs all the time through substances in any phases by means of conduction, convection and radiation. Therefore, it becomes the fundamental concept in thermodynamics and in mechanical and chemical engineering, which requires designers to be familiar with. A large quantity of correlation formulas dealing with distinct situations have already obtained by plenty of previous testing experience.

The Efficient Way to Design Cooling Sections for Heat Treatment of Long Steel Products.

To achieve the required mechanical properties in the heat treatment of steel, it is necessary to have an adequate cooling rate and to achieve the desired final temperature of the product. This should be achieved with one cooling unit for different product sizes. In order to provide the high variability of the cooling system, different types of nozzles are used in modern cooling systems. Designers often use simplified, inaccurate correlations to predict the heat transfer coefficient, resulting in the oversizing of the designed cooling system or failure to provide the required cooling regime. This typically results in longer commissioning times and higher manufacturing costs of the new cooling system. Accurate information about the required cooling regime and the heat transfer coefficient of the designed cooling is critical. This paper presents a design approach based on laboratory measurements. Firstly, the way to find or validate the required cooling regime is presented. The paper then focuses on nozzle selection and presents laboratory measurements that provide accurate heat transfer coefficients as a function of position and surface temperature for different cooling configurations. Numerical simulations using the measured heat transfer coefficients allow the optimum design to be found for different product sizes.

Condensation heat transfer of superheated vapour of R1234ze(E) and R134a inside a brazed plate heat exchanger

The recent restrictions on the use of refrigerants with high Global Warming Potential (GWP) have pushed the research to consider alternative solutions, such as the HydroFluoroOlefins, as possible replacements. Even though they are generally mildly flammable, their GWP is significantly lower compared to those of the HydroFluoroCarbons. This paper presents experimental measurements of R1234ze(E) and R134a heat transfer coefficients during condensation inside a brazed plate heat exchanger. Experimental tests during condensation heat transfer have been conducted considering different degrees of superheating, subcooling and outlet vapour quality. The condensation temperature varies between 34.6 °C and 42.3 °C and the refrigerant mass flux is between 9 kg m−2 s−1 and 49 kg m−2 s−1. The results showed that the heat transfer coefficients measured with R134a are between 4% and 8% higher than those of R1234ze(E). Complete condensation experiments showed that an increase in the liquid subcooling degree significantly reduces the thermal performance at low refrigerant mass velocities. In some cases, the plate area occupied by liquid refrigerant reached almost 30% of the overall heat transfer area, thus decreasing by 2.7 times the average heat transfer coefficient when passing from 3 K to 8 K subcooling degree. A comparison with the predictions of some empirical models is also presented to assess which one can better predict the experimental data.

Transient and periodic steady-state characteristics of the local heat transfer measurement by thermal perturbation with Gaussian power density distribution

The local heat transfer coefficient measurement with temperature oscillation induced by periodic thermal perturbation – usually via a Gaussian laser beam, was investigated for the impact of the spikiness (i.e., the standard deviation) elaborated in comparison with the analytical model for dimensional analysis. The statistically more robust technique that relies on the linearity of the spatial phase distribution of the test point array was favored when the target Biot number approaches unity in terms of its order of magnitude. The preferred upper limit for thermographic scanning was discussed as the simplification of later data processing is concerned. Nonetheless, the time elapsed for an acceptable periodic steady state, which in principle leans to the higher end of the target Biot number spectrum in a log scale, indicates the benefit from the time series of pointwise temperature measurement – as in the conventional single-blow testing, where the effect of spikiness, as well as that of the location of individual test point, holds. Note that the vicinity as the target Biot number approaches unity was again observed with higher preference.

Investigation and measurement of crude oil heat transfer coefficient in forced convection and subcooled flow boiling heat transfer

Investigation and measurement of crude oil heat transfer coefficient in forced convection and subcooled flow boiling heat transfer

Effect of particle size distribution on heat transfer in bubbling fluidized beds applied in thermochemical energy storage

The aim of the present work is to characterize the effect of particle size changes from cycling of the thermochemical energy storage material CaO/Ca(OH)2 on fluidizability and wall-to-bed heat transfer coefficients in the fluidized bed. Materials replicated from previous storage cyclization experiments are experimentally investigated in an ambient fluidization test rig. Differential pressure measurement and a horizontally immersed heat transfer probe are used to characterize fluidizability and wall-to-bed heat transfer between immersed cylinder and bed material. Sauter mean diameters of the replicated particle size distributions range from 27µm to 282µm. Particle size distributions with Sauter mean diameters significantly smaller than approximately 50µm were found to be difficult to fluidize due to excessive channeling, resulting in low heat transfer coefficients. In this case, fluidization quality and heat transfer is enhanced for higher gas velocities. Fluidizable particle size distributions ranged from Sauter mean diameters of 48µm to 282µm. Measured heat transfer coefficients show little dependency on the particle size. Observed maximum heat transfer coefficients range from 344Wm−2K−1 to 350Wm−2K−1.

Theoretical modeling of heat transfer in vertical upward and downward annular flow boiling

This work entails experimental measurement and theoretical modeling of heat transfer coefficient (HTC) for annular flow boiling in upward and downward flow configurations. The working fluid used was HFE-7000 and experimental measurements were carried out inside a 6 mm sapphire tube coated externally with indium-tin-oxide (ITO) for Joule heating. The range of vapor quality, mass flux, and heat flux investigated were 0.15 − 0.7, 75–400 kg/(m2s), and 0.5 − 3.0 W/cm2, respectively. Theoretical models for predicting HTC in upward and downward flows were developed using heat-flux-dependent wall shear stress correlations and roll-wave-velocity-based interfacial damping function. It was found that interfacial damping depends on the Reynolds number of the liquid film. The proposed models predicted over 96% of the measured HTC within ±20% in both upward and downward flows and reproduced the heat flux dependence of the HTC. The models also predicted over 96% of the measured liquid film thickness within ±30% in both upward and downward flows.

Heat Transfer Coefficient and Adiabatic Wall Temperature Measurements on High-Pressure Turbine Nozzle Guide Vanes With Representative Inlet Swirl and Temperature Distortions

Abstract Combustor exit conditions in modern gas turbines are generally characterized by significant temperature distortions and swirl degree, which in turn is responsible for very high turbulence intensities. These distortions have become particularly important with the introduction of lean combustion, as a mean to control NOx pollutant emissions. For this reason, combustor–turbine interaction studies have recently gained a lot of importance. Past studies have focused on the description of the effects of turbulence, swirl degree, and temperature distortions on the behavior of the high-pressure stages of the turbine, both considering them as separated aspects and accounting for their combined impact. Aspects like pressure losses, hot streaks migration, and film-cooling behavior have been widely investigated. Even if some studies have focused on the characterization of the heat transfer coefficient (HTC) on the nozzle guide vane external surface, none of them have addressed this aspect from a purely experimental point of view. Indeed, when inlet conditions are characterized by both swirl and temperature distortions, they represent a severe challenge for the commonly adopted measurement techniques. The work presented in this paper was carried out on a non-reactive, annular, three-sector test rig made by a non-reactive combustor simulator and a nozzle guide vane cascade; it is able to create a representative combustor outflow, characterized by all the flow characteristics described before. A novel experimental approach, which was developed in a previous work, was exploited to experimentally retrieve the heat transfer coefficient and the adiabatic wall temperature distributions on a non-cooled nozzle guide vane. Temperature measurements on the cascade inlet and outlet planes were also used to provide boundary conditions and achieve a better understanding of the investigated phenomena. The results allowed to evidence the effect of the inlet swirl on the heat transfer coefficient distribution, as well as the evolution of the temperature distribution on the vane surface moving through the cascade, constituting the first attempt to evaluate these aspects from a purely experimental point of view.

Prediction of effective heat transfer coefficients for vapour condensation inside horizontal tubes in stratified phase flow

In modern condensers of air conditioning systems, heat pumps, evaporators of seawater desalination systems, and heaters of power plants, the process of vapour condensation is carried out mainly inside the horizontal tubes and channels. Heat transfer processes occurring in condensers have a significant effect on the overall energy efficiency of the mentioned systems. In this paper, the experimental investigation of heat transfer during condensation of freons R22, R406a, and R407c in the plain smooth tube with d = 17 mm were carried out with the following parameters: ts = 35–40°C, G = 10–100 kg/(m2s), x = 0.8–0.1, q = 5–50 kW/m2, ΔT = 4–14 K. The unique measurements of circumferential heat fluxes and heat transfer coefficients were carried out with the thick wall method during different condensation modes. It can be inferred that with the increase of the heat flux, at the top part of the tube the thickness of the condensate film increases, which leads to the decrease in heat transfer. At the bottom of the tube, the increase in the heat flux enhances heat transfer coefficient, that is characteristic of the turbulent liquid flow in the tube. The obtained results allowed improving the prediction of effective heat transfer coefficients for vapour condensation, which takes into account the influence of condensate flow in the lower part of the tube on the heat transfer. This method generalises with sufficient accuracy (error ± 30%) the experimental data on condensation of freons R22, R134a, R123, R125, R32, R410a, propane, isobutene, propylene, dimethyl ether, carbon dioxide, and methane under stratified flow conditions. Using this method for designing heat exchangers, which utilise such types of fluids, will increase the efficiency of thermal energy systems.

Experimental Measurement and Theoretical Prediction of Bubble Growth and Convection Heat Transfer Coefficient in Direct Contact Heat Transfer

The measurement of the two-phase contact area is very important to determine the heat transfer coefficient in the process of direct contact heat transfer, but the direct measurement of the two-phase contact area is a difficult problem. The experiments are carried out utilizing a cylindrical Perspex tube of 100 cm in total height and 15 cm inner diameter. The active column height throughout the experiments is taken to be equal to 50 cm. Liquid Therminol®66 with four different initial temperatures (50 °C, 60 °C, 70 °C and 80 °C) is used as a continuous phase, while liquid R245fa at a constant temperature of 23 °C is used as a dispersed phase. In this paper, the empirical correlations between bubble growth and local convection heat transfer coefficient are obtained through modeling and measurement, and its correctness is verified by experiments. The results show that the bubble diameter is positively correlated with continuous phase temperature, flow rate ratio, and height, but the local convection heat transfer coefficient is negatively correlated with continuous phase temperature, flow rate ratio, and height. At the same time, it is found that the maximum error between the actual bubble diameter and the theoretical bubble diameter is 7%, and the error between the heat flux calculated by the local convection heat transfer coefficient and the actual heat flux is within 10%. This study provides theoretical guidance for an in-depth understanding of the direct contact heat transfer process and the development of high-efficiency waste heat recovery systems.

Evaluation of bed-to-tube heat transfer in a fluidized bed heat exchanger in a 75 MWth CFB boiler for municipal solid waste fuels

Bed-to-tube heat transfer has been investigated for a tertiary superheater in a 75 MWth Circulating Fluidized Bed (CFB) boiler in Norrköping, Sweden. The boiler is used for incineration of solid waste fuels. Two fluidized bed heat exchangers are located in loop seals, connecting the cyclones and the furnace. The heat exchangers are placed in series, with respect to the steam side, and in parallel, with respect to the particle side. The total heat transfer surface area is roughly 44 m2, distributed over 72 tubes. The total effect transferred most often is in the range 2–6 MW. The incoming steam temperature in the first superheater is 380–400 °C, while the exiting steam temperature from the second is around 450°, at 65 bar pressure. The bed temperature in the Fluidized Bed Heat Exchanger (FBHE) is 850–875 °C. The analysis is based on operational data from two time periods (2002–2005 and 2014–2021). The two periods use different heat exchanger designs, following a retrofit in 2005. The aim of the study is to establish the bed-to-tube heat transfer coefficient in an industrial FBHE unit and investigate how it varies over different time periods, for two different bed materials and for two different designs. Also, the experimentally determined heat transfer coefficients are compared with an established heat-transfer correlation, for prediction of heat transfer from bubbling fluidized bed to tubes. Operation with two bed materials were evaluated, namely silica sand and crushed and beneficiated ilmenite. Both materials are classified as Geldart B particles. Air is used as fluidization gas in the FBHE. The analysis show, with a few exceptions, comparably low heat-transfer coefficients from bed to tube of 100–150 W/(m2K). The results were similar for silica sand and ilmenite, but the highest measured heat transfer coefficient was for a period with ilmenite. The heat transfer was lower than expected based on literature data from FBHE units and fluidized bed boilers in general, and much lower than bed-to-tube heat transfer coefficients from lab-scale experiments and empirically derived predictive expressions. The difference could be related to one or more of several factors, such as the effect of very small tube spacing, unknown thermal conductivity of one of the layers in the tube bundle, the effect of lateral particle flow and the effect of fouling due to ash layers forming on the tube surfaces. It is suggested that it should be possible to significantly increase the bed-to-tube heat transfer by increasing the tube pitch, which is expected to improve bed mixing without increasing the risk of corrosion.

Effects of heat transfer coefficient variations on composite curing

Curing of composite laminates in a vessel was investigated in this study. The environment inside the processing vessel dictates the efficiency and ultimately drives the quality of thermoset composite parts. Experimental measurements of spatial heat transfer coefficients were conducted on industrial scale vessels, including autoclaves and large ovens, which ultimately drives the quality of thermoset composite parts. The final part quality was investigated using the experimental data as input to a coupled heat transfer and curing model. Measurements showed that heat transfer coefficients in autoclaves were greater in magnitude and spatial variability. The distribution in the autoclaves followed a pattern common in the literature, in contrast to that in the ovens which varied considerably between devices. Numerical predictions indicated autoclave measured heat transfer coefficients provide less lag to the imposed temperature history and smaller temperature overshoots. However, the greater robustness to variability at autoclave heat transfer coefficients was offset by the greater variability, resulting in comparable robustness across the ovens and autoclaves.

Hydrocarbon refrigerants boiling local heat transfer coefficients inside a Brazed Plate Heat Exchanger (BPHE)

This paper presents the local heat transfer coefficients of R290 (propane) and R1270 (propylene) boiling inside a Brazed Plate Heat Exchanger (BPHE). A new test section was designed for measuring the local heat transfer coefficient in refrigerant two-phase heat transfer in a BPHE. The test section includes two thick AISI316L stainless steel corrugated plates instrumented with 36 T-type thermocouples for measuring the local plate surface temperature, the local heat flux and the local heat transfer coefficient in 9 positions along the refrigerant channel. The experimental tests were carried out at a boiling temperature of 10 °C, in the refrigerant mass flux range 5 − 17.5 kg m−2 s−1 with an inlet vapour quality ranging from 0.24 to 0.37 and an outlet vapour quality around 1.00. The global uncertainty of the local heat transfer coefficient measurement is within ±12.7%, and ±15.6%, for R290, and R1270, respectively. Nucleate boiling seems to be the governing heat transfer regime both in R290 and R1270 tests. Consolidated correlations for nucleate boiling inside BPHE show a fair agreement with all the experimental data. R290 and R1270 exhibit boiling heat transfer performance between those of R404A and those of R410A.

Stratification and heat transfer characteristics of partially miscible lubricating oil-refrigerant mixtures during pool boiling

• Partially miscible oil-refrigerant mixture stratifies into four layers at boiling. • Newly proposed nucleate boiling model reflects effects of mixture stratifications. • Bubble dynamics and concentration diffusion determine mass transfer among layers. • Predicted thicknesses of separated layers agree well with observation results. • Mean deviation between predicted and measured heat transfer coefficients is 15.3% Applying partially miscible lubricating oil-refrigerant mixtures instead of completely miscible ones in refrigeration systems is a promising way to avoid the refrigerant insufficiency, and the application of partially miscible oil requires the clarity of nucleate pool boiling characteristics of partially miscible mixtures. As partially miscible mixtures at boiling conditions show an entirely different stratification process compared to completely miscible mixtures or partially miscible mixtures at non-boiling conditions studied by former researchers, the existing researches cannot be directly applied to evaluate pool boiling characteristics of the partially miscible mixture. The purpose of this study is to experimentally investigate and model both stratification and heat transfer performance of nucleate pool boiling of the partially miscible oil-refrigerant mixture. The stratification phenomenon at boiling of partially miscible mixtures is observed, and up to four layers (i.e., an oil-rich micro-layer, a middle oil-poor layer, a top oil-rich layer, and a gas phase layer) are found. The boiling process of partially miscible mixtures stratifying from a single layer into four layers is modelled, and calculation methods of mass transfer rates among layers are proposed. The model validation shows that predicted thicknesses of layers accord with experimental images, which covers the conditions of heat flux ranging from 5 to 15 kW/m 2 , initial oil mass fraction ranging from 5 to 15%, and initial height ranging from 80 to 120 mm. Predicted heat transfer coefficients agree with 80% of the experimental data within the deviation of ± 25%, and the mean deviation is 15.3%.