Search

• Resistively heated, gravity driven falling films were studied experimentally. • An optical technique featuring simultaneous PLIF, PTV and IR was developed and applied. • Spatiotemporally resolved thickness, velocity and interface temperature data was generated. • Wave/phase-locked averaged thickness, velocity and temperature information is recovered. • Local/instantaneous heat transfer data along the wave topologies are provided. We present an optical technique that combines simultaneous planar laser-induced fluorescence (PLIF), particle tracking velocimetry (PTV) and infrared (IR) thermography for the space-and time-resolved measurement of the film-height, 2-D velocity and 2-D free-surface temperature in liquid films falling over an inclined, resistively-heated glass substrate. Using this information and knowledge of the wall temperature, local and instantaneous heat-transfer coefficients (HTCs) and Nusselt numbers, Nu , are also recovered along the waves of liquid films with Kapitza number, Ka = 180 , and Prandtl number, Pr = 77. By employing this technique, falling-film flows are investigated with Reynolds numbers in the range Re = 18–66, wave frequencies set to f w = 7 , 12 and 17 Hz, and a wall heat flux set to q ̇ = 2.5 W cm−2. Complementary data are also collected in equivalent (i.e., for the same mean-flow Re) flows with q ̇ = 0 W cm−2. Quality assurance experiments are performed that reveal deviations of up to 2–3% between PLIF/PTV-derived film heights, interfacial/bulk velocities and flow rates, and both analytical predictions and direct measurements of flat films over a range of conditions, while IR-based temperature measurements fall within 1 °C of thermocouple measurements. Highly localized film height, velocity, flow-rate and interface-temperature data are generated along the examined wave topologies by phase/wave locked averaging. The application of a heat flux ( q ̇ = 2.5 W cm−2) results in a pronounced "thinning" of the investigated films (by 18%, on average), while the mean bulk velocities compensate by increasing by a similar extent to conserve the imposed flow rate. The axial-velocity profiles that are obtained in the heated cases are parabolic but "fuller" compared to equivalent isothermal flows, excluding any wave-regions where the interface slopes are high. As the Re is reduced, the heating applied at the wall penetrates through the film, resulting in a pronounced coupling between the HTC and the film height in thinner film regions. When the imposed wave frequency is increased, a narrower range of HTCs is observed, which we link to the evolution of the film topology and the associated redistribution of the fluid flow upstream of the imaging location, as the liquid viscosity decreases. The local and instantaneous Nu is strongly coupled to the film height and experiences variations that increase as f w is reduced. [ABSTRACT FROM AUTHOR]

The Up-THERM heat converter is an unsteady, two-phase thermofluidic oscillator that employs an organic working fluid, which is currently being considered as a prime-mover in small- to medium-scale combined heat and power (CHP) applications. In this paper, the Up-THERM heat converter is compared to a basic (sub-critical, non-regenerative) organic Rankine cycle (ORC) heat engine with respect to their power outputs, thermal efficiencies and exergy efficiencies, as well as their capital and specific costs. The study focuses on a pre-specified Up-THERM design in a selected application, a heat-source temperature range from 210 °C to 500 °C and five different working fluids (three n-alkanes and two refrigerants). A modeling methodology is developed that allows the above thermo-economic performance indicators to be estimated for the two power-generation systems. For the chosen applications, the power output of the ORC engine is generally higher than that of the Up-THERM heat converter. However, the capital costs of the Up-THERM heat converter are lower than those of the ORC engine. Although the specific costs (£/kW) of the ORC engine are lower than those of the Up-THERM converter at low heat-source temperatures, the two systems become progressively comparable at higher temperatures, with the Up-THERM heat converter attaining a considerably lower specific cost at the highest heat-source temperatures considered. [ABSTRACT FROM AUTHOR]

This paper deals with the dynamic modeling and multiparametric optimization of a thermally powered, reciprocating/oscillator fluid pumping technology known as the Non-Inertive-Feedback Thermofluidic Engine (NIFTE), with the aim of proposing pathways for the improvement of the capabilities of this technology. Analogies are drawn between the physical device and an electric circuit. All system parameters are perturbed simultaneously and independently, and configurations exhibiting the highest efficiencies, highest pumped volumetric flow rates, and lowest feedback gains (corresponding to lowest temperatures or heat inputs necessary for continuous operation) are examined. A sensitivity analysis is also undertaken. Low feedback tube resistances and inductances, low thermal resistances, and also low adiabatic volume capacitances all allow for high efficiencies and flow rates, as do somewhat elevated (relative to a preselected “nominal” design) load resistances and inductances, and displacer cylinder capacitances and inductances. The power cylinder inductance has its optimum near its selected nominal value. Generally, the power and displacer cylinder capacitances do not significantly affect the investigated performance indicators. The heat exchangers, vapor volume, and power cylinder are identified as important components that control the performance of the device. The optimal parameter values can be taken as useful indications toward the design of an improved NIFTE pumping device. [ABSTRACT FROM PUBLISHER]

Abstract: A two-phase thermofluidic oscillator was recently reported as being capable of undergoing sustained operation when a constant and low temperature difference is applied to the device, which consists of a network of tubes, compartments and two heat exchanger blocks. Within this arrangement a working fluid undergoes thermodynamic property oscillations that describe a heat engine cycle. Previous attempts to model the dynamic behaviour of this thermofluidic engine for performance predictions have been based on linear analyses. These have provided us with useful knowledge of the necessary minimum temperature difference for operation, and the resulting oscillation frequency and efficiency. However, experimental observations suggest a limit cycle operation associated exclusively with nonlinear systems. The present paper presents an effort to devise a nonlinear model for the device. Indicative results from this model are discussed, and the predictions are compared to those from the linear equivalents and experimental observations. The results reveal that although both linear and nonlinear models predict similar oscillation frequencies, the nonlinear model predicts lower exergetic efficiencies. This probably arises from the inability of the linear representation in the thermal domain to capture the saturation in the rate of heat exchange between the working fluid and the heat exchangers. The present effort aims to provide a better understanding of this device and to suggest improved design guidelines for increased efficiency and power density. [Copyright &y& Elsevier]

In this paper, we consider the flow and heat transfer behaviour of turbulent upward and downward flows of supercritical n-decane, in order to reveal the features of heat transfer deterioration (HTD) that would be expected in relevant active regenerative cooling systems for scramjet engines. Specific focus is placed on key velocity-field features that appear in these flows. Following the validation of six turbulence models, the SST k - ω and RNG k - ϵ models are found to be suitable for simulating the upward and downward flow cases, respectively. "M" type velocity profiles (a non-monotonicity of the velocity along the radial direction) are observed, which arise due to a spatially-varying interplay between the inertial and viscous forces in the flow domain, while larger velocity gradients in the buffer layer are also observed that contribute to the phenomenon of HTD. Furthermore, it is found that the secondary flows as well as the different mass fluxes that arise due to the velocity increase from the wall to the flow core zone (i.e., the influencing range and intensity of cross-sectional kinetic energy), respectively, are observed in the HTD development region, as well as the HTD peak area and degradation regions. A zone of higher thermal diffusion appears in the near-wall region, which acts as a thermal barrier and contributes to HTD. • The inertial force and the viscous force play a major role in the upward and downward flow, respectively. • A large velocity gradient in the buffer layer and the non-monotonicity of velocity contribute to the HTD. • A thermal barrier is generated under the large heat diffusion near the heated wall and thus causes the HTD. [ABSTRACT FROM AUTHOR]

The artucle introduces a special issue of the journal "Heat Transfer Engineering" which contains a collection of papers that were presented at UKHTC 2013: The 13th UK Heat Transfer Conference held at Imperial College London in England on September 2 and 3, 2013.

Several of the emerging technologies for electricity storage are based on some form of thermal energy storage (TES). Examples include liquid air energy storage, pumped heat energy storage and, at least in part, advanced adiabatic compressed air energy storage. Compared to other large-scale storage methods, TES benefits from relatively high energy densities, which should translate into a low cost per MW h of storage capacity and a small installation footprint. TES is also free from the geographic constraints that apply to hydro storage schemes. TES concepts for electricity storage rely on either a heat pump or refrigeration cycle during the charging phase to create a hot or a cold storage space (the thermal stores), or in some cases both. During discharge, the thermal stores are depleted by reversing the cycle such that it acts as a heat engine. The present paper is concerned with a form of TES that has both hot and cold packed-bed thermal stores, and for which the heat pump and heat engine are based on a reciprocating Joule cycle, with argon as the working fluid. A thermodynamic analysis is presented based on traditional cycle calculations coupled with a Schumann-style model of the packed beds. Particular attention is paid to the various loss-generating mechanisms and their effect on roundtrip efficiency and storage density. A parametric study is first presented that examines the sensitivity of results to assumed values of the various loss factors and demonstrates the rather complex influence of the numerous design variables. Results of an optimisation study are then given in the form of trade-off surfaces for roundtrip efficiency, energy density and power density. The optimised designs show a relatively flat efficiency vs. energy density trade-off, so high storage density can be attained with only a modest efficiency penalty. After optimisation, losses due to pressure drop and irreversible heat transfer in the thermal reservoirs are only a few percent, so round-trip efficiency is governed mainly by the efficiency of the compression and expansion processes: overall roundtrip efficiencies approaching those for pumped hydro schemes might be achievable whilst simultaneously attaining energy storage densities of around 200 MJ m –3 , but this is contingent upon attaining compression and expansion efficiencies for the reciprocating devices that have yet to be proven. [ABSTRACT FROM AUTHOR]

Abstract: Many energy conversion and other thermal-fluid systems exhibit unsteady convective heat exchange. In such systems, generic spatiotemporal variations in the flow give rise to variations in the heat flux for a given fluid–solid temperature difference, which can be interpreted as spatiotemporal fluctuations of the instantaneous heat transfer coefficient. These variations can lead to unsteady conjugate heat transfer, in which the exchanged heat flux arises from an interaction between the bulk fluid temperature and the temperature in the solid. Further, the nonlinear coupling between the fluctuating temperature differences and the heat transfer coefficient can lead to an effect we refer to as augmentation, which quantitatively describes the ability of a particular arrangement to have a different time-mean heat flux from the product between the mean heat transfer coefficient and the mean temperature difference across the fluid. It is important to be able to understand and to model in a simple framework the effects of the material properties, the geometry and the character of the heat transfer coefficient on the thermal response of the fluid–solid system, and consequently to predict the overall heat transfer performance of these systems. This paper, which follows on from its predecessor [1], is concerned with the phenomenon of augmentation in simple, one-dimensional, unsteady and conjugate fluid–solid systems. A simple semi-analytical one-dimensional model of heat transfer with a time-varying heat transfer coefficient, which was presented in Mathie and Markides [1], is applied herein to two different paradigm problems. Such a model can be an important tool in the design of improved heat exchangers and thermal insulation, through for example, the novel selection of materials to exploit these augmentation effects. The first flow considered is a thin, wavy fluid film flowing over a heated plate. This film flow exhibits a periodic fluctuation in the heat transfer coefficient, that is linked to the wavy interfacial deformations of free surface of the liquid film. The second flow considered concerns the heat transfer behind a backwards-facing step, which exhibits broadband fluctuations in the heat transfer coefficient due to the flow separation and turbulence behind the step. The model predictions of the augmentation ratio for these problems are also compared to direct measurements from each case. Good agreement is observed with the experimental results for the global heat transfer trends. In both cases the augmentation ratio was negative, reflecting a reduction in time-averaged heat transfer. For the backwards-facing step flow a low magnitude of augmentation ratio was observed, however, the thin film flows exhibited augmentation ratios of as high as 10%. [Copyright &y& Elsevier]

• Nanostructured and plain tubes tested under falling film-boiling conditions. • Heat transfer coefficients measured in refrigerants R-134a and R-245fa. • Critical heat flux due to departure from nucleate boiling identified. • Critical dryout identified. • Falling film heat transfer coefficients compared to those of pool-boiling. Falling film evaporators offer an attractive alternative to flooded evaporators as the lower fluid charge reduces the impact of leaks to the environment and associated safety concerns. A study was conducted of saturated falling film boiling of two refrigerants on one polished, one roughened and three nanostructured copper tubes in order to evaluate the potential of nanostructures in falling film refrigerant evaporators. Tubes were individually tested, placed horizontally within a test chamber and heated by an internal water flow with refrigerant distributed over the outside of the tubes. Wilson plots were used to characterise the internal water heat transfer coefficients (HTCs). A layer-by-layer (LbL) process was used to create the first nanostructured tube by coating the outside of a tube with silica nanoparticles. A chemical bath was used to create copper oxide (CuO) protrusions on the second nanostructured tube. The third tube was coated by following a commercial process referred to as nanoFLUX. R-245fa at a saturation temperature of 20 °C and R-134a at saturation temperatures of 5 °C and 25 °C were used as refrigerants. Tests were conducted over a range of heat fluxes from 20 to 100 kW/m and refrigerant mass film flow rates per unit length from 0 to 0.13 kg/m/s, which corresponds to a film Reynolds number range of 0 to approximately 1500 to 2500, depending on the refrigerant. Heat fluxes were increased further to test whether the critical heat flux (CHF) point due to a departure from nucleate boiling (DNB) could be reached. The CuO and nanoFLUX tubes had the lowest film Reynolds numbers at which critical dryout occurred at heat fluxes near 20 kW/m2, but as the heat fluxes were increased towards 100 kW/m2, critical dryout occurred at the highest film Reynolds numbers of the tubes tested. Furthermore, in some higher heat flux cases, CHF as a result of DNB for the CuO and nanoFLUX tubes was reached before critical dryout occurred, and DNB became the limiting operational factor. The refrigerant condition that had the worst dryout performance in terms of film Reynolds number was R-134a at 25 °C, followed by R-134a at 5 °C and R245fa at 20 °C. Tests across the heat flux range and refrigerant conditions revealed that compared to the polished tube, the roughened tube had HTCs between 60 and 100% higher, the LbL tube had HTCs between 20% lower and 20% higher, the CuO tube had HTCs between 20% lower and 80% higher and the nanoFLUX tube had HTCs between 40 and 200% higher than the polished tube. The falling film enhancement ratios for the plain and nanostructured tubes were found to be of a similar order of magnitude, typically between 1.3 and 0.8. [ABSTRACT FROM AUTHOR]

Highlights • A 3-D FEM and CFD model to simulate a hybrid PV-Thermal collector (PVT) is developed. • Several flat-box designs and materials for the absorber-exchanger are studied. • Flow-channel size or material seem not critical within the range of investigation. • Polymeric flat-box designs are expected to suffer lower strains than sheet and tube units. • Weight and investment cost reductions are achieved with polymeric PVT designs. Abstract In this paper, 26 alternative absorber-exchanger designs for hybrid PV-Thermal (PVT) solar collectors are proposed and compared against a reference-case, commercial sheet-and-tube PVT collector. The collectors involve different geometric design features based on the conventional sheet-and-tube configuration, and also on a newer flat-box structure constructed from alternative polymeric materials with the aim of maintaining or even improving heat transfer and overall (thermal and electrical) performance while achieving reductions in the overall weight and cost of the collectors. The main contributions of this research include: (i) the development and validation of a detailed 3-D computational finite-element model of the proposed PVT collector designs involving multi-physics processes (heat transfer, fluid dynamics and solid mechanics); (ii) results from comparative techno-economic analyses of the proposed PVT designs; and, (iii) further insights from thermal stress and structural deformation analyses of the proposed collectors, which are crucial for ensuring long lifetimes and especially important in the case of polymeric collectors. The results show that, in general, the flat-box designs (characterised by a thin absorber plate) are not sensitive to the flow-channel size or construction material, at least within the range of investigation. A PVT collector featuring a polycarbonate (PC) flat-box design with 3 × 2 mm rectangular channels appears to be a particularly promising alternative to commercial PVT collectors, achieving a slightly improved thermal performance compared to the reference case (with a 4% higher optical efficiency and 15% lower linear heat-loss coefficient), while also lowering the weight (by around 9%) and investment cost (by about 21%) of the collector. The structural analysis shows that the maximum von Mises stress experienced in the absorber-exchanger of the PC flat-box collector is considerably lower than that in the copper sheet-and-tube collector (<13% vs. 64% of the material's yield stress), which is attributed to the larger thermal expansion of the PC absorber-exchanger, that leads to lower stresses. Therefore, the proposed PC flat-box design is not expected to suffer higher strains than the commercially-available PVT collector. [ABSTRACT FROM AUTHOR]

• A novel indirect evaporative cooler is proposed and tested experimentally. • A temperature drop and heat transfer coefficient are achieved up to 20 °C, and 30 W/m2 K. • A novel heat transfer coefficient correlation is proposed with ±5% accuracy. • The proposed correlation can be used for commercial-scale system design. A remarkable surge in cooling demand is observed in the last decades. Currently, the cooling market is dominated by mechanical vapor compression chillers which are energy intensive and use harmful chemical refrigerants. Therefore, the current focus of the current research in cooling is the development of unconventional, sustainable cooling systems. In this regard, indirect evaporative coolers have shown significant potential (particularly under hot-dry climates) with high energy efficiency, low cost, water-based sustainable operation, and benign emissions. However, these systems are in the development stage and have not yet been fully commercialized because of certain design challenges. An innovative indirect evaporative cooler is proposed, fabricated, and experimentally tested in this study. Particularly, the study is focused on the development of heat transfer coefficient correlation for the system for commercial-scale design and expansion. This is because the earlier available correlation is based on simple airflow between parallel plates assumption and does not incorporate the effect of the evaporative potential of the system resulting in under/over-estimation of the heat transfer characteristics. The results showed that the proposed system achieved a temperature drop of 20 °C, a cooling capacity of around 180 W, and an overall heat transfer coefficient of up to 30 W/m2K. Moreover, the study presents an experiment-regression-based heat transfer coefficient correlation that satisfactorily captures the effect of outdoor air temperature and airflow rate ratio which are critical in the design of evaporative coolers. The proposed correlation showed a high (±5%) with experimental data thus making it suitable for the future design of IEC systems over assorted operating scenarios. [ABSTRACT FROM AUTHOR]

Thermo-hydraulic characteristics of supercritical CO 2 (SCO 2) flows in horizontal tubes with half-wall heat-flux conditions are investigated numerically, which is a common practice such as applications in solar parabolic trough collectors, while the heat transfer performance and the underlying mechanisms have not been fully understood. In heated flows, buoyancy acts to inhibit heat transfer when the top half of the tube wall is heated, however, when the bottom half of the tube wall is heated, this inhibition is alleviated, and the synergy between the temperature gradient and velocity fields improves thanks to the secondary flow in the near-wall region at the bottom wall. As a result, the heat transfer coefficient is ∼95% higher (on average) than in the case when the top half of the tube wall is heated. When the bottom half of the tube wall is cooled, buoyancy is expected to enhance heat transfer, while the synergy between the temperature gradient and velocity fields is supressed by the secondary flow in the near-wall region at the bottom of the tube. Conversely, when the top half of the tube wall is cooled, the buoyancy effect inhibits heat transfer, while the synergy between the temperature gradient and velocity fields is improved by the secondary flow in the near-wall region at the top of the tube, which eventually leads to an increase of ∼21% (on average) in the heat transfer coefficient relative to the case when the bottom half of the tube wall is cooled. Finally, the heat transfer discrepancy due to different heat flux conditions revealed in this study are employed in a heat exchanger model, indicating that the thermal performance of this device can be increased by ∼6% through an appropriate arrangement of the hot and cold flows without additional costs. • Thermal-hydraulic features of horizontal flows of supercritical CO 2 are investigated numerically. • Heat transfer coefficient is ∼21% higher when top half wall is cooled than bottom half. • Heat transfer coefficient is ∼95% higher when bottom half wall is heated than top half. • Heat transfer is affected by buoyancy and can be explained using the field synergy principle. • Optimisation of a heat exchanger can be realised by arranging flows without additional consumption. [ABSTRACT FROM AUTHOR]

• Vertical supercritical CO 2 flows at heating conditions are investigated numerically • HTC increases by ∼37% on average in downward flows relative to upward flows • Heat transfer is significantly deteriorated upstream of t pc in upward flows • Multiple wall temperature peaks appear at high heat-flux and low mass-flux conditions • Buoyancy suppresses turbulence and promotes gas-like film formation in upward flows The thermo-hydraulic characteristics of heated supercritical CO 2 (SCO 2) flows are investigated numerically in a vertical pipe from first- and second-law perspectives, and the influence of the flow direction, mass flux and heat flux (both distribution and average value) are evaluated. Two mass flux (254 kg/(m2∙s) and 400 kg/(m2∙s)) and three average heat flux (30 kW/m2, 50 kW/m2 and 70 kW/m2) conditions are simulated at an inlet temperature of 288 K and a pressure of 8.0 MPa (corresponding pseudo-critical temperature of 308 K) in a 4-mm diameter pipe. The simulation results reveal that the heat transfer is enhanced and the irreversibility is reduced in downward flows relative to flows without gravity, whereas the heat transfer deteriorates and the irreversibility is increased in upward flows. Both higher heat fluxes and lower mass fluxes also further hinder heat transfer in the upward flows, and multiple peaks are observed in the axial wall temperature profile. Moreover, it is found that the heat-flux distribution has a significant effect on the heat transfer performance of upward flows; the heat transfer further deteriorates and the irreversibility is further increased when a linearly decreasing heat-flux distribution is applied to the wall, while the heat transfer deterioration is alleviated when a linearly increasing heat-flux distribution is used. An analysis of the heat transfer mechanism indicates that the turbulence production in the core region of the supercritical flow is suppressed, and the accumulation of gas-like fluid in the near-wall region is promoted by the buoyancy effect in upward flows, leading to severe heat transfer deterioration and a sharp increase in the wall temperature, which is similar to the critical heat-flux phenomenon in subcritical boiling. The present study provides insights into the heat transfer characteristics of SCO 2 flows, as well as practical guidance on the design and optimisation of relevant components and equipment. [ABSTRACT FROM AUTHOR]

The thermohydraulic characteristics of supercritical CO 2 flows in a vertical tube at cooling conditions are numerically investigated, and the influence of the heat-flux condition and of the flow direction are evaluated. Constant (i.e., uniform), linearly increasing and linearly decreasing heat-flux conditions are considered as three typical heat-flux distributions over the pipe length. The simulation results show that there exists a maximum heat transfer coefficient at all heat-flux conditions when the fluid bulk temperature is slightly higher than the pseudo-critical temperature, but also that the heat-flux condition has little effect on the peak value of the heat transfer coefficient. From the viewpoint of the second law of thermodynamics, the influence of the heat-flux condition on the local entropy generation can be attributed to the distributed matching between the heat flux and the difference between the wall temperature and the fluid bulk temperature, as a better matching is associated with a higher uniformity of the local entropy generation and reduced overall irreversibilities. Upward and downward flows are considered, along with flows without gravity as a baseline case for comparison purposes, with the field synergy principle employed to explain the different phenomena in these flows. The buoyancy effect laminarises the downward flows and raises the temperature gradient; hence, the heat transfer deteriorates and the irreversibility increases. In the upward flows, the buoyancy effect augments the turbulence and alleviates the variations in temperature and velocity in the core region, consequently reducing the irreversible loss and enhancing heat transfer. The present study provides insights into the mechanisms of supercritical CO 2 heat transfer characteristics as well as practical guidance on the design and optimisation of relevant components. • Thermohydraulic features of supercritical CO 2 vertical flows were studied at cooling conditions. • Local entropy generation relies on the matching between heat flux and temperature difference. • A better uniformly distribution of local irreversibility leads to a less overall irreversibility. • Heat transfer deteriorates and irreversibility increases in downward flows due to buoyancy effect. • Heat transfer strengths and irreversibility decreases in upward flows due to buoyancy effect. [ABSTRACT FROM AUTHOR]

Abstract: The Non-Inertive-Feedback Thermofluidic Engine (NIFTE) is a two-phase thermofluidic oscillator which, by means of persistent periodic thermal-fluid oscillations when placed across a steady temperature difference, is capable of utilising low-grade (i.e., low temperature) heat to induce a fluid motion. Two linearised models of the NIFTE are presented in this paper, both containing a description of the phase-change convective heat transfer that takes place between the working fluid and the heat exchangers. The first model (LTP) imposes a steady linear temperature profile along the surface of the heat exchangers; whereas the second model (DHX) allows the solid heat exchanger blocks to store and release heat dynamically as they interact thermally with the working fluid. In earlier work [Solanki R, Galindo A, Markides CN. Appl Therm Eng; [24]] it was found that these models predict the oscillation (i.e., operation) frequency of an existing NIFTE prototype pump well, but significantly overestimate its reported efficiency. Specifically, the LTP and DHX models predicted exergetic efficiencies 11 and 30 times higher than those observed experimentally, respectively. In the present paper, a dissipative thermal loss parameter that can account for the exergetic losses due to the parasitic, cyclic phase change and heat exchange within the device is included in both models in an effort to make realistic predictions of the exergetic efficiencies. The LTP and DHX models, including and excluding the thermal loss parameter, are compared to experimental data. It is found that the inclusion of the thermal loss parameter increases the predicted oscillation frequencies in the DHX model, but has a negligible effect on the frequencies predicted by the LTP model. A more significant effect is observed with respect to the efficiencies, whereby the inclusion of the thermal loss parameter leads to a greatly improved prediction of the exergetic efficiencies of the prototype NIFTE pump by both the LTP and DHX models, both in trend and approximate magnitude. From the results it is concluded that, on accounting for thermal losses, the DHX model achieves the best predictions of the key performance indicators of the NIFTE, that is, of the oscillation frequency and exergetic efficiency of the device. [Copyright &y& Elsevier]

• Heterogeneous nucleation sites form at the initial stages of freezing • Droplets freeze inwards and along the direction of gravity • Recirculation regions inside the droplet are suppressed as freezing progresses • Temperature and pressure distributions diminish as the freezing front advances • Freezing time is strongly affected by the droplet size This study experimentally and numerically investigates the freezing characteristics and fluid dynamics of millimetre-sized sessile water droplets submerged in silicone oil at sub-zero temperatures under free convection. Individual water droplets were cooled to sub-zero temperatures (260-270 K) via interfacial heat transfer between the two liquid phases, in an approach different to studies in the literature where the cooling is done either from the solid substrate or from a low-temperature gas phase (such as air) surrounding the droplets. Laser-induced fluorescence was employed to perform spatiotemporally-resolved measurements of the phase distribution (from which interface distributions, freezing fronts, and rates were extracted). The particle image velocimetry was used to generate information on the velocity fields inside the liquid droplets. The experimental data are complemented by computational fluid dynamics (CFD) simulations, which showed acceptable qualitative and quantitative agreement with the experimental results. The experimental and simulation results indicated that prior to the initiation of freezing, two counteracting recirculation zones are generated in the central plane of the droplets, one on either side of the centreline, leading to a net upward flow at the edges and a downward flow in the centre due to the natural convection driven by internal temperature gradients. The nucleation sites appear on the external regions of the recirculation structures (which are locations with higher shear). Once freezing starts, the natural circulation patterns are suppressed, and instead, a sole downwards flow dominates, which is the result of the freezing layer suppressing the water phase. CFD results demonstrated a relatively wide temperature and pressure distribution in the water droplet at the beginning of the freezing stage, which gradually diminishes as the freezing process proceeds. The effect of droplet size and oil temperature on the freezing rates were investigated parametrically. The finding showed that when the droplet size doubles, the time required for the droplet to completely freeze extended by almost an order of magnitude. In addition, lower oil temperatures appeared to modify the normalised freezing rate, with sharper exponential growth rates and reduced initial and final stages at higher oil temperatures. [ABSTRACT FROM AUTHOR]