Search

Oscillating heat pipes (OHPs) represent a promising passive mechanism for the removal or spreading of heat. While simple to construct, the fluid and thermodynamics of these devices are still poorly understood. There is debate over whether the primary heat transfer mechanism is due to sensible heating of the liquid phase or due to latent heat transfer through phase change. To answer this question, an experimental apparatus was constructed to provide time- and space-resolved temperature and heat transfer data across the face of an operating OHP with HFE-7000 as the working fluid. This experiment utilized temperature sensitive paint (TSP) alongside visual recording of the fluid motion in order to determine the relative latent and sensible contribution to the overall heat transfer. The OHP was tested with input powers ranging from 2.6 W to 10.1 W. It found that latent heat transfer was the dominant heat transfer mechanism, accounting for between 65% and 83% of the total heat transferred in all cases.

Techniques based on temperature-sensitive paints (TSP) to measure time-resolved temperature and heat transfer distributions at the interface between a wall and fluid during pool and flow boiling are described. The paints are excited using ultraviolet (UV) light emitting diodes (LEDs), and changes in fluorescence intensity are used to infer local temperature differences across a thin insulator from which heat flux distribution is obtained. Advantages over infrared (IR) thermometry include the ability to use substrates that are opaque to IR (e.g., glass, plexiglass and plastic films), use of low-cost optical cameras, no self-emission from substrates to complicate data interpretation, high speed, and high spatial resolution. TSP-based methods to measure wall heat transfer distributions are validated and then demonstrated for pool and flow boiling.

Direct spraying of dielectric liquids has been shown to be an effective method of cooling high-power electronics. Recent studies have illustrated that even higher heat transfer can be obtained by adding extended structures, particularly straight fins, to the heated surface. In the current work, spray cooling of high-aspect-ratio open microchannels was explored, which substantially increases the total surface area and allows more residence time for the incoming liquid to be heated by the wall. Five such heat sinks were constructed, and their thermal performance was investigated. These heat sinks featured a projected area of 1.41 × 1.41 cm 2 , channel width of 360 μm, a fin width of 500 μm, and fin lengths of 0.25 mm, 0.50 mm, 1.0 mm, 3.0 mm, and 5.0 mm. The five enhanced surfaces and a flat surface with the same projected area were sprayed with a full cone nozzle using PF-5060 at 30°C and nozzle pressure differences from 1.36-4.08 atm (69-121 ml/min). In all cases, the enhanced surfaces improved thermal performance compared to the flat surface. Longer fins were found to outperform shorter ones in the single-phase regime. Adding fins also resulted in the onset of two-phase effects (and higher-heat transfer) at lower wall temperatures than the flat surface. The two-phase regime was characterized by a balance between added area, changing flow flux, flow channeling, and added conduction resistance. Spray efficiency calculations indicated that a much larger percentage of the liquid sprayed onto the enhanced surface evaporated than with the flat surface. Fin lengths between 1 mm and 3 mm appeared to be optimum for heat fluxes as high as 124 W/cm 2 (based on projected area) and the range of con ditions studied.

Detailed local measurements of wall heat flux during saturated pool boiling of FC-72 on a small heated area without sidewalls were made using an array of 96 temperature-controlled heaters. Data were obtained in the nucleate boiling, critical heat flux, and transition boiling regimes. The space and time resolved data were used to conditionally sample the heat flux according to whether or not boiling occurred on the surface, enabling the separation of the heat flux due to boiling from that due to natural convection or vapor contact. The heat transfer from the edge heaters was observed to be much higher than that for the inner heaters above the critical temperature. The heat transfer during liquid contact in transition boiling was constant for a given wall superheat for the inner heaters, and was observed to decrease with increasing wall superheat.

Experimental work was undertaken to investigate the process by which pool-boiling critical heat flux (CHF) occurs using an IR camera to measure the local temperature and heat transfer coefficients on a heated silicon surface. The wetted area fraction (WF), the contact line length density (CLD), the frequency between dryout events, the lifetime of the dry patches, the speed of the advancing and receding contact lines, the dry patch size distribution on the surface, and the heat transfer from the liquid-covered areas were measured throughout the boiling curve. Quantitative analysis of this data at high heat flux and transition through CHF revealed that the boiling curve can simply be obtained by weighting the heat flux from the liquid-covered areas by WF. CHF mechanisms proposed in the literature were evaluated against the observations. [DOI: 10.1115/1.4025697]

The relatively poor understanding of gravity effects on pool boiling heat transfer can be attributed to the lack of long duration high-quality microgravity data, g-jitter associated with ground-based low gravity facilities, little data at intermediate gravity levels, and a poor understanding of the effect of important parameters even at earth gravity conditions. The results of over 200 pool boiling experiments with n-perfluorohexane as the test fluid performed aboard the International Space Station (ISS) are presented in this paper. A flat, transparent, constant temperature microheater array was used to perform experiments over a wide range of temperatures (55 °C

Spray cooling heat transfer measurements of PF-5060 on smooth flat surfaces were obtained using a microheater array operated at constant temperatures using two nozzles (hollow cone and full cone), three nozzle-to-heater standoff distances (3, 5, and 7 mm), four nozzle operating pressures (207 kPa, 345 kPa, 483 kPa, and 689 kPa), and three subcooling levels (11 °C, 21 °C, 31 °C). A separate test setup was used to measure the local normal pressures produced by the sprays. The critical heat flux was found to depend primarily on the local normal pressure and the liquid subcooling. Furthermore, the temperature at which CHF occurred was within a narrow temperature band (about ±5 °C) for smooth flat surfaces over a wide range of spray conditions. The single-phase correlation previously proposed by the authors and the CHF correlation presented in this work were then combined to predict local spray cooling curve within ±25% of the measured values in the spray impingement zone.

One of the main challenges of spray cooling technology is the prediction of local and average heat transfer coefficients on the heater surface. It is hypothesized that the local heat transfer coefficient can be predicted from the local normal pressure produced by the spray. In this study, hollow cone, full cone, and flat fan sprays, operated at three standoff distances, five spray pressures, and two nozzle orientations, were used to identify the relation between the impingement pressure and the heat transfer coefficient in the single-phase regime. PF-5060, PAO-2, and PSF-3 were used as test fluids, resulting in Prandtl number variation between 12 and 76. A microheater array operated at constant temperature was used to measure the local heat flux. A separate test rig was used to make impingement pressure measurements for the same geometry and spray pressure. The heat flux data were then compared with the corresponding impingement pressure data to develop a pressure-based correlation for spray cooling heat transfer. The maximum deviation between the experimental data and prediction was within ±25%.

A pool boiling regime map demarcating the boundary between the surface tension and buoyancy dominated boiling regimes is developed based on heater size and gravity. For large heaters and/or high gravity conditions, boiling is dominated by buoyancy, and the ebullition cycle dominates the contribution to heat transfer. As the gravity level and/or heater size is decreased, surface tension forces become increasingly dominant, and a decrease in heat transfer is observed. The ratio of the heater size Lh (length of a side for a square heater) to the capillary length Lc is found to be a suitable parameter to define the transition criterion between these regimes. Based on the data obtained using FC-72 and pentane, the threshold value of Lh/Lc above which pool boiling is buoyancy dominated was found to be about 2.1. This transition criterion was found to be the same for gravity levels between ∼0g–1.7g and liquid subcoolings between 6.6°C and 26.6°C.

The effect of low gravity on pool boiling heat flux has been studied by researchers, but pool boiling heat flux data over a continuous range of gravity levels (0g–1.8g) was unavailable until recently. The current work uses the results of a variable gravity subcooled pool boiling experiment to develop a gravity scaling parameter for prediction purposes. The heat flux prediction at various gravity levels was found to be in good agreement with the measured heat flux data. The scaling parameter can be used as a tool to predict boiling heat flux at any gravity level if the data under similar experimental conditions are available at any other gravity level. The scaling parameter has been demonstrated to be valid for pool boiling of n-perfluorohexane in heater size independent boiling regime (Lh /Lc >2.8).Copyright © 2009 by ASME

An experimental investigation of the transition process on flat-plate and concave curved-wall boundary layers for various free-stream turbulence levels was performed. Results show that for transition of a flat-plate, the two forms of boundary layer behavior, identified as laminar-like and turbulent-like, cannot be thought of as separate Blasius and fully-turbulent profiles, respectively. Thus, simple transition models in which the desired quantity is assumed to be an average, weighted on intermittency, of the theoretical laminar and fully turbulent values is not expected to be successful. Deviation of the flow identified as laminar-like from theoretical laminar behavior is shown to be due to recovery after the passage of a turbulent spot, while deviation of the flow identified as turbulent-like from the full-turbulent values is thought to be due to incomplete establishment of the fully-turbulent power spectral distribution. Turbulent Prandtl numbers for the transitional flow, computed from measured shear stress, turbulent heat flux and mean velocity and temperature profiles, were less than unity. For the curved-wall case with low free-stream turbulence intensity, the existence of Gortler vortices on the concave wall within both laminar and turbulent flows was established using liquid crystal visualization and spanwise velocity and temperature traverses. Transition was found to occur via a vortex breakdown mode. The vortex wavelength was quite irregular in both the laminar and turbulent flows, but the vortices were stable in time and space. The upwash was found to be more unstable, with higher levels of u' and u'v', and lower skin friction coefficients and shape factors. Turbulent Prandtl numbers, measured using a triple-wire probe, were found to be near unity for all post-transitional profiles, indicating no gross violation of Reynolds analogy. No evidence of streamwise vortices was seen in the high turbulence intensity case.