Two-phase cooling of a square simulated electronic device surface of 21.3 mm side was successfully carried out without the need for a pump. This smart, passive cooling system incorporates a self-enhancing and self-sustaining mechanism, wherein the system inherently enhances its cooling capacity by increasing the velocity of the two-phase mixture along the boiling surface when an increase in heat flux is sensed. Other practical attributes of this pumpless loop are small liquid inventory requirements and absence of the incipient boiling temperature drop. It is shown small surface tension and contact angle render dielectric coolants such as FC-72 ideally suited for flow in narrow gaps. These unique properties are responsible for very small bubble size, precluding any appreciable blockage of the replenishment liquid flow even in narrow gaps. Critical heat flux (CHF) was found to generally increase with decreasing boiler gap. CHF for flat, micro-channel (with 0.2 mm rectangular fins) and mini-channel (with 1.98 mm rectangular fins) surfaces was 4.5, 5.9, and 5.7 times greater than for pool boiling from a flat surface for corresponding gaps. A pressure drop model was formulated to predict coolant mass flow rate, boiling surface inlet and exit velocities, and pressure drop components throughout the loop. The model predictions illustrate the pumpless loop's selfsustaining and self-enhancing attributes, and relate CHF trends to those of the two-phase mixture acceleration along the boiling surface. Index Terms-Boiling, critical heat flux, high heat flux, microchannel, passive cooling, phase change, pumpless loop, pressure drop. NOMENCLATURE Fin width. Flow area associated with boiler gap adjacent to boiling surface. Planform area of boiling surface, .

A compact cooling system is examined which capitalizes upon fluid density differences between two vertical, parallel, interconnected tubes to achieve a pumpless cooling loop. A heat-dissipating device is incorporated into a boiler at the bottom of the hot tube. The large density differences between the two tubes produces a substantial nonequilibrium in hydrostatic pressure, drawing liquid downwards through the cold tube as a two-phase mixture is released upwards in the hot tube. Cooling with this pumpless loop is fundamentally different from, and far superior to, pool boiling thermosyphons because of the former’s ability to separate the path of replenishment liquid from that of the released vapor. Experiments were performed to explore the effects of boiler gap (separation distance between the boiling surface and opposite insulating wall) on cooling performance and critical heat flux (CHF) for water and FC-72. The gap, which is the primary measure of boiler miniaturization, was varied from 0.051 to 21.46 mm. For large gaps, CHF showed insignificant dependence on the gap for both fluids. However, small gaps produced CHF variations that were both drastic and which followed opposite trends for the two fluids. Decreasing the gap below 3.56 mm produced a substantial rise in CHF for FC-72. For water, CHF was fairly insensitive down to 0.51 mm, below which it began to decrease sharply. These trends are shown to be closely related to the small surface tension and contact angle of FC-72 producing very small bubbles which can easily pass through narrow gaps in FC-72, while much larger bubbles in water obstruct liquid replenishment in narrow gaps. A numerical model is constructed to determine how the gap influences the various components of pressure drop, velocities, coolant flow rate, and hence system response to heat input.

The effect of the primary knock-on atom (PKA) spectrum in radiation damage and the subsequent defect structure formation and their impact in deuterium (D) trapping has been investigated using computer simulations and surrogate ion irradiation experiments. The neutron spectrum for an 'ITER-like' divertor shape and parameters has been generated using ATTILA and SPECTER codes to identify the relevant PKA energies. It has been observed that 10 MeV boron (B) produces a PKA spectrum similar to that obtained from a reactor-like neutron spectrum. Experiments have been carried out with ions of gold (Au), B, helium (He) and D with energies ranging from 0.1 MeV-80 MeV for a fluence range of 1.3 × 10 18 ions m −2 -5 × 10 21 ions m −2 , and distinctly different PKA spectra have been produced. While 80 MeV Au ions produced dense and small clusters of interstitial defects (<10 nm), B produced large dislocation loops up to 60 nm in size. At room temperature, the imprint of the cascade is well captured by the vacancies due to their low mobility, and the vacancy defects observed in Au and B irradiation showed significant differences. Molecular dynamics simulations show that at PKA energies exceeding 150 keV, the fragmentation of the cascades takes place, which tends to limit the size of individual defects in the case of 80 MeV Au irradiation. A mechanism based on the competitive capture of mobile interstitials has been proposed to explain the observed large dislocation loops as well as dislocation lines in different irradiation experiments.

Surface-shifted deuterium profiles are re-examined in deuterium-ion irradiation experiments by using a combined experimental and modelling approach. Recrystallized tungsten foil samples were irradiated with energetic deuterium ions and the defect and deuterium depth profiles were studied using positron annihilation spectroscopy and secondary ion mass spectroscopy. We report direct experimental evidence of trapping of deuterium at the vacancies created by the deuterium ions themselves during the implantation by using positron annihilation studies. The deuterium profile is simulated using a Monte-Carlo diffusion model by taking into account the defect-aided diffusion of deuterium due to the local strain field created by the vacancies. The simulations also elucidate the role of the anisotropy in the diffusion and trapping of deuterium in ion-implantation experiments in metals.

No abstract