Gravity effects in microgap flow boiling

Abstract

Increasing integration density of electronic components has exacerbated the thermal management challenges facing electronic system developers. The high power, heat flux, and volumetric heat generation of emerging devices are driving the transition from remote cooling, which relies on conduction and spreading, to embedded cooling, which facilitates direct contact between the heat-generating device and coolant flow. Microgap coolers employ the forced flow of dielectric fluids undergoing phase change in a heated channel between devices. While two-phase microcoolers are used routinely in ground-based systems, the lack of acceptable models and correlations for microgravity operation has limited their use for spacecraft thermal management. Previous research has revealed that gravitational acceleration plays a diminishing role as the channel diameter shrinks, but there is considerable variation among the proposed gravity-insensitive channel dimensions and minimal research on rectangular ducts. Reliable criteria for achieving gravity-insensitive flow boiling performance would enable spaceflight systems to exploit this powerful thermal management technique and reduce development time and costs through reliance on ground-based testing. In the present effort, the authors have studied the effect of evaporator orientation on flow boiling performance of HFE7100 in a 218 μm tall by 13.0 mm wide microgap cooler. Similar heat transfer coefficients and critical heat flux were achieved across five evaporator orientations, indicating that the effect of gravity was negligible.

Orientation Effects in Two-Phase Microgap Flow

“…For the present assembly, seven measurements were taken, which produced an average gap height of 1.02 mm, range of 0.99 to 1.04 mm, and uncertainty of ± 0.019 mm. The optical measurement technique was validated in an earlier effort [28] and the present measurements agree with the expected channel height based on individual component measurements, which produced an expected gap height of 1.03 mm.…”

“…Approved for public release; distribution is unlimited. 10 [19] and [20] with data from present study Figure 9: Dominant force maps with boundaries from [19] and [20] with data from [28] The Reynolds, Saad, and Satterlee transitions correctly predict the regime for five of the six data sets for the 1.01 mm tall channel and two of the three data sets for the 0.22 mm tall channel. The Baba et al transitions fail to predict correctly the regime for any of the nine available data sets.…”

“…Collectively, the CHF, HTC, and flow regime results suggest that the effect of gravity is small for mass fluxes of 400 kg/m 2 -s and higher for near-saturated flow boiling of HFE7100 in a 1.01 mm high by 13.0 mm wide by 12.7 mm long channel. In order to compare the results of the present study and those of an earlier study by the present authors [28] with the dominant force regime maps proposed by Reynolds, Saad, and Satterlee [19] and Baba et al [20], the relevant non-dimensional numbers were calculated via two methods: (1) using the hydraulic diameter for all length scales in Eq. 1-4 and (2) using the channel width for surface tension terms and the channel length in the direction of the gravity vector for gravity terms (subsequently referred to as the HW formulation).…”

“…Measurements of the surfaces via confocal microscopy revealed an average roughness of 0.035 μm ± 0.018 μm for the silicon TTC and 0.348 μm ± 0.018 μm for the polycarbonate surface of the microgap. The process for characterizing the thermal losses of the evaporator assembly to ambient are detailed in [28].…”

Orientation Effects in Two-Phase Microgap Flow

“…Flow boiling of HFE7100 in a 1.01 mm by 13.0 mm channel at mass flux 100 kg/m 2 -s and heat flux 53.6 kW/m 2 (flow direction is right to left; see Figure 3 for gravity vector orientation) Flow boiling of HFE7100 in a 1.01 mm by 13.0 mm channel at mass flux 300 kg/m 2 -s and heat flux 171.3 kW/m 2 (flow direction is right to left; seeFigure 3for gravity vector orientation) Dominant force maps with boundaries from[19] and[20] with data from present study Dominant force maps with boundaries from[19] and[20] with data from[28] …”