Effectiveness of Linear Spray Cooling in Microgravity Presented by Ben Conrad, John Springmann, Lisa McGill Undergraduates, Engineering Mechanics & Astronautics

Heat dissipation requirements Remove heat fluxes of 100-1000 W/cm2 Applicable to laser diodes, computer processors, etc. Laser Diode Array (Silk et al, 2008)

Heat dissipation requirements Current Solutions Flow boiling Microchannel boiling Jet impingement Spray cooling Spray cooling is the most promising because it achieves high heat transfer coefficients at low flow rates. Spray cooling: promising because high heat transfer coefficients for low flow rates. Problem: few examples of practical applications.

Limited previous microgravity research 14% variation in the critical heat flux from 0 to 1.8 Gs Sone et al. (1996): single spray perpendicular to heated surface (100 mm away) Yoshida, et al. (2001): single spray perpendicular to heated surface (100 mm away) Microgravity significantly effects critical heat flux Golliher, et al. (2005): single spray angled 55⁰ in 2.2 sec. drop tower Significant pooling on the heated surface due largely to surface tension ***NEED TO GET DETAILS ON BAYSINGER/YERKES The applicability of single spray cooling systems are limited by the difficulty in covering a large area (>1 cm2) and ensuring even spray density over the entire area (Shedd 2007) Noted a decrease in Nusselt number with acceleration Yerkes et al. (2004): single spray in micro- and enhanced-gravity.

Spray cooling – linear array Single-spray systems do not cover a large area (> 1 cm2) Regner and Shedd investigated a linear array of sprays directed 45o onto a heated surface Directs fluid flow towards a defined exit to avoid fluid management issues (Shedd, 2007)

Experiment basis & hypothesis Linear spray research showed performance independent of orientation Performance converged for heat fluxes greater than 20 W/cm^2 BUT still +/- 1G pull at all times (Regner, B. M., and Shedd, T. A., 2007)

Experiment basis & hypothesis Predict that with similar spray array, spray cooling will function independent of gravity Performance converged for heat fluxes greater than 20 W/cm^2 Emphasize entire system is in microgravity (pump, etc)

Experiment design Goal: determine variation of heat transfer coefficient h with gravity q’’: heat flux measured from heater power Ts: Temperature of heated surface Tin: Temperature of spray

Differential Pressure Sensor Closed-loop system Pump Flow Meter Pressure Sensor Bladder Filter 3 Axis Accelerometer Spray Box Heat Exchanger Differential Pressure Sensor Therm. Liquid coolant: FC-72 Add FC-72 to slide. Highlight the accelerometer.

Heater design Ohmite TGHG 1 Ω precision current sense resistor Four T-type thermocouples embedded in heater 20.6 mm 25.4 mm 8.0 mm 4.3 mm The material is thin copper, thermocouples embedded near the surface. We assumed thermocouples measure surface temperature.

Spray array design Made from microbore tubing: 3.2 cm Shedd, 2007 Picture of actual array, normal to array surface. Shedd, 2007

Spray array & spray box G Fluid inlet & outlet Top half: spray array Bottom half: heater Picture of actual array, normal to array surface. Z-direction

Microgravity environment 30 microgravity (nominally 0 g) parabolas lasting 20-25s each 1.8 g is experienced between microgravity

Microgravity environment Add accelerometer data in all directions

Procedure: Flow rate Q & heat flux q” Q (L/min): 0.67 2.67 3.81 q” (W/cm2): 24.9 25.8 26.6 Very conservative heat fluxes used due to experimental nature

Epoxy seal failure Epoxy cracked due to fluid pressure in pre-flight testing Spray Array Drain Epoxy Failure 3.2 cm

Epoxy seal failure

Visualization shows fluid behavior Heater Drain ********Grab two other pictures Camera

Complex fluid behavior 760,758,, 2229.6,2216.1,2961.8,3067.8

Flight data: flow rate dominates performance Flowrate determines the heat transfer coefficient low varies by 2.61% med by 1.42% hi by 1.35% Data fall around a constant heat transfer coefficient Strongly suggests independence System experienced acceleration

Δh is consistent with Δg for each flow rate h increases with microgravity Decreases with enhanced gravity Plotting the heat transfer coefficient in blue and the system acceleration in green versus flight time for each flow rate shows some interesting relationships: trends upward in microgravity downward during enhanced gravity

Possible Relationships h vs. jerk Increasing variability with flow rate: Flow rate: 0.67 L/min 2.67 L/min 3.81 L/min change in acceleration. increase in variation Greater turbulence in the spray box?

Shedd model for +/- 1 g Shedd (2007) found a correlation of the form: where the heat transfer coefficient, h, is a function of the average spray droplet flux, Q”, and constants: the fluid’s density, ρ, specific heat, cp, Prandtl number, Pr, an arbitrary constant, C in [m.5s-.5], for a linear spray array, and a constant power, a. Regner and Shedd earth-based orientation tests Shedd noted a relationship between: Constant C accounts for variations between spray arrays Plotting Regner and Shedd and applying the model,

Microgravity results fit trend Q” is believed to be 10-20% high due to the broken epoxy on the spray array follow the same relationship that Shedd described measured flow rates are 10-20% high improve our agreement with Shedd’s 1g model. This simple relationship may allow the performance of a microgravity system to be predicted by 1g earth-based orientation tests.

Future steps Effect of spray characteristics Fluid Inlet Nozzle diameter Nozzle length Nozzle edge type Effect of spray characteristics Spray hole diameter and length Hole entrance and exit design Enhanced surfaces with linear spray cooling? Spray nozzle design The extent to which this causes fluid cavitation. Enhanced surfaces More orderly draining (Kim, J. 2007)

Conclusion Flow rate Q largely determines h 2.61 % standard deviation of h Support for a simple relation between h and Q Ability to predict microgravity performance with a 1g test Unforeseen correspondence with jerk and Q Further microgravity studies are needed Flow rate is the primary driver of the heat transfer coefficient, which had a maximum standard deviation of 2.6%. Heat transfer coefficient and jerk Increase in variability with flow rate Flow loop components Static, microgravity environment Near CHF Linear spray cooling remains a promising method to cool high heat flux and large surface area devices in microgravity.

Thanks The authors are thankful to: the University of Wisconsin ZeroG Team the Multiphase Flow Visualization and Analysis Laboratory the UW Space, Science, and Engineering Center the UW Department of Engineering Physics the Wisconsin Space Grant Consortium NASA Reduced Gravity Student Flight Opportunities Program We would like to thank the University of Wisconsin Zero Gravity team Multiphase Flow Visualization and Analysis Laboratory, UW Space, Science, and Engineering Center, UW Department of Engineering Physics, Wisconsin Space Grant Consortium, and NASA’s Reduced Gravity Student Flight Opportunities Program.

Questions

FEA confirms broken array & uneven cooling FEA confirms the rupture caused uneven temperatures: Top down: Side with rupture, Side with spray cooling less cooling Despite broken spray array, delta_T is small along the spray direction--evidence of uniformity Cross-section: