Enhanced Boiling Heat Transfer by using micropin-finned surfaces for Electronic Cooling JinJia Wei State Key Laboratory of Multiphase Flow in Power Engineering Xi an Jiaotong University
Contents 1. Background and objective 2. Experimental apparatus and conditions 3 Effects of micro-pin-fins and submicron-scale roughness on boiling heat transfer 4 Effects of fin size on boiling heat transfer 5 Enhancement Mechanism for micro-pin-fins 6 Conclusions
Background Trend of sing-chip packaging technology 2006 2010 Performance 3.19 GH 9.54 GH Integration 0.2 billion transistors 1 billion transistors DRAM ½ pitch 80 nm 45 nm Power density 70W/cm 2 85 W/cm 2 Junction temp 85 o C 85 o C Temperature increases by 10 o C Fault probability increases by 50%
Air cooling technology Finned heat pipe radiator demerit 1 toward large volume and weight 2 nose increasing 3 Limited heat dissipation ability
Liquid cooling technology Toshiba Merit 1high efficiency 2 low noise Demerit Poor security and reliability
Liquid cooling technology Indirect liquid cooling Direct liquid cooling To use latent heat of phase change to get high heat dissipation
Direct Liquid cooling with phase change Pool boiling * No pump, less complex, easier to seal Flow boiling Pump, circulation loop
Treated Surfaces (a) Dendritic surface (Oktay et al. 1972) (b) Laser drilled cavity (Chu and Moran, 1977) (c) Micro-Pin-Fin (Mudawar and Anderson, 1989) (d) Hexagonally dimpled and trenched surface (Wright and Gebhart. 1989) (e) Micro-reentrant cavity (Kubo et al. 1999) (f) Diamond treated surface (O Connor et al., 1991)
Problem (1) Severe boiling heat transfer deterioration in high flux region (2) Critical heat flux occurs at T > 85 o C q treated T = 85 o C ONB smooth T sat
Objectives To develop a new surface microstructure for effective boiling heat transfer -----to solve the above problems for electronic cooling application q treated T = 85 o C ONB smooth T sat
Studies Micro-pin-fins and submicron-scale roughness were fabricated on the surface of silicon chip for enhancement of boiling heat transfer of FC-72. (1) Size and height of micro-pin-fins (2) Roughness (3) Subcooling (4) Dissolved gas content (5) Chip orientation
Contents 1. Background and objective 2. Experimental apparatus and conditions 3 Effects of micro-pin-fins and submicron-scale roughness on boiling heat transfer 4 Effects of fin size on boiling heat transfer 5 Enhancement Mechanism for micro-pin-fins 6 Conclusions
Experimental Apparatus 1. Test chip 17 16 10. Thermocouples 2. Glass plate 15 11. DC power supply 3. Vacuum chuck 4. Test vessel 11 D 14 12. Standard resistor 13. Scanner 5. Rubber bag 12 S 13 14. Digital multimeter 6. Water bath 7. Cooling unit kpa 9 8 10 15. Pen recorder 16. Computer 8. Condenser 9. Pressure gauge * *. * o C C 7 1 2 3 4 *.* o C 6 6 17. Power supply controller Schematic diagram of experimental apparatus Experimental Apparatus 5
Test Section 2 50 1 1. Silicon chip 2. Pyrex glass plate 3. Copper lead wire 3 4. Thermocouple 5. O ring 4 6. Vacuum chuck 0.5 10 1.2 2 φ30 φ34 5 6 Details of test section Experimental Apparatus
Micro-Pin-Finned Chips (a) Chip PF10-60 (b) Chip PF20-60 (c) Chip PF30-60 (d) Chip PF50-60 10 10 60 m 3 20 20 60 m 3 30 30 60 m 3 50 50 60 m 3 (d) Chip PF30-120 (e) Chip PF30-200 (f) Chip PF50-200 (g) Chip PF50-270 30 30 120 m 3 30 30 200 m 3 50 50 200 m 3 50 50 270 m 3 SEM images of micro-pin-fins Experimental Apparatus
Rough Surfaces Chip S Chip C RMS Roughness 2.3 nm RMS Roughness 72.7 nm Chip E Chip EPF50-60 RMS Roughness 23.8 nm RMS Roughness 32.0 nm Photographs of heater surface Experimental Apparatus
Experimental Conditions Test chips(12) 8 micro-pin-finned surfaces 2 roughed surfaces 1 roughed micro-pin-finned surface 1 smooth surface Chip orientation Horizontal and vertical Working fluid FC-72(Saturation temp. 56 o C) Liquid subcooling 45K (Liquid temp. 11 o C) 25K (Liquid temp. 31 o C) 3K (Liquid temp. 53 o C) 0K (Liquid temp. 56 o C) Air concentration 3-6 Vol. % (Degassed) 36-40 Vol. % (Gas dissolved) Experimental Apparatus
Contents 1. Background and objective 2. Experimental apparatus and conditions 3 Effects of micro-pin-fins and submicron-scale roughness on boiling heat transfer 4 Effects of fin size on boiling heat transfer 5 Enhancement Mechanism for micro-pin-fins 6 Conclusions
Effects of Micro-Pin-Fins and Submicron-Scale Roughness Effects of Roughness and Micro-Pin-Fin Effect of Liquid Subcooling Effect of Heater Orientation Effect of Dissolved Gas
Boiling Phenomena: Surface Effect Chip S (a) q =2.71W/cm 2 (b) q =5.98W/cm 2 (c) q =11.8 W/cm 2 Chip C (d) q =4.72W/cm 2 (e) q =12.2W/cm 2 (f) q = 30.8 W/cm 2 Chip EPF50-60 (g) q =4.83W/cm 2 (h) q =15.8W/cm 2 (i) q = 39.6 W/cm 2 T sat 4.0K T sat 11.0K T sat 19.0K Effects of Micro-Pin-Fins and Submicron-Scale Roughness
Boiling Curves: Orientation Effect 60 q (W/cm 2 ) 50 40 30 20 T sub =25K Gas dissolved Chip S C EPF50-60 Hor. Ver. 10 0-20 0 20 40 60 T sat (K) T w =85 o C Comparison of boiling curves for vertically and horizontally mounted chip S, C and EPF50-60 Effects of Micro-Pin-Fins and Submicron-Scale Roughness
Boiling Curves q (W/cm 2 ) 60 60 50 50 40 30 20 T sub =0K sub Degassed Chip PF50-60 Chip EPF50-60 Chip S Chip E Chip C Oktay Mudawar and Anderson O'Connor et al. 10 T w =85 o C Free convection 0-40 -20 0 20 40 60 80 T sat (K) Boiling curves; T sub =0 K, degassed Effects of Micro-Pin-Fins and Submicron-Scale Roughness
Boiling Curves q (W/cm 2 ) 60 50 40 30 20 10 0 T sub =25K Degassed Chip PF50-60 Chip EPF50-60 Chip S Chip E Chip C -20 0 20 40 60 T sat (K) T w =85 o C Free convection q (W/cm 2 ) 60 50 40 30 20 10 0 T sub =25K Gas dissolved Chip PF50-60 Chip EPF50-60 Chip S Chip E Chip C -20 0 20 40 60 T sat (K) T w =85 o C (a) Degassed (b) Gas dissolved Boiling curves; T sub =25 K Effects of Micro-Pin-Fins and Submicron-Scale Roughness
Boiling Curves q (W/cm 2 ) 80 60 40 T sub =45K Degassed Chip PF50-60 Chip EPF50-60 Chip S Chip C Mudawar and Anderson ( T sub sub =35K) q (W/cm 2 ) 80 60 40 T sub =45K Gas dissolved Chip PF50-60 Chip EPF50-60 Chip S Chip C O'Connor et al. ( T sub =41K) 20 0-40 -20 0 20 20 40 40 60 60 T sat sat (K) T w o w =85 o C Free Free convection 20 0-40 -20 0 20 40 60 60 T sat (K) T o w =85 o C (a) Degassed (b) Gas dissolved Boiling curves; T sub =45 K Effects of Micro-Pin-Fins and Submicron-Scale Roughness
Boiling Phenomena: Effect of Subcooling (a) q =2.055 W/cm 2 T w = 69.9 C (b) q =5.102 W/cm 2 T w = 73.8 C T sub = 3 K (c) q =30.53 W/cm 2 T w = 81.0 C (d) q =2.724 W/cm 2 T w = 66.1 C (e) q =6.641 W/cm 2 T w = 76.1 C T sub = 25 K Boiling phenomena; Chip PF50-60 (f) q =44.46 W/cm 2 T w = 81.1 C Effects of Micro-Pin-Fins and Submicron-Scale Roughness
Boiling Phenomena: Dissolved Gas Effect Degassed Chip PF30-60 (a) q =3.60W/cm 2 T w = 71.2 C (b) q =16.8 W/cm 2 T w = 73.0 C (c) q =47.5 W/cm 2 T w = 74.0 C Gas-dissolved (d) q =2.84 W/cm 2 T w = 66.6 C (e) q =16.6 W/cm 2 T w = 71.6 C (f) q =47.5 W/cm 2 T w = 75.6 C Effects of Micro-Pin-Fins and Submicron-Scale Roughness
(W/cm 2 2 ) q q Critical Heat Flux 100 Chip S E C PF EPF 90 Mudawar and and Gas dissolved Anderson Degassed 80 Micro-pin-fins O'Connor et et al. al. 70 q 85 o C Porous 60 O'Connor et et al. al. Smooth 50 q 40 85 o C 30 30 20 20 10 10 0 20 40 60 80 0 20 40 60 80 T sub (K) T sub (K) Variation of q with T sub Effects of Micro-Pin-Fins and Submicron-Scale Roughness
Contents 1. Background and objective 2. Experimental apparatus and conditions 3 Effects of micro-pin-fins and submicron-scale roughness on boiling heat transfer 4 Effects of fin size on boiling heat transfer 5 Enhancement Mechanism for micro-pin-fins 6 Conclusions
Size Effects of Micro-Pin-Fin Optimum size of micro-pin-fin? Effects of Cross-sectional Size and Height Effect of Liquid Subcooling Effect of Heater Orientation Effect of Dissolved Gas
Boiling Phenomena:Bubble Growth 0.6 mm 0.8 mm t = 0 ms t = 4 ms t = 8 ms t = 12 ms t = 14 ms t = 0 ms t = 2 ms t = 4 ms t = 6 ms t = 7 ms Sequence of boiling phenomena on vertically mounted chips PF30-200 (q=5.10 W/cm 2, T sat =13.5 K) and S (q=6.02 W/cm 2, T sat =9.2 K; gasdissolved Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Boiling Curves: Orientation Effect q (W/cm 2 ) 60 50 40 30 20 Chip PF50-200 T sub =25K Gas dissolved Hor. Ver. (c) (b) q (W/cm 2 ) 60 50 40 30 20 Chip PF30-200 T sub =25K Gas dissolved Hor. Ver. 10 (a) T w =85 o C 10 T w =85 o C 0-20 0 20 40 60 T sat (K) 0-20 0 20 40 60 T sat (K) (a) Chip PF50-200 (b) Chip PF30-200 Comparison of boiling curves for vertically and horizontally mounted chips Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Fin Cross-Sectional Size (a) Chip PF10-60 (b) Chip PF20-60 (c) Chip PF30-60 (d) Chip PF50-60 10 10 60 m 3 20 20 60 m 3 30 30 60 m 3 50 50 60 m 3 Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Boiling Curves:Fin Cross-Sectional Size q (W/cm 2 ) 40 40 30 20 10 T T sub =0K Degassed sub =0K Degassed Chip PF10-60 Chip Chip PF20-60 PF10 Chip PF20 Chip Chip PF30-60 PF30 Chip Chip PF50-60 PF50 Chip Chip S S O'Connor O'Connor et et al. al. Mudawar Mudawar and and Anderson Anderson Oktay Oktay T w =85 o C T w =85 o C 0-40 -20 0 20 40 60 T sat (K) Free convection Boiling curves; T sub =0K, degassed Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Boiling Curves:Fin Cross-Sectional Size q (W/cm 2 ) 80 60 40 20 T sub =45K Degassed Chip PF10-60 Chip PF20-60 Chip PF30-60 Chip PF50-60 Chip S Mudawar Mudawar and and Anderson. ( T sub =35K) q (W/cm 2 (W/cm ) 80 80 60 60 40 40 20 20 T T sub sub =45K Gas Gas dissolved Chip Chip PF10-60 Chip Chip PF20-60 Chip Chip PF30-60 Chip Chip PF50-60 Chip Chip S S O'Connor et al. et al. ( T ( T sub sub =41K) 0-40 -20 0 20 40 60 T sat (K) T w =85 o C Free convection 0 0-40 -40-20 -20 0 0 20 20 40 40 60 60 T T sat sat (K) (K) TT w =85 o w =85 o CC (a) Degassed (b) Gas dissolved Boiling curves; T sub =45K Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
: Fin Cross-Sectional Size Effect q (W/cm 2 ) 100 80 60 40 T sub (K) 0 3 25 45 Degassed Gas dissolved 20 Chip S PF50-60 PF30-60 PF20-60 PF10-60 A t /A 1.0 2.2 3.0 4.0 7.0 0 1 2 3 4 5 6 7 A t /A Variation of q with A t /A. A t : Total heat transfer area; A: Projected area. Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Maximum Heat Flux: Fin Cross-Sectional Size (W/cm 2 ) max max q 100 Chip S PF10-60 PF20-60 PF30-60 PF50-60 90 80 Mudawar and 70 Anderson Micro-pin-fins 60 O'Connor et al. 50 Porous O'Connor et al. 40 Smooth 30 20 10 0 0 20 40 60 80 T sub (K) q Definition of q max max q q T 85 C w T T w, w, 85 85 C C Variation of q max with T sub Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Effect of Fin Height (a) Chip PF30-60 (b) Chip PF30-120 (c) Chip PF30-200 30 30 60 m 3 30 30 120 m 3 30 30 200 m 3 (d) Chip PF50-60 (e) Chip PF50-200 (f) Chip PF50-270 50 50 60 m 3 50 50 200 m 3 50 50 270 m 3 Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Boiling Curves:Effect of Fin Height q (W/cm 2 ) 40 40 30 30 20 20 10 10 T T sub sub = 0 K Degassed Chip PF30-60 Chip PF30-60 PF30-120 Chip PF30-120 PF30-200 Chip PF30-200 PF50-60 Chip PF50-60 PF50-200 Chip PF50-200 PF50-270 Chip PF50-270 S Chip O'Connor S et al. O'Connor Mudawar et and al. Mudawar Andersonand Anderson Oktay Oktay T w =85 o C Free convection 0-40 -20 0 20 40 60 T sat (K) Boiling curves; T sub =0K, degassed Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Boiling Curves:Effect of Fin Height q (W/cm 2 ) 100 80 60 40 20 0 T sub = 45 K Degassed Chip PF30-60 Chip PF30-120 Chip PF30-200 Chip PF50-60 Chip PF50-200 Chip PF50-270 Chip S Mudawar and Anderson ( T sub = 35 K) T w =85 o C Free convection -40-20 0 20 40 60 T sat (K) (a) Degassed q (W/cm 2 ) 100 80 60 40 20 0 T sub = 45 K Gas dissolved Chip PF30-60 Chip PF30-120 Chip PF30-200 Chip PF50-60 Chip PF50-200 Chip PF50-270 Chip S O'Connor et al. ( T sub = 41 K) T w =85 o C -40-20 0 20 40 60 T sat (K) (b) Gas-dissolved Boiling curves; T sub =45K Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
100 80 : Effect of Fin Height T sub 0 K 25 K 45 K S PF30-h PF50-h T sub = 45K q (W/cm 2 ) 60 40 25K 0K 20 0 PF30 PF30 PF30 PF50 PF50 PF50 Chip S -60-120 -200-60 -200-270 A t /A 1.0 3.0 5.0 7.67 2.2 5.0 6.4 1 2 3 4 5 6 7 8 A t /A Variation of q with A t /A. A t : Total heat transfer area; A: Projected area. Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Maximum Heat Flux: Effect of Fin Height (W/cm (W/cm 2 ) q max max 100 100 80 80 60 60 40 40 20 20 Chip S, S, degassed Chip S, S, gas gas dissolved Chip E, gas dissolved Mudawar and Anderson Micro-pin-fins O'Connor et al. Porous O'Connor et al. Smooth h h ( m) 60 120 200 270 PF30-h PF50-h 00 00 20 20 40 60 80 T sub (K) Variation of q max with T sub Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
1.00 Fin Efficiency 0.98 0.96 0.94 0.92 T sub =45 K Degassed Chip PF50-270 Chip PF50-200 Chip PF30-200 Chip PF30-120 Chip PF50-60 Chip PF30-60 Chip PF20-60 Chip PF10-60 0.90 0 20 40 60 80 q (W/cm 2 ) Variation of with q Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Contents 1. Background and objective 2. Experimental apparatus and conditions 3 Effects of micro-pin-fins and submicron-scale roughness on boiling heat transfer 4 Effects of fin size on boiling heat transfer 5 Enhancement Mechanism for micro-pin-fins 6 Conclusions
Heat Transfer Process in Micro-Pin-Fins FC-72 (a) (b ) Micro-pin-fin Bubble Thin liquid film Evaporation of thin liquid film (c) Pumping action of bubble growing Micro-convection by capillary force Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Heat Transfer Process in Micro-Pin-Fins No natural convection heat transfer enhancement --- All micro-pin-fins are immersed in a boundary layer so that the fin side surfaces will not participate natural convection heat transfer, indicating fin side surfaces are not effective for non-boiling heat transfer enhancement. Natural convection q Micro-pin-fin T = 85 o C Boundary layer smooth Natural convection ONB T sat Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Heat Transfer Process in Micro-Pin-Fins Steep boiling curve ------- Almost simultaneous burn out of many bubbles on the pin fin side surfaces due to nearly the same surface conditions by fabrication with a small increase in wall temperature, resulting in an fast increase of effective boiling heat transfer surface area q T = 85 o C ONB T sat Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Heat Transfer Process in Micro-Pin-Fins No obvious deterioration at high heat flux --- Enough fresh bulk fluid supply through the inter-connect channels formed by micro-pin-fins, preventing the burnout at local area Effects of the Cross-Sectional Size and Height of Micro-Pin-Fin
Contents 1. Background and objective 2. Experimental apparatus and conditions 3 Effects of micro-pin-fins and submicron-scale roughness on boiling heat transfer 4 Effects of fin size on boiling heat transfer 5 Enhancement Mechanism for micro-pin-fins 6 Conclusions
Conclusions (I) Submicron-scale rough surface and micro-pin-finned surface (1) Submicron-scale roughness and micro-pin-fins were effective in enhancing heat transfer in the nucleate boiling region and increasing the. The boiling curve of the micro-pin-finned chip was characterized by a very small increase in wall superheat with increasing heat flux. While the micro-pinfinned chip showed a lower heat transfer performance than the chip with submicron-scale roughness in the low-heat-flux region, it showed a higher heat transfer performance than the latter in the high-heat-flux region. The roughened micro-pin-finned chip showed higher heat transfer performance than a corresponding single rough surface or micro-pin-finned surface. (2) The high boiling heat transfer performance was considered to be relevant to the micro-convection and evaporation of superheated liquid within the confined gaps between fins and the micro-convection caused by the suction of a bubble hovering on the top of micro-pin-fins.
(II) Effect of chip orientation Conclusions (3) For the smooth surface and rough surface, vertical orientation provides better heat transfer in the nucleate boiling regime with a very small decrease of value, whereas for the micro-pin-finned surface, nucleate boiling superheats was independent of orientation but the value was about 20% lower for the vertical orientation than that for the horizontal. (III) Effect of liquid subcooling (4) q increased almost linearly with increasing T sub. Liquid subcooling was very effective in elevating for all the micro-pin-finned chips as compared to the smooth surface and other treated surfaces. (IV) Effect of dissolved gas in FC-72 (5)The gas-dissolved FC-72 showed a marked decrease in the boiling incipience temperature. As a result, the heat transfer performance in the low-heat-flux region was higher than the case of degassed FC-72. However, the heat transfer performance in the high-heat-flux region was close to each other.
Thanks for your attention!