inverter-scale, passive two -phase cooling … · inverter-scale, passive two -phase cooling system...
TRANSCRIPT
INVERTER-SCALE, PASSIVE TWO-PHASE COOLING SYSTEM FOR AUTOMOTIVE POWER ELECTRONICS Gilbert Moreno, Jana Jeffers, and Sreekant Narumanchi National Renewable Energy Laboratory SAE 2014 Thermal Management Systems Symposium September 22–24, 2014 Denver, CO 14TMSS-0043
NREL/PR-5400-62917
NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.
SAE INTERNATIONAL
I. Project Motivation and Objectives
II. Why Two-Phase Cooling?
III. Approach
IV. Experimental Apparatus Description
V. Thermal Performance Results
VI. Conclusions
VII. Future Work
Outline
2
SAE INTERNATIONAL
Accelerate the adoption of electric-drive vehicles to reduce our nation’s dependence on oil The Problem: High cost of electric-drive vehicles
• Power electronics are a big contributor
to the overall cost Improved thermal management enables cost reductions
Power module
35%
Signal circuits30%
Capacitors14%
Bus bars10%
Heat sink and housing
5%
Current sensor
2%
Assembly4%
On-road 80 kW Manufactured Cost: $1,092
Source: Susan Rogers (U.S. DOE) 2013 AMR Presentation
Motivation
3
SAE INTERNATIONAL
Project Objective
Significantly improve thermal management of automotive power electronics by utilizing the high heat transfer rates of two-phase cooling
• Design and build an inverter-scale passive two-phase cooling system
• Demonstrate that the system dissipates automotive heat loads and provides superior thermal performance
Demonstrate more than 60% reduction in thermal resistance via two-phase immersion cooling. (Credit: Gilbert Moreno, NREL)
4
SAE INTERNATIONAL
Why Two-Phase Cooling?
• Two-phase cooling provides the highest heat transfer rates compared to other methods of cooling
• Near-isothermal system characteristics reduce temperature gradients
q″ (W
/cm
2 )
�T (°C)1 10 1,000100
100
1
0.01
0.1
1,000
10
10,000Experimentally measured heat transfer coefficients of ~20 W/cm2-K using passive two-phase cooling (R-134a)
Adapted from: Mudawar, I., 2001, "Assessment of High-Heat-flux Thermal Management Schemes," Components and Packaging Technologies, IEEE Transactions on, 24(2), pp. 122-141.
5
SAE INTERNATIONAL
When Does Two-Phase Cooling Make Sense?
0.1
1.0
100 1,000 10,000 100,000
Tota
l the
rmal
resi
stan
ce: R
th(K
/W)
Heat transfer coefficient (W/m2-K)
Cold-plate cooled
• Package resistance dominates
• Thermal grease is a major contributor to resistance
Direct-bond-copper substrate cooled
• Lower package resistance, more suited for high performance cooling strategies (i.e., two-phase)
6
SAE INTERNATIONAL
Challenges and Barriers
7
Two-phase cooling is not used to cool automotive power electronics. There are reliability and cost concerns associated with the technology.
Strategy:
• Propose a simple two-phase cooling solution (indirect cooling, passive, low-cost materials, minimize refrigerant requirements)
• Propose solution with pressures equivalent or lower than those used in vehicle air-conditioning systems
• Two-phase cooling technology is used in the vehicle air-conditioning systems. Technology is available to develop reliable refrigerant-based cooling systems.
SAE INTERNATIONAL
Two-Phase Cooling Concept
• Indirect-cooled: Easier to implement, electronics not in contact with refrigerant
• Passive: Increased efficiency through a passive (no pump or compressor) two-phase cooling approach
• Decrease the condenser size using a liquid-cooled condenser. Liquid-cooled condenser may allow use of waste heat for cabin heating (integrated thermal management)
Power electronics modules
Evaporator
Air-cooled condenser
Power electronics modules
Integrated condenser (liquid-cooled) and evaporator
8
SAE INTERNATIONAL
Refrigerant Selection
9
HFO-1234yf: Potential next generation air-conditioning refrigerant
Qualified for automotive use
Environmentally friendly (GWP = 4)
o Lower critical temperature (95 oC)
o Mildly flammable
HFC-245fa:
Higher critical temperature (150 oC)
Non-flammable
Lower pressure as compared with HFO-1234yf
o Higher GWP = 985
o Not used in automotive applications
1
10
100
1000
-40 -20 0 20 40 60 80 100
Pres
sure
(psi
a)
Temperature (°C)
P vs T
R-134a
HFO-1234yf
HFC-245fa
HFE-7000
HFE-7100
14.7 psia
– Evaluating two refrigerants, but the cooling concept can work with other refrigerants
GWP = global warming potential
SAE INTERNATIONAL
Approach
Fundamental Research
Module-Level Research
Inverter-Scale Demonstration
Vapor Liquid
Evaporator
Characterized performance of HFO-1234yf and HFC-245fa
Evaluate techniques to enhance two-phase cooling
Demonstrate two-phase immersion cooling of a power module
Quantified refrigerant volume requirements
Design a system to dissipate the required heat loads
Demonstrate improvements over conventional cooling systems
10
(Credit: Gilbert Moreno, NREL) (Credit: Gilbert Moreno, NREL)
(Credit: Bobby To, NREL)
SAE INTERNATIONAL
• Refrigerant charge: 250 mL (280 g – 330 g)
oComparison: 2010 Toyota Camry air-conditioning system uses 510 g of R-134a
• Maximum operating pressure: 1 MPa (150 psi)
• Automotive fan: 17.8 cm (7 in.) diameter, 38 W, 12 V
25.4 cm (10 in.)
20.5 cm (8.1 in.)
4.5 cm (1.75 in.)
25.4 cm (10 in.)
Ø 6.98 cm (2.75 in.) Tl
T v,P v
Ta, inletTa, outlet
Thtr
Experimental System: Designed and Fabricated a Proof-of-Concept Two-Phase Cooling System
11
SAE INTERNATIONAL
Experimental System: Designed and Fabricated a Proof-of-Concept Two-Phase Cooling System
Interchangeable cold plate design
Delphi discrete power module: Image courtesy of Delphi
Evaporator cross-sectional view
12
25.4 cm (10 in.)
20.5 cm (8.1 in.)
4.5 cm (1.75 in.)
25.4 cm (10 in.)
Ø 6.98 cm (2.75 in.) Tl
T v,P v
Ta, inletTa, outlet
Thtr
SAE INTERNATIONAL
Objectives
• Dissipate inverter-scale (< 3.5 kW for a 55-kW system) heat loads with a small amount of refrigerant without reaching critical heat flux (dry-out)
• Measure the condenser thermal resistance to size the condenser for automotive applications
Experimental System: Designed and Fabricated a Proof-of-Concept Two-Phase Cooling System
Nucleate boiling Critical heat flux (low heat transfer)
R-134a
13
(Credit: Gilbert Moreno, NREL) (Credit: Gilbert Moreno, NREL)
SAE INTERNATIONAL
Test Procedures
• Initial test conducted using six ceramic heaters attached to evaporator using thermal grease
• Six heaters represent six power switches (inverter-scale).
14
25 mm
15 mm
Condenser thermal resistance:
Evaporator thermal resistance:
T air, outlet
T vaporP vapor
T air, inlet
T liquid
T heaters Heater power
Fan power
𝑅𝑅𝑅𝑡𝑡𝑡 =𝑇𝑇𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣 − 𝑇𝑇�𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑡𝑡 𝑣𝑣𝑖𝑖𝑣𝑣
𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝐻𝐻𝐻𝐻𝐻𝐻𝑑𝑑 × Condenser frontal area
𝑅𝑅𝑅𝑡𝑡𝑡 =𝑇𝑇�𝑡𝑖𝑖𝑣𝑣𝑡𝑡𝑖𝑖𝑣𝑣𝑒𝑒 − 𝑇𝑇𝑖𝑖𝑖𝑖𝑙𝑙𝑙𝑙𝑖𝑖𝑙𝑙𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝐻𝐻𝐻𝐻𝐻𝐻𝑑𝑑 × Total heater area
(Credit: Gilbert Moreno, NREL)
SAE INTERNATIONAL
HFC-245fa (250 mL, 330 g)
• Dissipated 3.5 kW of heat in steady-state conditions with a 250- mL charge
• Limited by the heater power capacity
• Increase heat dissipation and decrease thermal resistance using enhanced surfaces
0.00
0.25
0.50
0.75
1.00
0 1,000 2,000 3,000 4,000
R" t
h(c
m2 -
K /
W)
Heat Dissipated (W)
HFC-245fa
HFC-245fa: Tsat = 75°C Psat = 0.7 MPa (106 psia)
Non-coated/plain evaporator
Evaporator Thermal Performance
15
3M microporous coating
(Credit: Bobby To, NREL)
SAE INTERNATIONAL
HFO-1234yf (250 mL, 280 g)
• Prototype limited heat dissipation to 1.3 kW with HFO-1234yf because of higher pressures. Expect we could dissipate more heat by increasing system pressure capacity
• HFO-1234yf’s higher heat transfer coefficients resulted in lower evaporator thermal resistance
0.00
0.25
0.50
0.75
1.00
0 1,000 2,000 3,000 4,000
R" t
h(c
m2 -
K /
W)
Heat Dissipated (W)
HFC-245fa HFO-1234yf
HFO-1234yf: Tsat = 42°C Psat = 1.1 MPa (154 psia)
Evaporator Thermal Performance
16
Non-coated/plain evaporator
SAE INTERNATIONAL
Advanced Evaporator Design
Identified techniques to increase performance and reduce the size of the evaporator
• Designed to cool six Delphi power modules
• Increased evaporation surface area to improve thermal performance
• Fabricated from low-cost materials (aluminum) using low-cost manufacturing techniques
• Reduced refrigerant requirements to 180 mL, (HFO-1234yf = 200 g, HFC-245fa = 240 g)
17.8 cm (7 in.)
7 cm (2.75 in.)
2.54 cm (1 in.)
17
SAE INTERNATIONAL
Advanced Evaporator Performance
• Dissipated 3.5 kW of heat with only 180 mL of HFC-245fa
• Reduced thermal resistance with improved design o Smaller aluminum evaporator
outperforms copper evaporator
• Pressure–temperature fluctuations at higher power levels (≥ 3 kW)
o Microporous coating may eliminate/reduce fluctuations
18
0.00
0.25
0.50
0.75
1.00
0 1,000 2,000 3,000 4,000
R" t
h(c
m2 -
K /
W)
Heat Dissipated (W)
Proof-of-concept (copper - 250 mL)
2nd generation (aluminum - 180 mL)
Non-coated/plain evaporator
aluminum copper
SAE INTERNATIONAL
R”th (mm2- K/W)
% R”th Reduction
Lexus Hybrid (2008) double-side cooled modules 33.2* -
Two-phase: aluminum evaporator (finite element analysis results) 13.9 58%
Two-phase: copper evaporator (finite element analysis results) 11.5 65%
based on die footprint
Indirect two-phase cooling (aluminum evaporator)
* Calculated using one side of the die (12.78 mm × 12.78 mm) * Performance from: Sakai, Yasuyuki, Hiroshi Ishiyama, and Takaji Kikuchi. Power control unit for high power hybrid system. No. 2007-01-0271. SAE Technical Paper, 2007.
Double-side cooled inverter
Indirect two-phase cooling (copper evaporator)
• Reduced thermal resistance by 58% to 65% with the advanced evaporator design compared with state-of-the-art cooling system
• Increased device heat flux by as much as 189%
Image courtesy of Oak Ridge National Lab
2.54 cm (1 in.)
7 cm (2.75 in.)
Length: 17.8 cm (7 in.)
Power devices
2.54 cm (1 in.)
7 cm (2.75 in.)
Length: 17.8 cm (7 in.)
Power devices
Advanced Evaporator Performance: Power Module Cooling
19
SAE INTERNATIONAL
Refrigerant R”th (cm2- K/W) [~250 cfm, 38 W fan]
Plain Tubes Rifled Tubes Reduced R”th with Rifled Tubes
HFC-245fa 9.30 7.58 18%
HFO-1234yf 8.12 6.06 25%
Rifled Tube
• Rifled tubes enhance condensation heat transfer and thus reduce condenser thermal resistance by 18% and 25% for HFC-245fa and HFO-1234yf, respectively
• HFO-1234yf provides lower condenser thermal resistance values compared with HFC-245fa
o 13% plain tubes
o 20% rifled tubes
Condenser Thermal Performance
20
𝑅𝑅𝑅𝑡𝑡𝑡 =𝑇𝑇𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣 − 𝑇𝑇�𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑡𝑡 𝑣𝑣𝑖𝑖𝑣𝑣
𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝐻𝐻𝐻𝐻𝐻𝐻𝑑𝑑× Condenser frontal area
(Credit: Gilbert Moreno, NREL)
SAE INTERNATIONAL
0
500
1,000
1,500
2,000
50 60 70 80 90 100
Con
dens
er F
ront
al A
rea
(cm
2 )
System Liquid/Vapor Temperature (°C)
HFC-245fa HFO-1234yf
• Estimated condenser frontal area/size requirements at various refrigerant temperatures for the operating condition:
o 3.3 kW of total heat, 43oC inlet air
• Higher refrigerant temperatures enable a more compact condenser
• HFO-1234yf’s higher performance allows for a smaller condenser
Condenser depth = 4.4 cm (1.75 in.)
Condenser Size versus System Temperature (Finned-Tube Condenser)
21
SAE INTERNATIONAL
• Estimated the size requirements for a brazed folded-fin condenser matched air-side surface area (assumed air-side was the dominant thermal resistance)
2.54 cm 2.54 cm
~36% more air-side surface area for folded-
fin condenser
Finned-tube Folded-fin
0
500
1,000
1,500
2,000
50 60 70 80 90 100
Con
dens
er F
ront
al A
rea
(cm
2 )
System Liquid/Vapor Temperature (°C)
finned tube
folded-fin
HFC-245fa
HFO-1234yfBarnes and Tuma (2009)
Condenser depth = 4.4 cm (1.75 in.)
Conditions: 3.3 kW of total heat, 43°C inlet air
Condenser Size versus System Temperature (Folded-Fin Condenser)
22
(Credit: Gilbert Moreno, NREL)
SAE INTERNATIONAL
• Demonstrated that a passive, two-phase cooling system can dissipate inverter-scale head loads (3.5 kW) with only 180 mL of refrigerant
• Improved the evaporator design to increase its performance and reduce its size
o Evaporator features enabled the aluminum evaporator to outperform the copper evaporator
o Increase power module heat flux by as much as 189% as compared with state-of-the-art automotive systems
• Measured the condenser thermal resistance and conducted analysis to size the condenser at various operating temperatures
Conclusions
23
SAE INTERNATIONAL
Future Work
• Evaluate the effect of cooling system inclination/orientation on thermal performance
• Experimentally quantify key system metrics (thermal resistance, coefficient of performance, volume, weight) and compare against conventional cooling systems
• Develop industry partnerships to demonstrate a two-phase cooled inverter system
24
For more information, contact:
Principal Investigator Gilbert Moreno [email protected] Phone: (303) 275-4450
APEEM Task Leader:
Sreekant Narumanchi [email protected] Phone: (303) 275-4062
Acknowledgments:
Susan Rogers and Steven Boyd, U.S. Department of Energy Team Members:
Jana Jeffers (NREL) Charlie King (NREL) Sreekant Narumanchi (NREL)