You are on page 1of 158

AIRCRAFT THERMAL MANAGEMENT USING LOOP HEAT PIPES

A thesis submitted in partial fulfillment
of the requirements for the degree of
Master of Science in Engineering





By





ANDREW JAMES FLEMING
B.S., Wright State University, 2004











2009
Wright State University
WRIGHT STATE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

March 20, 2009

I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY
SUPERVISION BY Andrew James Fleming ENTITLED Aircraft Thermal Management
Using Loop Heat Pipes BE ACCEPTED IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF Master of Science in Engineering.


____________________________________
Scott K. Thomas, Ph.D.
Thesis Director


____________________________________
George P.G. Huang, P.E., Ph.D.
Department Chair

Committee on
Final Examination


____________________________________
Scott K. Thomas, Ph.D.


____________________________________
Kirk L. Yerkes, Ph.D.


____________________________________
J. Mitch Wolff, Ph.D.


____________________________________
James A. Menart, Ph.D.


____________________________________
Joseph F. Thomas, Jr., Ph.D.
Dean, School of Graduate Studies
iii
ABSTRACT
Fleming, Andrew James. M.S.Egr., Department of Mechanical and Materials
Engineering, Wright State University, 2009. Aircraft Thermal Management using Loop
Heat Pipes.



The objective of this thesis was to determine the feasibility of using loop heat
pipes to dissipate waste heat from power electronics to the skin of a fighter aircraft and
examine the performance characteristics of a titanium-water loop heat pipe under
stationary and elevated acceleration fields. In the past, it has been found that the
boundary condition at the condenser can be a controlling factor in the overall
performance of this type of thermal management scheme. Therefore, the heat transfer
removed from the aircraft skin has been determined by modeling the wing as a flat plate
at zero-incidence as a function of the following parameters: airspeed: 0.8 ≤ Ma

≤ 1.4;
altitude: 0 ≤ H ≤ 22 km; wall temperature: 105 ≤ T
w
≤ 135°C. In addition, the effects of
the variable properties of air have been taken into account. Heat transfer due to thermal
radiation has been neglected in this analysis due to the low skin temperatures and high
airspeeds up to Ma

= 1.4. It was observed that flight speed and altitude have a
significant effect on the heat transfer abilities from the skin to ambient, with heat
rejection becoming more difficult with increasing Mach number or decreasing altitude.
An experiment has been developed to examine operating characteristics of a
titanium-water loop heat pipe (LHP) under stationary and elevated acceleration fields.
The LHP was mounted on a 2.44 m diameter centrifuge table on edge with heat applied
to the evaporator via a mica heater and heat rejected using a high-temperature
polyalphaolefin coolant loop. The LHP was tested under the following parametric
ranges: heat load at the evaporator: 100 ≤ Q
in
≤ 600 W; heat load at the compensation
chamber: 0 ≤ Q
cc
≤ 50 W; radial acceleration: 0 ≤ a
r
≤ 10 g. For stationary operation (a
z

= 1.0 g, a
r
= 0 g), the LHP evaporative heat transfer coefficient decreased monotonically,
iv
thermal resistance decreased to a minimum then increased, and wall superheat increased
monotonically. Heat input to the compensation chamber was found to increase the
evaporative heat transfer coefficient and decrease thermal resistance for Q
in
= 500 W.
Flow reversal in the LHP was found for some cases, which was likely due to vapor
bubble formation in the primary wick. Operating the LHP in an elevated acceleration
environment (a
z
= 1.0 g, a
r
> 0 g) revealed dry-out conditions from Q
in
= 100 to 400 W
and varying accelerations and the ability for the LHP to reprime after an acceleration
event that induced dry-out. Evaporative heat transfer coefficient and thermal resistance
was found not to be significantly dependent on radial acceleration. However, wall
superheat was found to increase slightly with radial acceleration.
v
TABLE OF CONTENTS
1.  Convective Heat Transfer from High-Speed Aircraft Skin .................................... 1 
1.1. Abstract ............................................................................................................ 1 
1.2. Introduction ...................................................................................................... 1 
1.3. Mathematical Model ........................................................................................ 3 
1.4. Results and Discussion .................................................................................... 5 
1.5. Conclusions ...................................................................................................... 6 
2.  Titanium-Water Loop Heat Pipe Characteristics Under Stationary and
Elevated Acceleration Fields .................................................................................... 14 
2.1. Abstract .......................................................................................................... 14 
2.2. Introduction .................................................................................................... 14 
2.3. Experimental Setup ........................................................................................ 19 
2.4. Results and Discussion .................................................................................. 27 
2.5. Conclusions .................................................................................................... 39 
2.6. Future Work ................................................................................................... 40 
References ........................................................................................................................ 74 
Appendix A. Operating Procedures .............................................................................. 77 
A.1. Standard Operating Procedure ...................................................................... 77 
A.2. Test Procedures ............................................................................................. 79 
Appendix B. Uncertainty Analysis ................................................................................ 89 
Appendix C. Calibration of Thermocouples and Flow Meter .................................... 92 
C.1. Thermocouple Calibration ............................................................................. 92 
vi
C.2. Flow Meter Calibration ................................................................................. 95 
Appendix D. Loop Heat Pipe Mounting ..................................................................... 114 
Appendix E. Brayco Micronic 889 Technical Data .................................................... 119 
Appendix F. Centrifuge Table Upgrades .................................................................... 121 
Appendix G. LabVIEW Programs .............................................................................. 130 
Appendix H. Centrifuge Wiring Tables ...................................................................... 136 
Appendix I. Sample Calculations ................................................................................ 139 
vii
LIST OF FIGURES
Figure 1.1. Comparison of atmospheric properties versus altitude: (a)
Temperature; (b) Density (DOD, 1997; Anderson, 2000). ........................................... 7 
Figure 1.2. Adiabatic wall temperature versus altitude for various Mach numbers
(1% hot day). ................................................................................................................. 8 
Figure 1.3. Temperature difference ) (
aw ∞
−T T versus altitude for various Mach
numbers (1% hot day). .................................................................................................. 8 
Figure 1.4. Temperature difference ) (
aw w
T T − versus altitude for various Mach
numbers (T
w
= 135ºC, 1% hot day). .............................................................................. 9 
Figure 1.5. Maximum Mach number before heat is transferred from the air to the
skin versus altitude for various wall temperatures (1% hot day). ................................. 9 
Figure 1.6. Average convective heat transfer coefficient versus altitude for
various Mach numbers (T
w
= 135ºC, L = 1.0 m, 1% hot day). ................................... 10 
Figure 1.7. Average heat flux dissipated over the plate versus altitude for various
Mach numbers (T
w
= 135ºC, L = 1.0 m, 1% hot day). ................................................ 10 
Figure 1.8. Local heat flux dissipated over the plate versus plate length for
various Mach numbers (T
w
= 135ºC, 1% hot day): (a) H = 0 km; (b) H = 10
km; (c) H = 20 km. ...................................................................................................... 11 
Figure 1.9. Average heat flux dissipated over the plate versus altitude for various
atmospheric conditions (T
w
= 135ºC, L = 1.0 m, Ma

= 0.98) (DOD, 1997;
Anderson, 2000). ......................................................................................................... 12 
Figure 1.10. Average heat flux dissipation versus altitude for various wall
temperatures (Ma

= 0.98, L = 1.0 m, 1% hot day). ................................................... 12 
Figure 2.1. Loop heat pipe operation. Adapted and reprinted with permission
from AIAA (Hoang and Ku, 2003). ............................................................................ 42 
Figure 2.2. Evaporator schematic: (a) Side view; (b) Cross-sectional view.
Adapted and reprinted with permission from AIAA (Hoang and Ku, 2003). ............ 43 
Figure 2.3. Schematic of Centrifuge Table Test Bed. ...................................................... 44 
Figure 2.4. Titanium-water loop heat pipe test article as delivered. ................................ 45 
Figure 2.5. Thermocouple locations on the LHP: (a) Locations of thermocouples
TC04 through TC15 across the LHP; (b) Locations of TC04 through TC07
within the evaporator. ................................................................................................. 46 
viii
Figure 2.6. Mounting of loop heat pipe to centrifuge table, front and top views:
(a) Evaporator and compensation chamber: (b) Transport lines; (c) Condenser
with cold plate; (d) Complete loop heat pipe. ............................................................. 47 
Figure 2.7. High temperature fluid loop: (a) Schematic; (b) Reservoir, pump,
filter, flowmeter, TC03, and liquid/liquid heat exchanger; (c) Cold plate,
TC00, and TC01. ......................................................................................................... 48 
Figure 2.8. Use of a cold-start test to determine when steady state occurred for
the stationary LHP (Q
in
= 600 W, Q
cc
= 0 W, a
r
= 0 g,
cp
m& = 0.0077 kg/s,
cp
T
= 67.7°C, T
amb
= 38.1°C): (a) Transient temperature traces; (b) Transient rate
of change of temperatures; (c) Transient thermal resistance and evaporative
heat transfer coefficient. .............................................................................................. 49 
Figure 2.9. Transient startup of the stationary LHP (Q
in
= 600 W, Q
cc
= 0 W, a
r
=
0 g,
cp
m&

= 0.0077 kg/s,
cp
T = 67.7°C, T
amb
= 38.1°C): (a) Initial startup; (b)
Complete startup until steady state. ............................................................................ 50 
Figure 2.10. Transient temperature profiles in the condenser and bayonet tube of
the stationary LHP (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T ≤
71.6°C, 31.7 ≤ T
amb
≤ 38.1°C): (a) Q
in
= 100 W; (b) Q
in
= 200 W; (c) Q
in
=
300 W; (d) Q
in
= 400 W; (e) Q
in
= 500 W; (f) Q
in
= 600 W. ...................................... 51 
Figure 2.11. Transient temperature profiles of the stationary LHP for Q
in
= 200
W (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s,
cp
T = 46.1°C, T
amb
= 31.7°C): (a)
Transient temperature profiles; (b) 2φ-1φ point oscillation in the condenser. ........... 52 
Figure 2.12. Steady state temperature distribution versus transported heat for the
stationary LHP (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T ≤ 71.6°C,
31.7 ≤ T
amb
≤ 38.1°C): (a) Evaporator section; (b) Condenser section. ..................... 53 
Figure 2.13. Steady state performance characteristics of the stationary LHP
versus transported heat (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T ≤
67.7°C, 27.6 ≤ T
amb
≤ 38.7°C): (a) Evaporative heat transfer coefficient; (b)
Thermal resistance; (c) Wall superheat ....................................................................... 54 
Figure 2.14. Steady state performance characteristics of the stationary LHP
versus compensation chamber heat input (Q
in
= 500 W, a
r
= 0 g,
cp
m& = 0.0077
kg/s, 63.4 ≤
cp
T ≤ 64.8°C, 36.1 ≤ T
amb
≤ 38.1°C): (a) Evaporator
temperatures; (b) Condenser temperatures; (c) Evaporative heat transfer
coefficient and thermal resistance. .............................................................................. 55 
Figure 2.15. Transient temperature profiles in the condenser and bayonet tube of
the stationary LHP for Q
cc
= 25 to 50 W (Q
in
= 500 W, a
r
= 0 g,
cp
m&

= 0.0077
kg/s, 63.4 ≤
cp
T ≤ 64.8°C, 36.1 ≤ T
amb
≤ 38.1°C): (a) Q
cc
= 25 W; (b) Q
cc
= 30
W; (c) Q
cc
= 35 W; (d) Q
cc
= 40 W; (e) Q
cc
= 45 W; (f) Q
cc
= 50 W. ......................... 56 
ix
Figure 2.16. Transient temperature traces of the LHP at elevated acceleration (Q
in

= 600 W, Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 55.2 ≤
cp
T ≤ 59.7°C, 27.9 ≤ T
amb

30.1°C): (a) a
r
= 0.1 g startup phase; (b) Transition to and steady state at a
r
=
10.0 g; (c) Transient rate of change of temperatures. ................................................. 57 
Figure 2.17. Effect of resultant acceleration vector direction on fluid distribution
within the LHP: (a) Resultant acceleration vector orientation versus radial
acceleration; (b) Liquid pooling in the evaporator, compensation chamber, and
condenser under elevated acceleration (to scale, top view); (c) Liquid pooling
in the condenser bends. ............................................................................................... 58 
Figure 2.18. Steady state performance characteristics of the LHP versus
transported heat at stationary and elevated acceleration (Q
cc
= 0 W,
cp
m& =
0.0077 kg/s, 37.2 ≤
cp
T ≤ 67.7°C, 25.1 ≤ T
amb
≤ 38.7°C): (a) Evaporative heat
transfer coefficient; (b) Thermal resistance; (c) Wall superheat. ............................... 59 
Figure 2.19. Transient temperature traces of the LHP at elevated acceleration
showing dry-out behavior (Q
in
= 400 W, Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 37.2 ≤
cp
T ≤ 59.7°C, T
amb
= 28.0°C): (a) T
e,max
= 150°C; (b) T
e,max
= 175°C; (c)
T
e,max
= 200°C. ............................................................................................................ 60 
Figure 2.20. Quasi-steady state temperature traces of the LHP and cold plate at
elevated acceleration for Q
in
= 200 W (Q
cc
= 0 W,
cp
m& = 0.0077 kg/s,
cp
T =
41.9°C, T
amb
= 26.4°C): (a) Transient temperature trace at a
r
= 0.1 g and t =
13834 s; (b) Transient temperature trace at a
r
= 4.0 g and t = 31240 s. ..................... 61 
Figure 2.21. Steady state performance map of the LHP relating radial
acceleration and heat transported (Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 37.2 ≤
cp
T ≤
59.7°C, 25.1 ≤ T
amb
≤ 30.2°C). ................................................................................... 62 
Figure A.1. Centrifuge table main power breaker: (a) Electrical panel MCC-6;
(b) Centrifuge table main power breaker. ................................................................... 81 
Figure A.2. Centrifuge table control panel box. .............................................................. 82 
Figure A.3. Sample LabVIEW control program. ............................................................. 83 
Figure A.4. Centrifuge table power switch. ..................................................................... 84 
Figure A.5. Neslab recirculating chiller. .......................................................................... 85 
Figure A.6. Chill bath plumbing schematic. .................................................................... 86 
Figure A.7. Booster pump control panel. ......................................................................... 87 
Figure A.8. Centrifuge table motor control power switch. .............................................. 88 
Figure C.1. LabVIEW sub-VI wire diagram for RTD read. .......................................... 100 
Figure C.2. LabVIEW sub-VI wire diagram for calibration bath temperature set. ....... 100 
Figure C.3. LabVIEW sub-VI wire diagram for calibration bath temperature read. ..... 101 
x
Figure C.4. LabVIEW VI for controlling the automatic thermocouple calibration:
(a) Front panel; (b) Wire diagram ............................................................................. 102 
Figure C.5. LabVIEW VI for manual thermocouple calibration: (a) Front panel;
(b) Wire diagram. ...................................................................................................... 103 
Figure C.6. RTD temperature vs. time from the thermocouple calibration
procedure. .................................................................................................................. 104 
Figure C.7. Sample RTD vs. thermocouple plot for TC00. ........................................... 105 
Figure C.8. LabVIEW VI for flow meter calibration program: (a) Front panel;
(b) Wire diagram. ...................................................................................................... 106 
Figure C.9. Schematic of flow meter calibration loop ................................................... 107 
Figure C.10. Temperature and flow meter voltage versus mass flow rate
calibration curve for the high-temperature fluid loop flow meter. ........................... 108 
Figure C.11. Sample data collected during one time run for the flow meter
calibration. (a) “Shotgun Blast” good data set; (b) “Trend” bad data set. ............... 109 
Figure D.1. Mounting of LHP to minimize acceleration gradient. ................................ 116 
Figure D.2. LHP survey locations.................................................................................. 117 
Figure E.1. Brayco Micronic 889 properties vs. temperature. (a) ρ vs. T; (b) k vs.
T; (c) C
p
vs. T. ........................................................................................................... 120 
Figure F.1. Updated wiring on the centrifuge table. ...................................................... 126 
Figure F.2. Wiring panel from centrifuge table to the centrifuge table control
room. ......................................................................................................................... 127 
Figure F.3. Wiring panel for the new data acquisition system. ..................................... 128 
Figure F.4. Centrifuge table voltage versus
+
r
a . ........................................................ 129 
Figure G.1. LabVIEW VI for the LHP experiment: (a) Front panel; (b) Wire
diagram. .................................................................................................................... 131 
Figure G.2. LabVIEW sub-VI wire diagram for voltage output control: (a)
Output on; (b) Output off. ......................................................................................... 132 
Figure G.3. LabVIEW sub-VI wire diagram for data acquisition communication. ...... 133 
Figure G.4. LabVIEW sub-VI wire diagram for data analyzing. .................................. 134 
Figure G.5. LabVIEW sub-VI wire diagram for data recording. ................................... 135 
xi
LIST OF TABLES
Table 1.1. Regression equations for air properties versus altitude for 1% hot
(DOD, 1997). .............................................................................................................. 13 
Table 1.2. Regression equations for air properties versus temperature (Incropera
and DeWitt, 2002). ...................................................................................................... 13 
Table 2.1. AFRL/RZPS design requirements. ................................................................. 63 
Table 2.2. ACT LHP geometric design parameters. ........................................................ 64 
Table 2.3. Summary of LHP thermocouple locations ...................................................... 65 
Table 2.4. Summary of uncertainties. .............................................................................. 66 
Table 2.5. Steady state operating characteristics for the stationary LHP (Q
cc
= 0
W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T ≤ 71.6°C, 27.6 ≤ T
amb
≤ 38.7°C)
showing effect of startup path. .................................................................................... 67 
Table 2.6. Steady state operating characteristics for the stationary LHP showing
effect of heat input to the compensation chamber (Q
in
= 500 W, a
r
= 0 g,
cp
m&
= 0.0077 kg/s, 63.4 ≤
cp
T ≤ 64.8°C, 36.1 ≤ T
amb
≤ 38.1°C). ....................................... 69 
Table 2.7. The effect of compensation chamber temperature control on LHP
operation (Q
in
= 500 W, a
r
= 0 g,
cp
m& = 0.0077 kg/s,
cp
T = 52.5°C, T
amb
=
26.4°C) ........................................................................................................................ 70 
Table 2.8. Steady state operating characteristics of the rotating LHP (Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 37.2 ≤
cp
T ≤ 59.7°C, 25.1 ≤ T
amb
≤ 30.2°C). .............................. 71 
Table 2.9. Comparison of quasi-steady states for Q
in
= 200 W (Q
cc
= 0 W,
cp
m& =
0.0077 kg/s)................................................................................................................. 73 
Table C.1. Maximum deviation of calculated RTD and experimental RTD
corresponding to each order of polynomial for thermocouple TC00. ...................... 110 
Table C.2. Coefficients for the trend line of each thermocouple. .................................. 111 
Table C.3. Maximum deviation and total error for each thermocouple. ........................ 112 
Table C.4. 3-D paraboloid regression equation for high-temperature fluid loop
flow meter. ................................................................................................................ 113 
Table D.1. Loop heat pipe mounting survey data. ......................................................... 118 
Table H.1. E1418A 8/16-CH D/A Converter wiring. .................................................... 137 
xii
Table H.2. Data acqusition terminal board wiring. ........................................................ 138 
xiii
NOMENCLATURE
a Speed of sound, m/s; acceleration, m/s
2
; flow meter calibration constant
b Flow meter calibration constant
B Experimental constant related to Eq. (F.2)
c Flow meter calibration constant
f
C Skin friction coefficient, 2С(µu
2
A)
C
p
Specific heat, J/(kg-K)
d Flow meter calibration constant
D Diameter, m
f Frequency, Hz
g Acceleration due to standard gravity, 9.81 m/s
2

h Heat transfer coefficient, W/(m
2
-K)
H Altitude, m
k Thermal conductivity, W/(m-K)
L Length, m
m Mass, kg
Ma Mach number, a U /
n Number of data points
Nu Nusselt number, bСk
Pr Prandtl number, k¡(µC
p
)
Q Heat transfer rate, W
q Heat flux, W/m
2

r Recovery factor; radial coordinate, m
R Particular gas constant, m
2
/(s
2
-K); thermal resistance, K/W
R
2
Coefficient of determination
Ra Rayleigh number, g[(I
s
-I
«

3
¡vo
xiv
Re Reynolds number, µux¡µ
St Stanton number, b¡(µuC
p
)
t Time, s; t-distribution
T Temperature, K
U Velocity, m/s
V Voltage, V; volume, L
y
0
Flow meter calibration constant

Greek Letters
α Thermal diffusivity, m
2
/s
β Inverse temperature, K
-1

γ Ratio of specific heats
ΔT Temperature difference, K
ε Emissivity
θ Angle, degrees
μ Absolute or dynamic viscosity, (N-s)/m
2

ν Kinematic viscosity, m
2
/s
ρ Density, kg/m
3

σ Standard deviation; Stefan-Boltzmann constant, 5.67×10
-8
W/(m
2
-K
4
)
φ Fluid phase
ω Angular velocity, rad/s

Superscripts
* Film condition
+ Normalized
1φ Single-phase
2φ Two-phase

Subscripts
1 Heat into vaporization of fluid
2 Heat leak to the compensation chamber
xv
∞ Freestream condition
a Actual
amb Ambient
aw Adiabatic wall
c Condenser
cc Compensation chamber
cl Centerline
cm Condenser midpoint
conv Convection
cp Cold plate
ct Centrifuge table
dev Deviation
D Diameter
e Evaporator
eg Ethylene glycol
e/cc Evaporator/compensation chamber junction
ie Inner edge
in Cold plate inlet; heat in to evaporator
L Local
max Maximum value
min Minimum value
m/t Mass/time
oe Outer edge
out Cold plate outlet; heat out of condenser
p Predicted
PAO Polyalphaolefin
r Radial
R Reference condition
rad Radiation
s Surface
sh Superheat
xvi
surr Surroundings
TC Thermocouple
tot Total
v Vapor
V/T Voltage/temperature
w Wall
z Axial
θ Azimuthal

xvii
ACKNOWLEDGEMENTS
This research effort was conducted as a part of the in-house program at the Air
Force Research Laboratory, Propulsion Directorate, Energy/Power/Thermal Division,
Thermal and Electrochemical Branch, AFRL/RZPS, Wright-Patterson Air Force Base,
Dayton, OH. I would like to thank Dr. Scott Thomas, Dr. Kirk Yerkes, and Dr. Quinn
Leland for the mentoring, wisdom, and knowledge you have given me over the past
several years. Thank you to Mr. Dave Courson for all of the technical assistance and
knowledge you have given to me throughout the duration of experimentation. Thank you
to the thermal crew for your guidance, help, and friendship: Mr. Travis Michalak, Mr.
Levi Elston, Dr. Larry Byrd, Ms. Cindy Obringer, and Ms. Bekah Puterbaugh. Thank
you to my parents, Ed and Judy, for your love, support, and guidance in helping me be
everything I am. Thank you to my wife and best friend, Jennifer, for your unfailing love
and support through our years together. Last, but most certainly not least, thank you to
my Lord and Savior Jesus Christ for all the blessings and gifts You have bestowed on me.
You are truly worth all my praise.

xviii
DEDICATION
To my beautiful wife, Jennifer, and son, Ethan. I love you.
1
1. CONVECTIVE HEAT TRANSFER FROM HIGH-SPEED AIRCRAFT SKIN
1.1. Abstract
The objective of the present analysis was to determine the feasibility of using loop
heat pipes to dissipate waste heat from power electronics to the skin of a fighter aircraft.
In the past, it has been found that the boundary condition at the condenser can be a
controlling factor in the overall performance of this type of thermal management scheme.
Therefore, the heat transfer removed from the aircraft skin has been determined by
modeling the wing as a flat plate at zero-incidence as a function of the following
parameters: airspeed: 0.8 ≤ Ma

≤ 1.4; altitude: 0 ≤ H ≤ 22 km; wall temperature: 105
≤ T
w
≤ 135°C. In addition, the effects of the variable properties of air have been taken
into account. Heat transfer due to thermal radiation has been neglected in this analysis
due to the low skin temperatures and high airspeeds up to Ma

= 1.4. It was observed
that flight speed and altitude have a significant effect on the heat transfer abilities from
the skin to ambient, with heat rejection becoming more difficult with increasing Mach
number or decreasing altitude.
1.2. Introduction
The More Electric Aircraft initiative (MEA) is the concept for future aircraft
including warfighter, transport, helicopters, and commercial aircraft. This approach has
been adopted by the United States Air Force since the early 1990’s with the purpose of
reducing or removing as many of the hydraulic, mechanical, and pneumatic power
components and replacing them with electrically driven devices. This approach to
aircraft design was first envisioned during World War II. However, at that time, the
power generation capability and power conditioning equipment required was not feasible
due to volume requirements. As a result, hydraulic, pneumatic, and mechanical systems
became the norm for aircraft until this initiative. Under the MEA paradigm, power for
systems such as flight control actuation, anti-ice, braking, environmental control, engine
starting, and fuel pumping will be provided by a starter/generator driven by the gas
2
generator spool of the aircraft engine (Quigley, 1993). The MEA initiative has been
analytically proven to improve aircraft reliability, maintainability, support, and operations
cost as well as reduce weight, volume, and enhance battle damage reconfigurability
(Cloyd, 1997).
While the reduction of hydraulic, pneumatic, and mechanical systems in favor of
electrical systems is beneficial, it presents a problem in terms of thermal management.
Replacing the centralized hydraulic system with an electrical based system removes a
primary method of transporting and removing waste heat (Vrable and Yerkes, 1998). A
separate cooling fluid system for thermal management would be contrary to the goals of
the MEA initiative. Therefore, thermal management would need to be distributed over
the entire aircraft. As a result, a new approach to thermal management involves handling
heat loads on a local level. This means taking individual components in the aircraft and
locally handling their heat rejection requirements.
The operating envelope for military aircraft places stringent limitations on any
proposed thermal management system. The on-board electrical flight control actuation
system operates at altitudes from sea level to above 12 km, airspeeds from stationary to
supersonic speeds, transient body forces up to 9 g due to maneuvering, and ambient
temperatures from -68 to 58ºC. MEA has resulted in the development of high-
temperature, high-efficiency, and high-density power electronic component technologies.
The next-generation power electronics will be capable of operating at cold plate
temperature excursions up to 200ºC, which presents an opportunity to reject heat through
the aircraft skin to the ambient using passive cooling. In addition, the actuation system
rejects heat continuously at a rate of Q = 500 W (q = 3 W/cm
2
) and has transient heat
rejection rates of Q = 5000 W over a period of one second. Possible thermal
management scenarios include direct connection of the electronics package to the skin,
high-thermal conductivity graphite straps, or the use of a loop heat pipe between the
package and the skin to provide mounting flexibility. The objective of this analysis is to
determine the external heat transfer possibilities of the aircraft skin. The heat flux and
heat transfer coefficient have been found as functions of the skin and ambient
temperatures, the altitude, and airspeed.
3
1.3. Mathematical Model
The temperature and density of air vary considerably with altitude and also vary
day-to-day depending on weather conditions. In order to be conservative in the
calculation of heat transfer coefficients, data for the highest temperature recorded with a
frequency-of-occurrence of 1% were used to generate equations for temperature and
density versus altitude (DOD, 1997) as shown in Figure 1.1 and Table 1.1. Also
presented are data for the lowest temperature recorded with a frequency-of-occurrence of
1% (DOD, 1997) and data for the “standard atmosphere” (Anderson, 2000).
The film temperature was used as the reference temperature to evaluate the air
properties (White, 1988)
I
-
= I

(u.S + u.uS9Na
«
2
) + u.SI
w
(1.1)
The air density at the film temperature and at altitude was evaluated using the perfect gas
law

µ
-
= µ

|
I

I
-
1
(1.2)
The freestream speed of sound is
o

= .yRI

(1.3)
The freestream velocity is
u

= Na

o

(1.4)
The absolute viscosity of air is given by the following relation (NACA, 1953)

µ = µ
R
|
I
I
R
1
0.76

(1.5)
where μ
R
is a reference viscosity evaluated at a known reference temperature T
R
.
The Reynolds number for a plate of length L is determined by evaluating the
properties of air at the freestream condition.
Re
L
=
µ

u

I
µ

(1.6)
4
Regression equations for the specific heat and Prandtl number were determined as
functions of temperature using data from Incropera and DeWitt (2002), as shown in Table
1.2.
The adiabatic wall temperature is (White, 1988)

I
aw
= I

|1 + r |
y - 1
2
1 Na

2
|
(1.7)
where the recovery factor is
r = |
Pi
1¡2
foi laminai flow
Pi
1¡S
foi tuibulent flow
(1.8)
For the purposes of this analysis, Reynolds numbers less than 500,000 were considered to
be laminar, greater than 500,000 were turbulent. The local skin friction coefficient at the
end of the plate was found by evaluating the air properties at the film temperature. For
laminar flow, the skin friction coefficient is given by (White, 1988)

C
f,L
-
=
u.664
I
µ
-
u

I
µ
-
]
1 2 ⁄

(1.9)
and for turbulent flow

C
f,L
-
=
u.4SS
ln
2
I
u.u6µ
-
u

I
µ
-
]

(1.10)
The local Stanton number at the end of the plate for laminar flow is given by (White,
1988)
St
L
-
= u.SS2Re
L
1¡2
Pi
-2¡3
(1.11)
and for turbulent flow

St
L
-
=
b
L
µ
-
u

C
p
-
=
C
f,L
2
2 ⁄
1 +12.7(Pi
-2¡3
- 1)(C
f,L
-
2) ⁄ )
1¡2

(1.12)
The local heat transfer coefficient at the end of the plate is
b
L
= St
L
-
µ
-
u

C
p
-
(1.13)
5
The local heat transfer coefficient was calculated using the appropriate skin friction
coefficient and Stanton number based on laminar or turbulent flow. The average heat
transfer coefficient over the length of the plate is approximated by (White, 1988)
b

= 1.1Sb
L
(1.14)
The heat flux dissipated over the plate, both local and average, is defined in terms of the
adiabatic wall temperature (White, 1988)
o
w
= b(I
w
- I
aw
) (1.15)
Thermal radiation was neglected in this analysis as it contributed less than 1.6% to the
total heat rejected from the plate surface.
1.4. Results and Discussion
The adiabatic wall temperature is shown in Figure 1.2 as a function of altitude and
Mach number. The overall trend of the adiabatic wall temperature with altitude follows
the freestream air temperature in Figure 1.1 and increases with Mach number as
expected. Figure 1.3 presents the temperature difference, ΔT = (T
aw
- T

), versus altitude.
This temperature difference demonstrates the increase in the adiabatic wall temperature
over the freestream due to aerodynamic heating. The temperature difference ΔT = (T
w
-
T
aw
) is given in Figure 1.4. Of interest is the portion of the curves in which this
difference is negative, which indicates that heat is transferred from the air to the aircraft
skin. The maximum Mach number achievable before heat is transferred from the air to
the skin is given by

Na
∞,max
= |
1
r
|
I
w
I

- 11 |
2
y - 1
1|
1¡2

(1.16)
and is plotted in Figure 1.5 over a range of wall temperatures. The maximum Mach
number increases with altitude and wall temperature up to a maximum at approximately
18 km. In Figure 1.6, the average convective heat transfer coefficient decreases
monotonically with altitude due to the continual decrease in the air density. In general,
the convective heat transfer coefficient increases with Mach number, as expected. The
average heat flux dissipated from the plate is shown in Figure 1.7. For low Mach
numbers, the heat flux is positive for all values of altitude, which indicates that heat is
6
transferred from the aircraft skin to the air. At high Mach numbers, however, the heat
flux is negative at low altitudes due to the negative ΔT as shown in Figure 1.4. This
means that the adiabatic wall temperature is higher than the skin temperature due to
aerodynamic heating effects. The effect of heated plate length on the local heat flux for
H = 0, 10, and 20 km is shown in Figure 1.8. The local heat flux starts low and decreases
in the laminar region of the plate, and then increases as the flow transitions to turbulent
where it once again decreases. In general, the average heat flux follows the behavior of
the local heat transfer coefficient, where h
L
is high at the leading edge and at the
beginning of turbulent flow and decreases as the boundary layer grows. One item to note
is that Figure 1.9 shows the average heat flux dissipated over the plate versus altitude for
the 1% hot day, the 1% cold day, and the standard atmosphere data as presented in Figure
1.1. At low altitudes,
w
q

is significantly higher for the 1% cold day due to the combined
effects of the lower atmospheric temperature and the higher air density. The effect of
wall temperature

on average heat flux for a given airspeed is shown in Figure 1.10. The
heat flux increases dramatically with altitude and wall temperature for low altitudes.
1.5. Conclusions
An analysis of the heat transfer from a heated plate has provided important
insights for the possible use of the aircraft skin to reject heat from electric actuator
systems. It was found that the altitude and speed of the aircraft significantly affected the
amount of heat that could be rejected from the skin. Aerodynamic heating of the skin
reduced the heat transfer, and if the Mach number was high enough, heat transfer from
the skin to the air went to zero. A performance map of this phenomenon was provided.
The altitude of the aircraft affected the freestream temperature and density, which in turn
affected the overall heat transfer coefficient. It was also shown that the assumption of a
“standard atmosphere” could result in significant errors in the prediction of the heat
dissipation as compared to the data for the 1% hot day or the 1% cold day. The analysis
showed that the aircraft skin temperature, which is directly influenced by the actuator
thermal management system, has a strong effect on the heat dissipation rate, especially at
low altitudes.
7


Figure 1.1. Comparison of atmospheric properties versus altitude: (a) Temperature; (b)
Density (DOD, 1997; Anderson, 2000).
180
200
220
240
260
280
300
320
340
T

(
K
)
1% Hot Day
Standard Atmosphere
1% Cold Day
(a)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0 5 10 15 20
ρ

(
k
g
/
m
3
)
H (km)
(b)
8


Figure 1.2. Adiabatic wall temperature versus altitude for various Mach numbers (1%
hot day).





Figure 1.3. Temperature difference ) (
aw ∞
−T T versus altitude for various Mach numbers
(1% hot day).

0
50
100
150
200
250
300
350
400
450
0 5 10 15 20
H (km)
T
a
w



(
K
)
Ma = 0.8
Ma = 0.98
Ma = 1.2
Ma = 1.4
0
20
40
60
80
100
120
140
0 5 10 15 20
Δ
T

=

[
T
a
w
-
T

]




(
K
)
H (km)
Ma = 0.8
Ma = 0.98
Ma = 1.2
Ma = 1.4
9

Figure 1.4. Temperature difference ) (
aw w
T T − versus altitude for various Mach numbers
(T
w
= 135ºC, 1% hot day).





Figure 1.5. Maximum Mach number before heat is transferred from the air to the skin
versus altitude for various wall temperatures (1% hot day).

0
20
40
60
80
100
120
140
0 5 10 15 20
Δ
T

=

[
T
a
w
-
T

]




(
K
)
H (km)
Ma = 0.8
Ma = 0.98
Ma = 1.2
Ma = 1.4
0
0.5
1
1.5
2
2.5
0 5 10 15 20
M
a


,

m
a
x
H (km)
Tw = 105 ºC
Tw = 115 ºC
Tw = 125 ºC
Tw = 135 ºC
10

Figure 1.6. Average convective heat transfer coefficient versus altitude for various Mach
numbers (T
w
= 135ºC, L = 1.0 m, 1% hot day).





Figure 1.7. Average heat flux dissipated over the plate versus altitude for various Mach
numbers (T
w
= 135ºC, L = 1.0 m, 1% hot day).


0
100
200
300
400
500
600
700
800
0 5 10 15 20
h




(
W
/
m
2
-
K
)
H (km)
Ma = 0.8
Ma = 0.98
Ma = 1.2
Ma = 1.4
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
3
0 5 10 15 20
q
w
(
W
/
c
m
2
)
H (km)
Ma = 0.8
Ma = 0.98
Ma = 1.2
Ma = 1.4
11



Figure 1.8. Local heat flux dissipated over the plate versus plate length for various Mach
numbers (T
w
= 135ºC, 1% hot day): (a) H = 0 km; (b) H = 10 km; (c) H = 20 km.

-4
-3
-2
-1
0
1
2
3
4
q
w
(
W
/
c
m
2
)
Ma = 0.8
Ma = 0.98
Ma = 1.2
Ma = 1.4
(a)
0
0.5
1
1.5
2
2.5
3
q
w
(
W
/
c
m
2
)
(b)
0
0.2
0.4
0.6
0.8
1
1.2
0 0.5 1 1.5 2
q
w
(
W
/
c
m
2
)
L (m)
(c)
12

Figure 1.9. Average heat flux dissipated over the plate versus altitude for various
atmospheric conditions (T
w
= 135ºC, L = 1.0 m, Ma

= 0.98) (DOD, 1997; Anderson,
2000).





Figure 1.10. Average heat flux dissipation versus altitude for various wall temperatures
(Ma

= 0.98, L = 1.0 m, 1% hot day).


0
1
2
3
4
5
6
7
8
0 5 10 15 20
q
w
(
W
/
c
m
2
)
H (km)
1% Hot Day
Standard Atmosphere
1% Cold Day
0
0.5
1
1.5
2
2.5
0 5 10 15 20
q
w
(
W
/
c
m
2
)
H (km)
Tw = 105 ºC
Tw = 115 ºC
Tw = 125 ºC
Tw = 135 ºC
13
Table 1.1. Regression equations for air properties versus altitude for 1% hot (DOD,
1997).

y = a
0
+ a
1
H + a
2
H
2
+ a
3
H
3
+ a
4
H
4

(H in km)

Property a
0
a
1
a
2
a
3
a
4
R
2

T

(ºC) 4.8507E+1 -9.5033E+0 5.3483E-1 -2.8994E-2 7.7664E-4 0.99779
ρ

(kg/m
3
) 1.0868E+0 -8.9917E-2 2.0898E-3 -4.9336E-6 — 0.99954





Table 1.2. Regression equations for air properties versus temperature (Incropera and
DeWitt, 2002).

y = a
0
+ a
1
T + a
2
T
2
+ a
3
T
3

(T in K)

Property a
0
a
1
a
2
A
3
R
2

c
p
(J/kg-K) 1.0187E+3 -6.9921E-2 -3.3333E-5 4.4444E-7 0.99916
Pr

8.6418E-1 -9.4177E-4 1.7778E-6 -1.2593E-9 0.99725

14
2. TITANIUM-WATER LOOP HEAT PIPE CHARACTERISTICS UNDER
STATIONARY AND ELEVATED ACCELERATION FIELDS
2.1. Abstract
An experiment has been developed to examine operating characteristics of a
titanium-water loop heat pipe (LHP) under stationary and elevated acceleration fields.
The LHP was mounted on a 2.44 m diameter centrifuge table on edge with heat applied
to the evaporator via a mica heater and heat rejected using a high-temperature
polyalphaolefin coolant loop. The LHP was tested under the following parametric
ranges: heat load at the evaporator: 100 ≤ Q
in
≤ 600 W; heat load at the compensation
chamber: 0 ≤ Q
cc
≤ 50 W; radial acceleration: 0 ≤ a
r
≤ 10 g. For stationary operation (a
z

= 1.0 g, a
r
= 0 g), the LHP evaporative heat transfer coefficient decreased monotonically,
thermal resistance decreased to a minimum then increased, and wall superheat increased
monotonically. Heat input to the compensation chamber was found to increase the
evaporative heat transfer coefficient and decrease thermal resistance for Q
in
= 500 W.
Flow reversal in the LHP was found for some cases, which was likely due to vapor
bubble formation in the primary wick. Operating the LHP in an elevated acceleration
environment revealed dry-out conditions from Q
in
= 100 to 400 W and varying
accelerations and the ability for the LHP to reprime after an acceleration event that
induced dry-out. Evaporative heat transfer coefficient and thermal resistance was found
not to be significantly dependent on radial acceleration. However, wall superheat was
found to increase slightly with radial acceleration.
2.2. Introduction
Loop heat pipes (LHP's) are two-phase thermal transport devices that operate
passively using the latent heat of vaporization to transport heat from one location to
another. The LHP was invented in 1972 by Gerasimov and Maidanik (Maidanik, 2005)
in the former Soviet Union, and was later patented in the United States (Maidanik et al.,
1985). The LHP consists of an evaporator, compensation chamber, liquid and vapor
15
transport lines made of smooth tubing, and a condenser as shown in Figure 2.1. Heat is
applied directly to the exterior wall of the evaporator, which often has a circular cross-
section. The majority of the input heat is used to vaporize the working fluid within the
primary wick structure, which is an inverted meniscus wick in direct contact with the
exterior evaporator wall. The vapor is captured in the axial vapor grooves in the primary
wick and is directed via a manifold at the end of the evaporator to the vapor line due to
the increased pressure within the evaporator. Due to evaporation, menisci are developed
in the primary wick which establishes a capillary pressure head that returns liquid to the
evaporator from the condenser. This capillary head must be greater than the total system
pressure drop in order for the LHP to continue to operate without drying out.
The vapor from the evaporator section travels via the vapor line to the condenser
section, which is also made of smooth tubing. Heat is rejected from the condenser to the
ultimate heat sink. The working fluid enters the condenser as a superheated vapor. After
sufficient heat is rejected, the vapor becomes a saturated vapor, a two-phase mixture, a
saturated liquid, and, depending on the amount of heat rejection, it may or may not
become a subcooled liquid. The location of the point at which the working fluid becomes
a subcooled liquid (2φ-1φ) is dependent on the heat input at the evaporator, the heat
rejection at the condenser, and the saturation temperature in the compensation chamber.
After exiting the condenser section, the liquid will continue to lose heat due to convection
and/or thermal radiation to the ambient. The subcooled liquid returns to the evaporator
via the bayonet tube, which delivers the liquid to the end of the evaporator where the
vapor manifold resides.
As stated previously, most of the evaporator heat input evaporates liquid in the
primary wick. The rest of the heat is transferred by conduction through the primary wick,
where liquid is evaporated into vapor channels leading to the compensation chamber
(Figure 2.2). Part of this vapor stream condenses onto the secondary wick, which is in
intimate contact with the bayonet tube. This heat transfer to the bayonet tube raises the
temperature of the subcooled liquid entering the compensation chamber to the saturation
temperature as it travels to the end of the evaporator. The rest of the vapor condenses
onto the wick lining the compensation chamber. This latent heat is then rejected from the
compensation chamber to the ambient. The condensate in the compensation chamber is
16
drawn back to the evaporator section through the secondary wick by capillary action. In
this way, the secondary wick and the compensation chamber behave similar to a
conventional heat pipe.
The compensation chamber allows the LHP to automatically regulate itself during
transient situations like startup, shutdown, or a change in the operating conditions. The
compensation chamber provides for storage of excess liquid when the evaporator heat
input is high, where the majority of the condenser section is free of subcooled liquid. The
compensation chamber can also be used to control the location of the 2φ-1φ point in the
condenser. Controlling the heat transfer through the shell of the compensation chamber
can adjust the saturation point in the condenser, thereby changing the amount of
subcooling of the liquid returning to the evaporator.
There has been limited experimentation on the acceleration effects on loop heat
pipes and heat pipes. Ku et al. (2000a) performed experiments on a miniature
aluminum/anhydrous ammonia LHP by using a spin table to examine the effects of
varying acceleration on start-up. Four mounting configurations were examined: (1)
horizontally with the compensation chamber and liquid line outboard on the table, (2)
horizontally with the evaporator and vapor line outboard on the table, (3) vertically with
evaporator above the compensation chamber with no radial acceleration, and (4)
vertically with evaporator below the compensation chamber with no radial acceleration.
Several different experiments were conducted, including LHP startup before acceleration
was applied and vice versa, as well as varying heat load inputs up to Q
in
= 100 W.
Several acceleration profiles were examined, including a
r
= 0.0 g, constant a
r
= 1.2 g,
constant a
r
= 4.8 g, combination of constant a
r
= 1.2 and 4.8 g, constant a
r
= 1.2 g for 30
seconds followed by a
r
= 0.0 g for 300 seconds periodically, constant a
r
= 4.8 g for 30
seconds followed by a
r
= 0.0 g for 300 seconds periodically, and combinations of a
r
=1.2
and 4.8 g followed by a
r
= 0.0 g for 300 seconds periodically. Their experimental results
indicated that the wall superheat, defined as the difference between the evaporator and
compensation chamber wall temperatures, appeared to be independent of input heat load
and acceleration. When temperature overshoot in the evaporator was examined, for heat
loads greater than Q
in
= 50 W, there was essentially no overshoot. For smaller heat loads,
such as at Q
in
= 5 W, a temperature overshoot of a few degrees was always observed, but
17
at Q
in
= 25 W, the temperature overshoot ranged from 0 to 45°C. In every experiment,
the LHP started successfully.
Ku et al. (2000b), in an extension of the previous experimental study, examined
the temperature stability of the same miniature LHP under varying heat loads and
acceleration levels. Their experimental results showed that the radial acceleration caused
a redistribution of fluid in the evaporator, condenser, and compensation chamber. This in
turn changed the LHP operating temperature. The effect was not universal, in the sense
that all the operating conditions needed to be taken into account. With sufficient time,
constant acceleration could either increase or decrease the LHP operating temperature.
Periodic acceleration led to a quasi-steady operating temperature. Temperature hysteresis
could also be caused by the radial acceleration. In all of the experiments the LHP
continued to operate without problems.
Similar research has been conducted to examine body force effects on heat pipes.
Ponnappan et al. (1992) examined a flexible copper-water arterial wick heat pipe
subjected to transverse acceleration using a centrifuge table. Evaporator heat loads up to
Q
in
= 150 W and steady state radial accelerations up to a
r
= 10.0 g were investigated.
Transport capacity of the heat pipe dropped from Q
out
= 138 W at radial accelerations of
a
r
= 1.0 g to Q
out
= 60 W at a
r
= 10.0 g. The temperature difference between the
evaporator and condenser remained fairly constant up to a
r
= 4 g then decreased from a
r
=
4 to 10 g. This decrease was due to a more uniform distribution of fluid within the wick
at the higher radial acceleration.
Yerkes and Beam (1992) examined the same flexible copper-water arterial wick
heat pipe as Ponnappan et al. under transient transverse and axial acceleration forces with
periodic and burst transverse accelerations from f = 0.01 to 0.03 Hz and magnitudes from
a
r
= 1.1 to 9.8 g peak-to-peak and evaporator heat inputs up to Q
in
= 83 W. It was
observed that pooling of excess fluid had a significant effect on the heat transport of the
heat pipe at steady state transverse acceleration. Heat transport potential decreased with
increasing transverse acceleration causing partial dry-out of the artery and pooling in the
condenser. The heat pipe was able to reprime after dry-out events with subsequent
reduction of transverse acceleration. Under cyclic transverse acceleration, significant
fluid slosh was thought to create a cyclic variation in heat pipe temperature. Temperature
18
rise was lower at the onset of dry-out conditions when compared to steady state
transverse acceleration. Frequency of the steady periodic burst transverse acceleration
had no effect on the heat pipe temperature and tended to delay the onset of dry-out.
Thomas and Yerkes (1996) examined the same flexible copper-water arterial wick
heat pipe as Ponnappan et al. with evaporator heat loads from Q
in
= 75 to 150 W,
condenser temperatures of T
c
= 3, 20, and 35°C, and sinusoidal acceleration frequencies
of f = 0, 0.01, 0.05, 0.1, 0.15, and 0.2 Hz. The amplitude of the radial acceleration ranged
from a
r
= 1.1 to 9.8 g. The effects of the previous dry-out history of the heat pipe were
also examined. It was discovered that the thermal resistance increased and then
decreased with respect to increasing acceleration frequency. The thermal resistance also
increased with increasing evaporator heat loads. The previous dry-out history adversely
affected the thermal resistance of the heat pipe when dry-out occurred prior to increasing
the acceleration frequency.
Thomas et al. (1998) examined a helically grooved copper-ethanol heat pipe as a
function of evaporator heat input and transverse radial acceleration. Heat loads ranging
from Q
in
= 20 to 250 W were applied to the evaporator. At Q
in
= 20 W the heat pipe did
not experience any dry-out conditions when the radial acceleration was increased and
then decreased stepwise from a
r
= 0 to 10 g. At Q
in
= 50 W, the heat pipe experienced
dry-out conditions at a
r
= 0 and 2 g, but quickly reprimed at the higher radial
accelerations. This indicated the elevated body forces actually aided the performance of
the heat pipe by increasing the capillary limit due to the forces generated from
acceleration gradients down the length of the helical groove. The thermal resistance of
the heat pipe was noted to decrease then increase with increasing heat transported when
dry-out started.
Zaghdoudi and Sarno (2001) examined the body force effects on a flat copper-
water heat pipe via a centrifuge setup. The heat pipe was mounted such that the
accelerating forces were opposite to the liquid flow, or in an “unfavorable” mounting
condition. Three types of accelerations were performed in this study: A parabolic profile
from a
r
= 0 to 10 to 0 g with a 5 second stabilization at a
r
= 10 g, a step increase from a
r
=
0 to 10 to 0 g with a 10 second stabilization at each step, and increasing then decreasing
the acceleration from a
r
= 0 to 10 g after thermal stabilization. Heat loads of Q
in
= 20, 40,
19
and 60 W were applied to examine the effect on evaporator and condenser temperature as
well as thermal resistance. For the first two types of acceleration profile, it was observed
there was a delayed increase in evaporator temperature and decrease in condenser
temperature. This was likely due to the pooling of fluid in the condenser. Thermal
resistance also experienced a delayed increase in onset and remained elevated even in the
absence of an accelerating force. For the third type of acceleration profile, there was a
much more gradual increase in evaporator temperature and nearly negligible decrease in
condenser temperature, quickly returning to normal in the absence of the accelerating
force. Thermal resistance had a similar trend, quickly returning to normal after the
acceleration burst. This suggested that the heat pipe quickly reprimed after the
acceleration event. These tests demonstrated the importance of prior operation history
when the heat pipe was subjected to elevated body forces.
The objective of the present experiment was to determine the operating
characteristics of a titanium-water loop heat pipe subjected to varying heat loads and
accelerations. Transient temperature distributions, the evaporative heat transfer
coefficient, and the thermal resistance have been found in terms of the heat input at the
evaporator, heat input at the compensation chamber, and radial acceleration field. In
addition, the transient behavior during startup and steady operation has been examined.
A performance map has been developed that relates dry-out to the heat load and radial
acceleration for the experimental conditions described. The experimental parametric
ranges were as follows: heat load at the evaporator: 100 ≤ Q
in
≤ 600 W; heat load at the
compensation chamber: 0 ≤ Q
cc
≤ 50 W; radial acceleration: 0 ≤ a
r
≤ 10 g.
2.3. Experimental Setup
The Centrifuge Table Test Bed at Wright-Patterson Air Force Base (AFRL/RZPS)
was used to determine the heat transfer characteristics of the titanium-water LHP under
stationary and elevated acceleration fields. A schematic of this test bed can be seen in
Figure 2.3. The test bed consisted of a 2.44 m diameter horizontal rotating table driven
by a 20 hp DC electric motor. The test bed was able to deliver the following to devices
mounted to the rotating table: Conditioned DC electrical power through three separate
power supplies, 120 VAC power, temperature-controlled ethylene glycol coolant, and
electrical signals for analog or digital control. In addition, electrical signals were
20
collected from instruments on the table and stored in a data acquisition computer. The
radial acceleration could exceed a
r
= 12 g, with a maximum onset of approximately
r
a& =
10 g/s, inducing a tangential acceleration. The acceleration field could be varied
manually using a potentiometer, or controlled digitally using a signal generator in the
data acquisition system. The acceleration field was measured using an orthogonal triaxial
accelerometer (Columbia SA-307HPTX) with an uncertainty of ± 0.01 g.
Power was supplied to heaters on the table by three precision power supplies
(Kepco ATE150-7M, Kepco ATE150-3.5M, and HP 6290A) through power slip rings.
These slip rings were separated from the instrumentation slip rings to reduce electrical
noise. The heater power was calculated by multiplying the voltage drop across the heater
by the current. The current was determined from the voltage drop across a precision
resistor in series with the heater. This type of measurement was required due to the
voltage drop between the control room and the table. The uncertainty in this
measurement was less than 2.0%
Heat was rejected from the centrifuge table using an ethylene-glycol/water
mixture that was delivered to the rotating centrifuge table via a double-pass hydraulic
rotary coupling (Deublin 1690-000-115). The temperature of the coolant was maintained
at a constant setting by a recirculating chiller (Neslab HX-300). The volumetric flow rate
of the coolant mixture was controlled using a high-pressure booster pump, which aided
the low-pressure pump in the recirculating chiller. Throughout experimentation the flow
rate was held constant at
eg
V
&
= 2.4 L/min.
Instrumentation signals generated on the table were acquired through a custom-
built forty-channel instrumentation slip ring using a data acquisition system.
Temperatures, mass flow rates, accelerations, and voltages were all measured using a
data acquisition mainframe (Agilent VXI E8408A) with a command module (Agilent
E1406A), 5½ digit multimeter module (Agilent E1411B), and a 64-channel 3-wire
multiplexer module (Agilent E1476A). The rotational speed of the centrifuge table,
heater power, and other low voltage control devices on the table were controlled using an
8/16-channel D/A converter module (Agilent E1418A). Communication between the data
acquisition unit and the computer was established using a general purpose interface bus
(GPIB) coupled with a custom-designed LabVIEW virtual instrument.
21
Gathering temperature data from rotating machinery using slip rings presents
unique problems. First, when the thermocouple wires are connected to the wires leading
to a slip ring, at least one extra junction is created, depending on the materials of the
thermocouple wires. To avoid this problem, a Type E thermocouple amplifier was
installed on the centrifuge table (Omega OM7-47-E-07-2-C) with internal cold junction
compensation. This converted the millivoltage signals from the thermocouples to 0 to 10
V signals without the creation of extra junctions. Another problem that is present when
slip rings are used is electrical noise. This problem was reduced (not eliminated) by the
use of a low-pass filter for each of the thermocouple signals coming from the table before
the data acquisition system.
The test article, a titanium-water loop heat pipe, was developed for AFRL/RZPS
by Advanced Cooling Technologies (ACT), Inc., in Lancaster, PA, under contract
FA8601-06-P-0076. Initial design parameters set by AFRL/RZPS were to develop a loop
heat pipe capable of a minimum heat load of 500 W and minimum heat flux of 3 W/cm
2
.
The minimum transport line length was 2 m to simulate relevant aircraft geometries. An
evaporator operating temperature of 200°C and condenser operating temperature between
5 and 140°C were selected to match relevant acquisition and rejection temperatures
aboard aircraft. The evaporator and condenser dimensions were selected to be 20.32 ×
10.16 cm and 30.48 × 28.56 cm, respectively, to match commercial off-the-shelf heaters
and cold plates. A summary of the requested design parameters can be seen in Table 2.1.
After several design iterations, ACT delivered the loop heat pipe shown in Figure 2.4. A
summary of the loop heat pipe specifications can be seen in Table 2.2. The LHP was
instrumented with twelve type E exposed tip thermocouples as seen in Figure 2.5. A
summary of their locations can be seen in Table 2.3.
The loop heat pipe was mounted onto the centrifuge table such that the centerline
of the tubing coincided with the outer table radius as much as possible. Small deviations
existed since the condenser section and the evaporator/compensation chamber were both
straight. This induced a non-uniform radial acceleration field over the lengths of these
sections that needed to be quantified. Stands were designed using G-7 phenolic to mount
the loop heat pipe with support at the compensation chamber, evaporator, condenser, and
transport lines (Figure 2.6). The tops of these stands were anchored to the table to reduce
22
deflection when the table was rotating. A survey was taken at 22 locations on the loop
heat pipe to determine how far various portions of the loop heat pipe were from the
centerline radius. The loop heat pipe had a minimum radius to centerline of 119.2 cm
and a maximum radius to centerline of 123.3 cm. The entire loop heat pipe fitted within
4.6 cm for a percent acceleration difference of 3.8%. Complete survey data can be seen
in Appendix D. To minimize heat loss to the environment, the entire assembly was
thoroughly insulated using Kaowool blankets and aluminum foil. The assembly was
placed inside an aluminum frame (80/20, Inc.) for structural support and enclosed with
sheet metal sides to minimize convective heat losses.
During operation, heat was applied to the LHP at the evaporator while the heat
transfer to the compensation chamber was independently controlled. A mica heater
(Minco) was located between the evaporator body and a ceramic fiber insulative layer,
followed by the evaporator stand. A flexible electric heat tape (Thermolyne) was wound
around the compensation chamber and surrounded by Kaowool insulation and aluminum
foil to minimize heat losses. In normal operation, the compensation chamber is not
insulated and the temperature is closely controlled during operation. For these
experiments, insulating the LHP, including the compensation chamber, was selected to
mimic a typical configuration of a LHP in an aircraft environment where bay
temperatures could be higher than the LHP temperatures. This would minimize parasitic
heat gain, and reduce the use of external heaters or coolers on the compensation chamber.
As a result, the LHP compensation chamber was allowed to “float” into equilibrium with
the evaporator and condenser, rather than controlling the temperature of the evaporator
by controlling the compensation chamber temperature.
As previously mentioned, the centrifuge table was equipped with an on-board
fluid loop for dissipating heat from sources on the table, which used ethylene glycol as its
working fluid. In the present experiment, it was desired to have the option of operating
the LHP condenser section at elevated temperatures, so a high-temperature fluid loop was
constructed and mounted to the centrifuge table to act as an interface between the LHP
and the low-temperature fluid loop, as shown in Figure 2.7. The high-temperature
working fluid (Brayco Micronic 889 polyalphaolefin or PAO oil) flowed from the
custom-made copper reservoir into a positive displacement gear pump (Tuthill). After
23
passing through a filter and a flow-straightening section, the PAO was directed through
the turbine flow meter (Omega FTB-9506). An electrical tape heater was mounted to the
copper tubing after the flow meter to allow for preheating the PAO prior to reaching the
calorimeter on the condenser section, which consisted of three heat exchangers plumbed
in series and mounted to the condenser section. Type E thermocouple probes were
installed at the inlet and outlet of the three heat exchangers for calorimetry (TC00 and
TC01), and another was placed prior to the flow meter (TC03). This was needed due to
the dependence of the viscosity of PAO on temperature. After the PAO exited the three
heat exchangers on the condenser, it flowed to a liquid/liquid heat exchanger that
transferred heat from the high-temperature coolant loop to the low-temperature ethylene
glycol loop. The PAO then returned to the reservoir.
Four grounded probe thermocouples for the high temperature loop and twelve
exposed tip type E thermocouples mounted on the LHP were used in the experiment.
Thermocouple calibrations were conducted over two temperature ranges depending on
the anticipated operating temperatures. The grounded probe thermocouples were used for
calorimetry, coolant flow meter calibration and the measurement of the ambient
temperature, where the error needed to be minimized. These four thermocouples were
calibrated over the anticipated range of 20 to 145°C in 5°C intervals. The twelve
exposed tip thermocouples were mounted on the LHP in various locations and needed to
be calibrated over the full range of 20 to 230°C in 5°C increments. The calibration
procedure consisted of using two separate recirculating chiller baths (Brinkmann Lauda
RCS 20-D, T = 20 to 140°C; Hart Scientific 6330, T = 40 to 230°C) with PAO as the
working fluid to achieve the required temperature range. The temperature readings from
the sixteen thermocouples were compared to a NIST-traceable platinum resistance
temperature detector (Hart Scientific RTD 1502A) with a resolution of ± 0.009°C. To
ensure that the bath had reached steady state at a given temperature, the RTD temperature
was continuously monitored. When the standard deviation of 100 readings dropped
below the specified threshold of 0.005°C, 100 readings from the thermocouples were
sampled, stored in an array, and the bath temperature was changed. For repeatability, the
bath temperature was first incremented from the lowest temperature to the highest
temperature, and then decremented from highest to lowest, and the two sets of 100 data
24
points collected for each thermocouple at a given temperature were used to determine
two average readings. Plots of the RTD temperature versus each thermocouple
temperature were generated, and polynomial trend lines were fitted for each
thermocouple as can be seen in Appendix C. A fifth-order polynomial was selected since
it reduced the maximum deviation from the data by approximately a factor of four over a
first-order trend line. The uncertainty associated with each thermocouple was determined
by accounting for four sources of error: the stated uncertainty of the RTD, the confidence
interval of the RTD average reading at a confidence level of 0.95, the confidence interval
of the thermocouple average reading at a confidence level of 0.95, and the maximum
deviation of the temperature calculated using the polynomial curve fit from the actual
measured temperature.
The turbine flow meter used in the high-temperature fluid loop was calibrated to
achieve accurate results for the amount of heat extracted from the LHP. This was critical
for the calculation of the evaporative heat transfer coefficient and the thermal resistance
of the LHP. Since the viscosity of the PAO, used in the high-temperature fluid loop,
changes significantly with temperature, a “calibration surface” was generated that related
the output voltage of the flow meter and the temperature of the PAO at the entrance of
the flow meter to the mass flow rate. The calibration setup consisted of a recirculating
chiller bath (Brinkmann Lauda RCS 20-D) filled with PAO from the same source as used
in the high-temperature fluid loop. The gear pump, inline filter (Whitey SS56S6 140
micron) and a calibrated grounded thermocouple probe, from the high-temperature fluid
loop, were installed in a line from the bath to the turbine flow meter (Omega FTB-9506)
and signal conditioner (Omega FLSC-61). Flow straightening sections upstream and
downstream were placed according to the manufacturer's instructions. A three-way valve
was installed after the flow meter, which allowed the entire flow system to reach a steady
temperature. Once the temperature was steady, the flow was diverted to a catch basin for
a specified amount of time. The voltage from the flow meter and the temperature from
the thermocouple were recorded during this time, and when the basin was full, the flow
was again diverted to recirculating the PAO back to the chiller bath. All of the data was
collected through the instrumentation slip rings on the centrifuge table to the data
acquisition system to capture all errors inherent to the centrifuge table test bed. A lab
25
scale (Mettler PC4400) was used to determine the mass collected during a given test run
to within ± 0.3 gm. During each measurement, as many data points as possible were
collected across the time span with the limiting factor being the iteration time on the
LabVIEW software. The minimum number of data points collected for any given run
was 437. The voltages and temperatures were averaged and a confidence interval was
calculated based on a confidence level of 0.95 for each test run. The test was repeated for
a total of five averaged data points for each nominal temperature and flow rate. These
tests were completed over the range of T = 20 to 120°C in intervals of 25°C and flow
rates ranging from m& = 0.0064 to 0.025 kg/s in intervals of approximately 0.002 kg/s. A
3-D paraboloid regression equation was generated using SigmaPlot to relate temperature,
flow meter voltage, and mass flow rate, and was given by
m
cp
= y
0
+ oI + bI + cI
2
+JI
2
(2.1)
where y
0
, a, b, c, and d are calibration constants (Appendix C). The general root-sum-
square uncertainty equation used for all uncertainties was given by

Δy = |||
oy
ox
1
1 Δx
1
·
2
+ ||
oy
ox
2
1 Δx
2
·
2
+ ·|
1¡2

(2.2)
where y = f(x
1
, x
2
, …). The uncertainty of the mass flow rate measurement was affected
by the maximum deviation of the regression equation from the actual data, the confidence
interval for the temperature and flow meter voltage measurements, the root-sum-square
total error associated with the scale and stopwatch given by

Δm
m¡t
= |||
1
t
1 Δm+
2
+ |I-
m
t
2
] Δt+
2
|
1¡2

(2.3)
and the root-sum-square error associated with the temperature and voltage measurements
given by

Δm
v¡T
= |((o +2cI)ΔI)
2
+ ((b + 2JI)ΔI)
2
|
1¡2

(2.4)
26
The percent error on the mass flow rate decreased with increasing flow rate. Since the
mass flow rate was kept constant at
cp
m&

= 0.0077 kg/s, the uncertainty associated with
that setting was 4.0%.
The heat transferred from the LHP condenser to the cold plate, Q
out
, was defined
as
0
out
= m
cp
C
p,PA0
(I
out
- I
in
) (2.5)
A linear fit equation for
PAO p,
C as a function of temperature was developed by Ghajar et
al. (1994) and used in equation (2.5) (Appendix E). The average evaporative heat
transfer coefficient was defined as
b

÷
0
out
nÐI(I

e
- I
v
)
(2.6)
where D is the inside diameter of the evaporator shell, L is the length of the evaporator,
e
T is the average evaporator temperature measured by the four thermocouples embedded
in the wall between the heater and the wick (Figure 2.5(b)), and T
v
is the external
temperature of the vapor line at the outlet of the evaporator. The heat rejected to the cold
plate, Q
out
, was selected as it was the best estimate of heat actually transported by the
LHP. The thermal resistance of the loop heat pipe, R, was determined using the average
evaporator temperature and the average temperature of the cold plate, and was defined as

R ÷
I

e
- I

cp
0
out

(2.7)
where
cp
T is the average cold plate temperature. The root-sum-square uncertainty of Q
out
,
h , and R are given by

Δ0
out
= |(C
p,PA0
(I
out
- I
in
)Δm
cp
)
2
+(m
cp
(I
out
- I
in
)ΔC
p,PA0
)
2
+ (m
cp
C
p,PA0
ΔI
out
)
2
+ (-m
cp
C
p,PA0
ΔI
in
)
2
!
1¡2

(2.8)
27

Δb

= ||
1
nÐI(I

e
- I
v
)
Δ0
out
1
2
+ |
0
out

2
I(I

e
- I
v
)
ΔÐ1
2
+ |
0
out
nÐI
2
(I

e
- I
v
)
ΔI1
2
+|
0
out
nÐI(I

e
- I
v
)
2
ΔI

e
1
2
+ |
0
out
nÐI(I

e
- I
v
)
2
ΔI
v
1
2
|
1¡2

(2.9)

ΔR = ||
(I

e
- I

cp
)
0
out
2
Δ0
out
+
2
+ |
1
0
out
ΔI

e
1
2
+ |
1
0
out
ΔI

cp
1
2
|
1¡2

(2.10)
The uncertainty of
PAO p,
C was estimated by Ghajar et al. to be 0.5% of the value. For
each steady state condition, 151 data points were collected from each sensing device
representing five minutes of data. Measured values were averaged and uncertainties were
calculated based on the fixed error of each instrument and the confidence interval for the
average at a confidence level of 0.95. A summary of the uncertainties for this experiment
can be found in Table 2.4. Details of the uncertainty analysis can be found in Appendix
B.
2.4. Results and Discussion
The purpose of this series of experiments was to determine the operating
characteristics of a titanium-water loop heat pipe subjected to changes in evaporator heat
input, compensation chamber heat input, and radial acceleration. Steady state and
transient temperature data were collected which provided insight into the fluid-thermal
behavior of the LHP. The raw data was reduced to obtain the evaporative heat transfer
coefficient, thermal resistance, and evaporator wall superheat in terms of the heat
transported and radial acceleration level. Quasi-steady phenomena and dry-out of the
LHP were observed and quantified in a performance map.
Figure 2.8 presents a typical stationary (a
z
= 1.0 g, a
r
= 0.0 g) cold-start test of the
LHP, which consisted of the following: With the LHP at ambient conditions, the
recirculating chiller in the low-temperature fluid loop was set to T
eg
= 35°C. Heat was
applied as a step function to the evaporator section (in this case, Q
in
= 600 W) while the
pump for the high-temperature fluid loop was simultaneously turned on (
cp
m&

= 0.0077
28
kg/s). The mass flow rate of the high-temperature fluid loop was maintained constant at
this value throughout this series of experiments to minimize the uncertainty associated
with the calorimetry of the cold plate. Figure 2.8(a) shows the transient temperature
response of the evaporator, vapor line, and calorimeter inlet and outlet. The temperatures
appear to become steady after approximately 6000 s. However, in order to determine
when steady state occurred the time rate of change of the temperatures was averaged over
15 min. intervals and plotted with respect to time as shown in Figure 2.8(b). It was
observed that dT/dt approached zero shortly after 6000 s, but for times greater than 6000
s, significant oscillations occurred. The oscillations in dT/dt were not apparent in the raw
temperature traces, but steady state was found to occur at approximately 18,000 s. This
was further demonstrated by calculating the thermal resistance and heat transfer
coefficient for this test at different times, as shown in Figure 2.8(c). This methodology
was used throughout testing to ensure that a repeatable steady state was reached.
Figure 2.9 also shows transient temperature traces during the Q
in
= 600 W test
described in the previous paragraph. In Figure 2.9(a), the evaporator temperature
(TC04) increased very quickly while the rest of the LHP did not react. After
approximately 60 s, the thermocouple located on the vapor line nearest to the exit of the
evaporator (TC08) suddenly increased. This was followed in turn by increases in
temperature reflected by the thermocouples located throughout the condenser section.
This shows the progression of the saturated vapor clearing the condenser section of
liquid, which was subsequently displaced into the evaporator section and the
compensation chamber via the bayonet tube. Figure 2.9(b) shows that the evaporator
temperature was significantly higher than the condenser temperatures, which led to a
relatively high value of thermal resistance, which will be discussed in detail below.
Figure 2.10 shows temperature traces in the condenser (TC09 through TC13) and
at the bayonet inlet (TC14). Each figure shows the transient temperature after the
stationary LHP reached steady state conditions at heat inputs ranging from 100 ≤ Q
in

600 W. In Figure 2.10(a), with Q
in
= 100 W, the liquid entering the bayonet tube was
highly subcooled at approximately 40°C. At this heat input level, the majority of the
condenser was flooded with subcooled liquid. In fact, only TC09 (condenser inlet)
indicated two-phase flow. Figure 2.10(b), with Q
in
= 200 W, was a unique case that is
29
described further in the following paragraph (Figure 2.11). Figure 2.10(c) to Figure
2.10(f) shows that the 2φ-1φ point progressed through the liquid line as heat input
increased until it reached the bayonet inlet. If the heat input at the evaporator is high
enough, saturated vapor will pass through the bayonet tube and reach the evaporator
section. This point represents a performance limit to the LHP operation because if vapor
enters the evaporator, the wick will dry out and the LHP will overheat.
Figure 2.11 shows the oscillatory behavior of the LHP for the heat input of Q
in
=
200 W. Initially, at t = 0, the evaporator temperatures (TC04, TC05, TC06, and TC07)
ranged from 66 to 68°C. The evaporator temperature nearest to the bayonet tube outlet
(TC07) was the lowest, which indicated that the subcooled liquid that entered the
evaporator tended to reduce the evaporator temperature at this point. The vapor line and
condenser temperatures (TC08 through TC13) ranged from 46 to 58°C. The vapor line
(TC08) was the highest, with the first three thermocouples in the condenser (TC09,
TC10, TC11) decreasing slightly. The vapor became saturated within the condenser, and
condensation formed on the interior walls of the tubing. From the point at which the
quality of the working fluid was x = 1 (saturated vapor) to where it reached x = 0
(saturated liquid), the temperature should have been constant, except for the fact that the
pressure dropped slightly due to viscous losses. This drop in the saturation pressure in
turn decreased the saturation temperature. Past TC11, the other condenser temperatures
(TC12, TC13) dropped significantly. This showed that the 2φ-1φ point, where x = 0,
occurred between TC11 and TC12. The working fluid after this point became a
subcooled liquid, where the temperature drop was due to sensible heat extraction by the
cold plates. Interestingly, at t = 0, the temperature at the bayonet inlet (TC14) was higher
than the outlet of the condenser. Under typical operation, this was not the case due to
convective losses from the liquid lines.
As time progressed from t = 0 (Figure 2.11(a)), several things occurred nearly
simultaneously. The evaporator thermocouple nearest to the vapor manifold (TC07)
suddenly decreased, which indicated movement of subcooled liquid from the exit of the
bayonet tube into the evaporator. The junction between the evaporator and the
compensation chamber (TC15) increased and then decreased in temperature over a
relatively short period. This was due to warm liquid in the evaporator section being
30
pushed through the grooves into the compensation chamber, followed by cooler liquid
from the bayonet tube exit. The inlet of the bayonet tube (TC14) decreased, and the two
thermocouples measuring the subcooled liquid in the condenser increased (TC12 and
TC13). Again, this was indicative of movement of the slug of liquid that existed from the
2φ-1φ point in the condenser to the meniscus within the grooves of the secondary wick
inside the evaporator section, as shown in Figure 2.2(a). The dramatic increase in the
condenser section (TC12) shows that the 2φ-1φ point moved from between TC11 and
TC12, across the TC12 location, and then between TC12 and TC13 as shown
schematically in Figure 2.11(b). In fact, TC12 increased to the saturated vapor
temperature existing within the first half of the condenser.
At approximately t = 80 s, the temperatures in the evaporator and the bayonet tube
inlet (TC14) started to increase, while the condenser temperatures TC12 and TC13
decreased. This behavior indicated that the liquid slug had reversed direction; i.e. the 2φ-
1φ point re-crossed thermocouple location TC12 in the condenser. The significant rise in
the bayonet inlet temperature TC14 shows that warm liquid originally in the evaporator
was now flooding back through the bayonet tube into the liquid line. This movement of
liquid out of the evaporator may be due to the sudden appearance of a vapor bubble
within the wick structure of the evaporator section which would tend to drive the heated
liquid in the evaporator in the opposite direction. As can be seen in Figure 2.11(a), the
period of the oscillation was approximately 150 s. This type of percolation is not typical
of a fully operational LHP, but is actually closer to the behavior of a pulsating heat pipe.
Discussion of flow reversal within LHPs in the literature was limited to startup and
shutdown operation. Douglas et al. (1999) discussed flow reversal in LHPs as a
phenomenon that occurred during startup and continued until the capillary pressure in the
secondary wick could no longer maintain the system pressure drop. Cimbala et al. (2004)
used neutron radiography to visualize LHP operation and observed flow reversal only
occurred when the heat input was reduced to Q
in
= 0 W. It was concluded that with no
heat input, convective and radiative heat transfer from the LHP to the ambient caused the
flow reversal. In general, flow reversal was not discussed as part of normal operation.
However, in the present experiment, flow reversal was found at some operating points.
31
Figure 2.12 shows the various steady state LHP temperatures versus transported
heat for the stationary case. The four evaporator temperatures in Figure 2.12(a) increased
monotonically with heat transported, but diverged from the vapor outlet temperature.
The behavior of the condenser temperatures with heat transported was slightly different,
as shown in Figure 2.12(b). At the lowest heat input value (Q
in
= 100 W), a significant
temperature drop was present between the inlet of the condenser (TC09) and the
thermocouples within the condenser. This shows that the 2φ-1φ point resided between
TC09 and TC10, which means that very little of the available condenser was being used
for two-phase condensation. This condition also shows that the liquid returning to the
evaporator section (TC14) was highly subcooled. As the heat input increased to Q
in
=
200 W, the temperatures measured at TC10 and TC11 rose to match that at TC09, which
means that the time averaged location of the 2φ-1φ point moved farther into the
condenser (between TC11 and TC12). At a heat input of Q
in
= 300 W, the 2φ-1φ point
traveled past the end of the condenser into the liquid lines such that all of the condenser
temperatures matched the evaporator outlet temperature (TC08). As the heat input
increased, the condenser temperatures continued to rise. However, the evaporator outlet
temperature increased at a faster rate, which is indicative of an increased superheat
penalty.
Figure 2.13 shows the thermal performance of the stationary LHP for heat inputs
ranging from Q
in
= 100 to 600 W. The evaporative heat transfer coefficient, Figure
2.13(a), decreased monotonically with transported heat. This behavior was controlled by
the slope of the average evaporator temperature versus that of the evaporator outlet, as
shown in Figure 2.12(a). The temperature difference ( )
v e
T T − defined in equation (2.6)
increased more rapidly than Q
out
, which resulted in an overall decrease in h . As dry-out
was approached, more of the wick in the evaporator section was depleted of liquid, which
tended to increase the evaporator temperature. The thermal resistance of the stationary
LHP versus heat transported is presented in Figure 2.13(b), where it is seen to decrease,
reach a minimum, and then increase. At low power inputs, the relatively large
temperature drop defined by equation (2.7), ( )
cp e
T T − , drives the thermal resistance to a
high value. This temperature drop was a result of the fact that most of the condenser
32
section was flooded by subcooled liquid which was close to the cold plate temperature.
As the 2φ-1φ point moved through and then exited the condenser, the temperature drop
decreased with transported heat, which decreased the thermal resistance. The minimum
R corresponds to the point in Figure 2.12(b) where the 2φ-1φ point just exited the
condenser. Past this point, the evaporator section increased in temperature more rapidly
than the condenser section, which resulted in the thermal resistance increasing with
transported heat. The wall superheat, defined as the difference between the average
evaporator temperature and the temperature of the evaporator/compensation chamber
junction, was found to monotonically increase with an increasing amount of transported
heat. With respect to the evaporative heat transfer coefficient, thermal resistance, and
wall superheat, no notable difference was observed between starting the LHP while the
unit was at ambient temperature versus a step change in the evaporator heat input from a
lower to higher value or a higher to lower value. A summary of the stationary steady
state data points and the path to reach steady state can be seen in Table 2.5.
Figure 2.14 shows the operating characteristics and performance of the stationary
LHP for an evaporator heat input of Q
in
= 500 W while varying the compensation
chamber heat input from Q
cc
= 0 to 50 W. For this particular test, the LHP was allowed
to achieve steady state conditions for the given evaporator heat input, after which the
compensation chamber heat input was incremented in steps of 5 W. In Figure 2.14(a)
and Figure 2.14(b), for Q
cc
= 0 W, the evaporator temperatures were relatively uniform,
where the vapor exiting the evaporator was slightly superheated and the 2φ-1φ point was
out of the condenser. When a small amount of heat was input to the system through the
compensation chamber (Q
cc
= 5 W), the evaporator temperatures and the evaporator exit
temperature both decreased while the condenser temperatures remained constant. This
trend continued until approximately Q
cc
= 15 W, at which point the evaporator
temperature leveled off, the evaporator exit temperature decreased to the saturation
temperature within the condenser, and the condenser outlet temperature dropped below
the saturation temperature. The decrease in the average evaporator temperature
significantly affected the evaporative heat transfer coefficient and the thermal resistance,
as shown in Figure 2.14(c). In fact, h increased by 68% with an increase in the overall
heat input of only 3%. The drop in the condenser outlet temperature indicated that the
33
2φ-1φ point moved from the liquid line into the condenser section. For Q
cc
≥ 20 W, the
evaporator temperatures increased and the 2φ-1φ point continued to move toward the
evaporator which resulted in an increase in the thermal resistance. Ku (1999) indicated
that operating the compensation chamber at a higher temperature by using an external
heater in effect increases the amount of subcooling in the condenser and liquid return
line. According to Ku (1999), this subcooling is necessary to balance the additional heat
input and results in underutilizing the condenser and a degradation of the thermal
conductance. In the present experiment, this conclusion held true for Q
cc
≥ 20 W, as the
amount of subcooling increased the thermal resistance and decreased the evaporative heat
transfer coefficient by way of an increased superheat penalty. This did not hold true for
Q
cc
< 20 W. When the LHP operated at Q
in
= 500 W, the 2φ-1φ interface was located in
the liquid return line. Increasing the heat input to the compensation chamber moved the
2φ-1φ interface to the condenser outlet at Q
cc
= 15 W. Operation at this point maximized
the amount of heat transfer due to condensation with the added benefit of cooler liquid in
the compensation chamber and evaporator which decreased the thermal resistance and
increased the evaporative heat transfer coefficient.
Also of interest is the temperature increase at the bayonet inlet (TC14) starting
when Q
cc
= 35 W seen in Figure 2.14(b). Figure 2.15 shows the transient temperature
traces of the condenser, bayonet tube, and evaporator/compensation chamber junction for
Q
cc
= 25 to 50 W. In Figure 2.15(a) and Figure 2.15(b), with Q
cc
= 25 and 30 W,
subcooled liquid moved through the bayonet inlet as seen in typical operation. In Figure
2.15(c) through Figure 2.15(f), with Q
cc
= 35 through 50 W, a sudden increase in
temperature at the bayonet inlet (TC14) showed that flow reversal occurred in the
evaporator section. This was similar to the oscillating phenomena described for Q
in
=
200 W except that the liquid-vapor meniscus in the secondary wick was driven backward
by the elevated vapor pressure within the compensation chamber, which was due to the
heat input at the shell of the compensation chamber. In addition, the temperature of the
evaporator/compensation chamber interface (TC15) did not vary appreciably, which was
different than that seen at Q
in
= 200 W. The liquid-vapor meniscus moved backward due
to the increased pressure within the compensation chamber until a point at which the
pressure was balanced. Forward flow then resumed and heat was lost through the liquid
34
line, shown by the slow decrease in temperature at the bayonet inlet (TC14) to the
temperature of the subcooled liquid at the condenser outlet (TC13). A summary of the
steady state data points for Q
cc
= 0 to 50 W can be seen in Table 2.6.
To further explore compensation chamber heat input and heat loss to the ambient,
Table 2.7 shows the effect of operating the stationary LHP for Q
in
= 500 W with the
compensation chamber uninsulated, insulated, temperature controlled to T
cc
= 72.8°C via
simultaneous heat input to the compensation chamber (Q
cc
= 20 W) and evaporator, and
preconditioning the temperature to T
cc
= 72.8°C via heat input (Q
cc
= 100 W reduced to
Q
cc
= 20 W) prior to heat input to the evaporator. For this series of tests, thermocouple
TC15 was relocated to the top side of the compensation chamber to directly monitor its
operating temperature. It was observed that the average evaporator temperature
increased, the evaporative heat transfer coefficient decreased, and the thermal resistance
increased in the uninsulated state when compared to the insulated case. The uninsulated
compensation chamber operated at a temperature 3.6°C lower than the insulated case as
expected due to free convection and radiative heat loss. To estimate the amount of heat
loss to the ambient in the uninsulated case, the exposed compensation chamber was
modeled as a cylinder in free convection with radiation. The average Nusselt number for
free convection was given by (Incropera and DeWitt, 2002)

Nu

D
= {u.6u +
u.S87Ra
D
1¡6
|1 + (u.SS9¡Pi)
9¡16
]
8¡27
¦
2

(2.11)
where

Ra
D
=
g[(I
s
-I
«

3
vo

(2.12)
with air properties evaluated at the average temperature of the freestream and the surface.
The average heat transfer coefficient was given by

b

=
k
Ð
Nu

D

(2.13)
The total heat loss per unit length from the compensation chamber was given by
o
tot
i
= o
conv
i
+ o
rad
i
= b

nÐ(I
s
- I
«
) + enÐo(I
s
4
- I
surr
4
) (2.14)
35
The emissivity of grade 2 titanium used in this calculation was ε = 0.3 (Boyer et al.,
1994). The heat loss from the compensation chamber for the uninsulated case was found
to be Q
cc
= -6.2 W. When the uninsulated case was included with the previous
compensation chamber heat input data, it was found that evaporative heat transfer
coefficient and thermal resistance followed the trends shown in Figure 2.14(c). These
results were expected since removing the insulation from the compensation chamber in
effect provided cooling, which moved the 2φ-1φ point away from the condenser. The
average evaporator and cold plate temperatures were significantly different which was
likely due a 10°C higher ambient temperature during the test involving the variation of
compensation chamber heat input from Q
cc
= 0 to 50 W. As a result, for this particular
case, it was advantageous to operate the LHP compensation chamber insulated for
improved performance. For controlling the temperature of the compensation chamber,
the evaporative heat transfer coefficient, thermal resistance, and operating temperatures
were nearly identical between simultaneous compensation chamber and evaporator heat
input startup and compensation chamber temperature preconditioning, demonstrating that
the startup procedure had no impact on steady state conditions. However,
preconditioning the compensation chamber required approximately one hour less time to
reach steady state conditions over the simultaneous heat input startup.
Figure 2.16 presents transient LHP temperatures for a typical test at elevated
acceleration (a
z
= 1.0 g, a
r
> 0 g). With the LHP at ambient conditions, the recirculating
chiller in the low temperature loop was set to T
eg
= 35°C. Heat was applied as a step
function (in this case, Q
in
= 600 W) while simultaneously starting the pump for the high-
temperature loop (
cp
m&

= 0.0077 kg/s). In addition, the radial acceleration was increased
to a
r
= 0.1 g, which was a nominally small value to prevent damage to the power slip
rings (Figure 2.16(a)). The LHP was allowed to achieve steady state conditions at a
r
=
0.1 g, indicated by dT/dt (Figure 2.16(c)) decreasing to below the threshold of 0.01
K/min, then the acceleration was increased to the next desired radial acceleration value
(in this case, a
r
= 10.0 g). The LHP was again allowed to achieve steady state conditions
at the given acceleration (Figure 2.16(b)), then the acceleration was reduced back to a
r
=
0.1 g for a minimum of thirty min. If another elevated acceleration was desired, steady
state at a
r
= 0.1 g was reached before increasing the acceleration level. When the
36
acceleration was increased to a
r
= 10.0 g at t = 15,000 s in Figure 2.16(b), the average
evaporator temperature increased by 11°C. The 2φ-1φ point moved to the condenser
outlet from the liquid line with increasing acceleration, indicated by the small oscillations
in temperature at the TC13. The amount of subcooling increased overall as indicated by
the decrease in temperature at the bayonet inlet (TC14). These phenomena may be due in
part to fluid redistribution in the LHP and is discussed in the following paragraph.
As the rotational velocity of the centrifuge increased, the resultant acceleration
vector magnitude and direction changed (Figure 2.17(a)) which influenced the
distribution of fluid in the LHP. Subcooled liquid entering the primary wick of the
evaporator was forced to the outboard side of the evaporator body, opposite of the heat
source, and perhaps leading to a partial dry-out of the wick (Figure 2.17(b)). The
elevated acceleration also hindered the ability of the secondary wick in the compensation
chamber to supply the evaporator with liquid due to pooling. In the condenser, pooling
occurred in the bends of the condenser coil, again due to the acceleration gradient.
Depending on the acceleration vector direction, this pooling could either open or close
the passage to vapor flow (Figure 2.17(c)). All of these phenomena are a result of
centrifuge testing. Due to the short radius, strong acceleration gradients occur that could
have advantageous or adverse effects on the LHP operation. Operation in an aircraft
environment, with significantly larger radii during turns, will provide a more uniform
acceleration gradient across the LHP and potentially yield different temperature profiles,
evaporative heat transfer coefficients, and thermal resistances.
Figure 2.18 shows the thermal performance of the LHP for radial accelerations
ranging from a
r
= 0.1 to 10.0 g and heat inputs ranging from Q
in
= 100 to 600 W. The
evaporative heat transfer coefficient, Figure 2.18(a), again decreased with transported
heat, similar to the trend in Figure 2.13(a) for the stationary LHP. The thermal resistance
of the LHP (Figure 2.18(b)), was found to decrease to a minimum, then increase, again
similar to the stationary test results shown in Figure 2.13(b). In fact, when combining the
stationary and elevated acceleration test data, it was found that the evaporative heat
transfer coefficient and thermal resistance data were in close agreement with each other,
regardless of the radial acceleration. This indicated that bench top testing of the LHP was
a reliable method for determining the evaporative heat transfer coefficient and thermal
37
resistance of a LHP in an elevated acceleration environment. However, it will be shown
that this was not true with respect to finding the dry-out limit. The wall superheat (Figure
2.18(c)) was higher at elevated accelerations when compared to a
r
= 0 g. This was
possibly due to fluid redistribution in the evaporator forcing liquid away from the heater.
A summary of all elevated acceleration steady state data points can be seen in Table 2.8.
Ku et al. (2000a and 2000b) observed that radial acceleration changed the fluid
distribution throughout the LHP which changed operating temperatures and that
acceleration could either increase or decrease LHP operating temperatures. In addition,
temperature overshoots were observed for mid-range heat inputs and the wall superheat,
defined as the temperature difference between the evaporator and compensation chamber,
was independent of heat input and acceleration during startup. In the present elevated
acceleration tests, it was observed that the accelerating force changed the fluid
distribution within the LHP, causing the operating temperatures to change. However, in
all instances, it was observed that elevated acceleration forces increased operating
temperatures over those at a
r
= 0.1 g. Significant temperature overshoots were not
observed in any of the elevated acceleration tests.
Figure 2.19 shows the transient response of the LHP during a series of dry-out
events. Dry-out was indicated by a steady increase in the evaporator temperature and a
decrease in the heat extracted by the calorimeter Q
out
. In addition, the position of the 2φ-
1φ point in the condenser moved toward the evaporator as indicated by a sequential
decrease in the condenser temperatures. This occurred because the evaporator no longer
generated a sufficient flow of vapor which changed the operating point of the LHP. In
Figure 2.19(a), the LHP reached steady state while rotating slowly at a
r
= 0.1 g and Q
in
=
400 W. The rotational speed of the centrifuge table was increased until the radial
acceleration reached a
r
= 8.0 g at t = 300 s. After the evaporator temperature TC06
reached T
e,max
= 150°C, the radial acceleration was reduced back to a
r
= 0.1 g. At this
time, the evaporator temperature continued to increase, but then leveled off and then
decreased back to nearly the same temperature as the previous steady state. In fact, all of
the LHP temperatures returned to within 1°C of the original steady state except for TC13
(condenser outlet), which returned to within 4°C of the previous steady state. This larger
temperature difference in TC13 was attributable to a slight change in the location of the
38
2φ-1φ point in the condenser. This recovery behavior shows that the LHP was capable of
repriming at the end of an acceleration burst even if the heat input remained constant. In
Figure 2.19(b) and Figure 2.19(c), the radial acceleration was again increased from a
r
=
0.1 to 8.0 g with the same heat input (Q
in
= 400 W). In fact, all of the experiments
presented in Figure 2.19 were performed sequentially. In Figure 2.19(b), the evaporator
temperature TC06 was allowed to reach T
e,max
= 175°C before decreasing the radial
acceleration to a
r
= 0.1 g, and in Figure 2.19(c), the evaporator temperature TC06 reached
T
e,max
= 200°C before decelerating. In each instance, the evaporator temperature
continued to increase, reached a maximum, and then decreased to the original steady
state. However, the intensity of dry-out did seem to have an impact on the ability of the
LHP to reprime. In Figure 2.19(c), with a maximum evaporator temperature at
deceleration of T
e,max
= 200°C, the evaporator temperature reached two maximums before
finally decreasing back to the previous steady state, whereas in Figure 2.19(a) and Figure
2.19(b), the maximum evaporator temperatures reached a peak and then monotonically
decreased. This indicated that if the evaporator temperature were much higher than
200°C, the LHP may not have recovered, which would have required that the heat input
be reduced to zero.
Figure 2.20 shows the temperature traces associated with the test at Q
in
= 200 W
and a
r
= 0.1 and a
r
= 4.0 g. Following the previously mentioned startup procedures, the
LHP reached a quasi-steady state while the centrifuge table rotated slowly for a
r
= 0.1 g,
as shown in Figure 2.20(a). Similar to the stationary case at this heat input, the LHP
temperatures oscillated, showing that the heat pipe was operating during reversals in the
liquid flow due to the liquid-vapor meniscus in the secondary wick moving back and
forth. Overall, the temperatures shown in Figure 2.20(a) were quite close to the case
shown in Figure 2.11(b), as presented in Table 2.9. In addition, the period of the
oscillation of the a
r
= 0.1 g case was nearly identical to the a
r
= 0 g case (approximately
175 s). The only significant differences in the independent variables between the two
tests were the ambient temperature (ΔT
amb
= 5.3°C), and the relatively small value of the
radial acceleration. Of note, however, was the location of the 2φ-1φ point in the
condenser: For a
r
= 0 g, this point resided close to TC12, whereas for the case in which
a
r
= 0.1 g, the 2φ-1φ point was near TC10. The linear distance between these two points
39
was approximately 143 cm. While it was impossible to know the exact location of the
2φ-1φ point due to the coarse resolution of the thermocouples in the condenser, it was
obvious that the location had changed significantly between the two cases. In addition,
the evaporative heat transfer coefficient decreased and the thermal resistance increased
from a
r
= 0 to 0.1 g. It was believed that this small value of the radial acceleration
resulted in a significant change in location of the 2φ-1φ point due to the pooling of liquid.
After achieving the quasi-steady state at a
r
= 0.1 g, the radial acceleration was
increased to a
r
= 4.0 g, and the LHP again reached a quasi-steady state, as shown in
Figure 2.20(b). The average evaporator temperature increased by more than 30°C, and
the temperature in the bayonet inlet ranged from 37 ≤ T
bayonet inlet
≤ 70°C, which was a
much larger range than that for a
r
= 0.1 g. Oscillations were again seen at this
acceleration level, but the period of the oscillations increased to approximately 350 s.
This may be due to the distance that the meniscus travelled within the evaporator, which
resulted in wider swings in the evaporator temperatures and significant oscillations of the
cold plate outlet temperature, which was nearly steady in the a
r
= 0.1 g case.
Figure 2.21 shows the steady state performance map for the LHP relating radial
acceleration and heat transported for a
r
= 2.0 to 10.0 g and Q
in
= 100 to 600 W. It was
observed that dry-out conditions occurred at varying radial accelerations for Q
in
= 100 to
400 W. Dry-out conditions were not observed through a
r
= 10.0 g at Q
in
= 500 and 600
W. Quasi-steady state conditions were observed at Q
in
= 200 W and a
r
= 4.0 g. This
demonstrated that bench-top testing cannot be used to determine the dry-out limit with
respect to elevated acceleration.
2.5. Conclusions
The effect of changes in evaporator heat input, compensation chamber heat input,
and radial acceleration on a titanium-water loop heat pipe were investigated for Q
in
= 100
to 600 W, Q
cc
= 0 to 50 W, and a
r
= 0.0 to 10.0 g. A transient temperature rate of change
method was developed to ensure steady state had been achieved. For evaporator heat
input Q
in
= 100 to 600 W, it was observed that the evaporative heat transfer coefficient
decreased monotonically, thermal resistance decreased to a minimum, then increased
over the same range, and wall superheat monotonically increased. Flow reversal was
observed at Q
in
= 200 W due to vapor bubble generation in the evaporator.
40
When examining the effect of compensation chamber heat input for Q
in
= 500 W,
it was found that the average evaporator temperatures dropped by 15°C and evaporative
heat transfer coefficient improved by 68% with only a 3% increase in heat load. These
results differ from Ku (1999) in that an improvement was observed for compensation
chamber heat input up to the point where subcooling was occurring in the condenser.
Flow reversal was observed starting at Q
cc
= 35 W due to the increased pressure in the
compensation chamber driving the liquid/vapor meniscus backwards. Operating the LHP
compensation chamber uninsulated at Q
in
= 500 W was found to degrade the LHP
performance for this particular case and preconditioning the compensation chamber
temperature prior to evaporator heat input shortened the time to steady state.
When examining the effect of radial acceleration, it was found that dry-out
conditions occurred more readily at lower heat inputs (Q
in
= 100 to 400 W) than at higher
heat inputs (Q
in
= 500 to 600 W). The LHP was found to be able to reprime after an
acceleration event that caused dry-out without the heat input being reduced to zero. It
was also observed that radial acceleration had little effect on the evaporative heat transfer
coefficient and thermal resistance of the LHP. Wall superheat was found to be higher at
steady state elevated accelerations when compared to a
r
= 0 g. This led to conclusion that
bench top testing of the LHP is a reliable method for determining the evaporative heat
transfer coefficient and thermal resistance of a LHP in an elevated acceleration
environment induced by a centrifuge table, but is not sufficient for determining the wall
superheat and dry-out limit. These results may or may not actually occur in an aircraft
environment as centrifuge operation can induce artifacts in the data due to the short
radius of operation.
2.6. Future Work
Experimentation in this thesis has been conducted using strict regimented
procedures for repeatability and to allow comparisons across data sets. Typical operation
of LHPs is not along regimented schedules but in transient environments where heat
sources, heat sinks, and accelerating forces are varying with respect to time. As such,
experimentation should be conducted using transient profiles to more closely mimic
actual aircraft environments. Experimentation should also be conducted with tighter
41
control over the compensation chamber temperature to enhance repeatability and tailor
operation to specific heat sources and sinks.
42

Figure 2.1. Loop heat pipe operation. Adapted and reprinted with permission from
AIAA (Hoang and Ku, 2003).

Liquid Line Vapor Line
Heat Exchanger
Compensation
Chamber
Evaporator
Bayonet Tube
Liquid-Phase
(Subcooler)
Vapor-Phase
(Condenser)
Primary Wick

out
Q

out
Q
in
Q
cc
Q
2
Q
1
Q
Liquid
43

Figure 2.2. Evaporator schematic: (a) Side view; (b) Cross-sectional view. Adapted and
reprinted with permission from AIAA (Hoang and Ku, 2003).

Liquid from
Compensation
Chamber
Secondary
Wick
Primary Wick
Vapor to
Compensation
Chamber
Liquid from
Liquid Line
Non-wick
Flow Path
Bayonet Tube
Q
1
Q
2
Liquid
Flow
Bayonet Tube
Non-wick
Flow Path
Secondary
Wick
Primary
Wick
Q
1
Q
2
(a)
(b)
Meniscus
44

Figure 2.3. Schematic of Centrifuge Table Test Bed.

Gear
Box
Cool
Bath
TV
Monitor
Data
Acquisition
& Control
0 - 1 kW
DC Power
Supplies
Hydraulic Rotary
Coupling
Instrumentation
Slip Rings
TV Camera
Counterbalance
Weights
Thermocouple
Signal
Conditioner
Power
Slip Rings
Centrifuge
Table
20 HP
DC Motor
Isolation
Transformer
Motor
Controller
Centrifuge Table Room Control Room
120
VAC
Computer w/
LabVIEW
Triaxial
Accelerometer
Test Article
45

Figure 2.4. Titanium-water loop heat pipe test article as delivered.

46

(a)

(b)
Figure 2.5. Thermocouple locations on the LHP: (a) Locations of thermocouples TC04
through TC15 across the LHP; (b) Locations of TC04 through TC07 within the
evaporator.

4 5 6 7 8 15 14 9 10 11 12 13
4 5 6
7 Q
in
47

(a)

(b)

(c)

(d)
Figure 2.6. Mounting of loop heat pipe to centrifuge table, front and top views: (a)
Evaporator and compensation chamber: (b) Transport lines; (c) Condenser with cold
plate; (d) Complete loop heat pipe.

48

(a)

(b)

(c)
Figure 2.7. High temperature fluid loop: (a) Schematic; (b) Reservoir, pump, filter,
flowmeter, TC03, and liquid/liquid heat exchanger; (c) Cold plate, TC00, and TC01.


Cold Plate
TC00
TC01
Liquid-Liquid
Heat Exchanger
Gear Pump
Flow meter
LHP Condenser
Reservoir
Recirculating
chiller
TC03
LHP Evaporator
Electric Heater
PAO Heater
Filter
Reservoir
Pump
Filter
TC03
L/L HX
Flow meter
Cold plate
TC00
TC01
49



Figure 2.8. Use of a cold-start test to determine when steady state occurred for the
stationary LHP (Q
in
= 600 W, Q
cc
= 0 W, a
r
= 0 g,
cp
m& = 0.0077 kg/s,
cp
T = 67.7°C, T
amb

= 38.1°C): (a) Transient temperature traces; (b) Transient rate of change of temperatures;
(c) Transient thermal resistance and evaporative heat transfer coefficient.


0
20
40
60
80
100
120
140
160
T

(
C
)
TC00 - CP In
TC01 - CP Out
TC04 - Evap 1
TC08 - Evap Out
TC09 - Cond In
(a)
(TC00) (TC01)
(TC04)
(TC08) (TC09)
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
d
T
/
d
t

(
K
/
m
i
n
)
TC00 - CP In
TC01 - CP Out
TC04 - Evap 1
TC05 - Evap 2
TC06 - Evap 3
TC07 - Evap 4
TC08 - Evap Out
TC09 - Cond In
(b)
0
0.05
0.1
0.15
0.2
0.25
0.3
0
100
200
300
400
500
600
700
800
900
0 5000 10000 15000 20000 25000 30000
R

(
K
/
W
)
h

(
W
/
m
2
-
K
)
t (s)
h
R
(c)
50


Figure 2.9. Transient startup of the stationary LHP (Q
in
= 600 W, Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s,
cp
T = 67.7°C, T
amb
= 38.1°C): (a) Initial startup; (b) Complete startup
until steady state.

0
20
40
60
80
100
120
0 200 400 600 800 1000 1200
T

(
C
)
t (s)
TC04 - Evap 1
TC08 - Evap Out
TC09 - Cond In
TC10 - Cond 1
TC11 - Cond 2
TC12 - Cond 3
TC13 - Cond Out
TC14 - Bayonet In
TC15 - Evap/CC
(a)
(TC04)
(TC08) (TC09) (TC10)
(TC11) (TC12)
(TC13) (TC14)
(TC15)
0
20
40
60
80
100
120
140
160
0 5000 10000 15000 20000 25000 30000
T

(
C
)
t (s)
(b)
(TC04) (TC08) (TC09-TC15)
51






Figure 2.10. Transient temperature profiles in the condenser and bayonet tube of the
stationary LHP (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T ≤ 71.6°C, 31.7 ≤ T
amb

≤ 38.1°C): (a) Q
in
= 100 W; (b) Q
in
= 200 W; (c) Q
in
= 300 W; (d) Q
in
= 400 W; (e) Q
in
=
500 W; (f) Q
in
= 600 W.
35
40
45
50
55
60
65
T

(
C
)
TC09 - Cond In TC10 - Cond 1
TC11 - Cond 2 TC12 - Cond 3
TC13 - Cond 4 TC14 - Bayonet In
(a) (TC09)
(TC10) (TC11) (TC12-TC14)
35
40
45
50
55
60
65
T

(
C
)
(b)
(TC09- TC11)
(TC14) (TC12) (TC13)
45
50
55
60
65
70
75
T

(
C
)
(c)
(TC09- TC13) (TC14)
55
60
65
70
75
80
85
T

(
C
)
(d)
(TC09- TC13) (TC14)
70
75
80
85
90
95
100
T

(
C
)
(e)
(TC09- TC13) (TC14)
80
85
90
95
100
105
110
T
(
C
)
(f)
(TC09- TC13) (TC14)
52


Figure 2.11. Transient temperature profiles of the stationary LHP for Q
in
= 200 W (Q
cc
=
0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s,
cp
T = 46.1°C, T
amb
= 31.7°C): (a) Transient
temperature profiles; (b) 2φ-1φ point oscillation in the condenser.

40
45
50
55
60
65
70
0 50 100 150 200 250 300
T

(
C
)
t (s)
(TC14) (TC12) (TC13) (TC15)
(TC04) (TC06) (TC07) (TC05)
(TC08) (TC10) (TC11) (TC09)
Evaporator Condenser
Vapor Line Liquid Line
Evap / CC Cold Plate
(TC00) (TC01)
(a)
9
10
11
12
13
0 ≤ t ≤ 30 30 ≤ t ≤ 80 t = 80 80 ≤ t ≤ 160 t = 160


C
o
n
d
e
n
s
e
r
(b)
53


Figure 2.12. Steady state temperature distribution versus transported heat for the
stationary LHP (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T ≤ 71.6°C, 31.7 ≤ T
amb

≤ 38.1°C): (a) Evaporator section; (b) Condenser section.

0
20
40
60
80
100
120
140
160
T

(
C
)
TC04 Evap 1
TC05 Evap 2
TC06 Evap 3
TC07 Evap 4
TC08 Evap Out
TC15 Evap/CC
(a)
0
20
40
60
80
100
120
0 100 200 300 400 500 600
T

(
C
)
Q
out
(W)
TC08 Evap Out
TC09 Cond In
TC10 Cond 1
TC11 Cond 2
TC12 Cond 3
TC13 Cond Out
TC14 Bayonet In
(b)
54



Figure 2.13. Steady state performance characteristics of the stationary LHP versus
transported heat (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T ≤ 67.7°C, 27.6 ≤ T
amb

≤ 38.7°C): (a) Evaporative heat transfer coefficient; (b) Thermal resistance; (c) Wall
superheat

0
500
1000
1500
2000
2500
h

(
W
/
m
2
-
K
)
(a)
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
R

(
K
/
W
)
(b)
0
10
20
30
40
50
60
0 100 200 300 400 500 600
Δ
T
s
h
(
K
)
Q
out
(W)
(c)
55




Figure 2.14. Steady state performance characteristics of the stationary LHP versus
compensation chamber heat input (Q
in
= 500 W, a
r
= 0 g,
cp
m& = 0.0077 kg/s, 63.4 ≤
cp
T ≤
64.8°C, 36.1 ≤ T
amb
≤ 38.1°C): (a) Evaporator temperatures; (b) Condenser temperatures;
(c) Evaporative heat transfer coefficient and thermal resistance.

80
90
100
110
120
130
140
T

(
C
)
TC04 - Evap 1
TC05 - Evap 2
TC06 - Evap 3
TC07 - Evap 4
TC08 - Evap Out
TC15 - Evap/CC
(a)
60
70
80
90
100
110
T

(
C
)
TC08 - Evap Out
TC09 - Cond In
TC10 - Cond 1
TC11 - Cond 2
TC12 - Cond 3
TC13 - Cond Out
TC14 - Bayonet In
(b)
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0
200
400
600
800
1000
1200
1400
1600
1800
2000
-10 0 10 20 30 40 50 60
R

(
K
/
W
)
h

(
W
/
m
2
-
K
)
Q
cc
(W)
h
h - Uninsulated
R
R - Uninsulated
(c)
56






Figure 2.15. Transient temperature profiles in the condenser and bayonet tube of the
stationary LHP for Q
cc
= 25 to 50 W (Q
in
= 500 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 63.4 ≤
cp
T ≤ 64.8°C, 36.1 ≤ T
amb
≤ 38.1°C): (a) Q
cc
= 25 W; (b) Q
cc
= 30 W; (c) Q
cc
= 35 W; (d)
Q
cc
= 40 W; (e) Q
cc
= 45 W; (f) Q
cc
= 50 W.

60
70
80
90
100
110
T

(
C
)
TC09 - Cond In TC10 - Cond 1
TC11 - Cond 2 TC12 - Cond 3
TC13 - Cond Out TC14 - Bayonet In
(TC09- TC12) (TC13)
(a)
(TC14) (TC15)
60
70
80
90
100
110
T

(
C
)
(TC09- TC12) (TC13)
(b)
(TC14) (TC15)
60
70
80
90
100
110
T

(
C
)
(TC09- TC11) (TC13)
(c)
(TC14) (TC15) (TC12)
60
70
80
90
100
110
T

(
C
)
(TC09- TC11)
(TC13)
(d)
(TC14) (TC15) (TC12)
60
70
80
90
100
110
T

(
C
)
(TC09- TC10) (TC13)
(e)
(TC14) (TC15) (TC12) (TC11)
60
70
80
90
100
110
0 50 100 150 200 250 300
T

(
C
)
t (s)
(TC09- TC10) (TC13)
(f)
(TC14) (TC15) (TC11) (TC12)
57



Figure 2.16. Transient temperature traces of the LHP at elevated acceleration (Q
in
= 600
W, Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 55.2 ≤
cp
T ≤ 59.7°C, 27.9 ≤ T
amb
≤ 30.1°C): (a) a
r
=
0.1 g startup phase; (b) Transition to and steady state at a
r
= 10.0 g; (c) Transient rate of
change of temperatures.
20
40
60
80
100
120
140
0 3000 6000 9000 12000 15000
T

(
C
)
t (s)
TC04 - Evap 1
TC08 - Evap Out
TC09 - Cond In
TC10 - Cond 1
TC11 - Cond 2
TC12 - Cond 3
TC13 - Cond Out
TC14 - Bayonet In
TC15 - Evap/CC
(a)
(TC04)
(TC08) (TC09-TC13)
(TC14) (TC15)
20
40
60
80
100
120
140
15000 18000 21000 24000 27000
T

(
C
)
t (s)
(b)
(TC04)
(TC13)
(TC09-TC12)
(TC14) (TC15)
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0 6000 12000 18000 24000
d
T
/
d
t

(
K
/
m
i
n
)
t (s)
TC00 - CP In
TC01 - CP Out
TC04 - Evap 1
TC05 - Evap 2
TC06 - Evap 3
TC07 - Evap 4
TC08 - Evap Out
TC09 - Cond In
(c)
58

(a)

(b)

(c)
Figure 2.17. Effect of resultant acceleration vector direction on fluid distribution within
the LHP: (a) Resultant acceleration vector orientation versus radial acceleration; (b)
Liquid pooling in the evaporator, compensation chamber, and condenser under elevated
acceleration (to scale, top view); (c) Liquid pooling in the condenser bends.
0
10
20
30
40
50
60
70
80
90
0 2 4 6 8 10
θ
(
a
r
c

d
e
g
r
e
e
s
)
a
r
(g)
θ
a
z
a
r
|a|
V
L Liquid
Pooling
59



Figure 2.18. Steady state performance characteristics of the LHP versus transported heat
at stationary and elevated acceleration (Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 37.2 ≤
cp
T ≤
67.7°C, 25.1 ≤ T
amb
≤ 38.7°C): (a) Evaporative heat transfer coefficient; (b) Thermal
resistance; (c) Wall superheat.

0
200
400
600
800
1000
1200
1400
1600
1800
2000
h

(
W
/
m
2
-
K
)
(a)
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
R

(
K
/
W
)
(b)
0
10
20
30
40
50
60
70
80
90
0 100 200 300 400 500 600
Δ
T
s
h
(
K
)
Q
out
(W)
a = 0 g
a = 0.1 g
a = 2.0 g
a = 4.0 g
a = 6.0 g
a = 8.0 g
a = 10.0 g
r
r
r
r
r
r
(c)
r
60



Figure 2.19. Transient temperature traces of the LHP at elevated acceleration showing
dry-out behavior (Q
in
= 400 W, Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 37.2 ≤
cp
T ≤ 59.7°C, T
amb

= 28.0°C): (a) T
e,max
= 150°C; (b) T
e,max
= 175°C; (c) T
e,max
= 200°C.
0
50
100
150
200
250
300
350
400
450
500
25
45
65
85
105
125
145
165
185
205
225
0 1000 2000 3000 4000 5000 6000
Q
o
u
t
(
W
)
T

(
C
)
t (s)
TC06 - Evap 3
TC09 - Cond In
TC10 - Cond 1
TC11 - Cond 2
TC12 - Cond 3
TC13 - Cond Out
Qout
(TC11) (TC10) (TC12) (TC09) (TC13)
(Q
out
)
(TC06)
(a)
8 g 0.1 g
0
50
100
150
200
250
300
350
400
450
500
25
45
65
85
105
125
145
165
185
205
225
6000 7000 8000 9000 10000 11000 12000 13000
Q
o
u
t
(
W
)
T

(
C
)
t (s)
(TC11) (TC10) (TC12) (TC09) (TC13)
(Q
out
)
(TC06)
(b)
8 g 0.1 g 0.1 g
0
50
100
150
200
250
300
350
400
450
500
25
45
65
85
105
125
145
165
185
205
225
13000 15000 17000 19000 21000
Q
o
u
t
(
W
)
T

(
C
)
t (s)
(TC11) (TC10) (TC12) (TC09) (TC13)
(Q
out
)
(TC06)
(c)
8 g 0.1 g 0.1 g
61


Figure 2.20. Quasi-steady state temperature traces of the LHP and cold plate at elevated
acceleration for Q
in
= 200 W (Q
cc
= 0 W,
cp
m& = 0.0077 kg/s,
cp
T = 41.9°C, T
amb
=
26.4°C): (a) Transient temperature trace at a
r
= 0.1 g and t = 13834 s; (b) Transient
temperature trace at a
r
= 4.0 g and t = 31240 s.
35
40
45
50
55
60
65
70
75
0 50 100 150 200 250 300
T

(
C
)
t (s)
(TC14) (TC12) (TC13)
(TC15)
(TC04) (TC06) (TC07) (TC05)
(TC08)
(TC10) (TC11)
(TC09)
Evaporator Condenser
Vapor Line Liquid Line
Evap / CC Cold Plate
(TC00) (TC01)
(a)
30
40
50
60
70
80
90
100
110
0 100 200 300 400 500 600 700
T

(
C
)
t (s)
(TC14) (TC12) (TC13)
(TC15)
(TC04) (TC06) (TC07) (TC05)
(TC08)
(TC10) (TC11)
(TC09)
Evaporator Condenser
Vapor Line Liquid Line
Evap / CC Cold Plate
(TC00) (TC01)
(b)
62

Figure 2.21. Steady state performance map of the LHP relating radial acceleration and
heat transported (Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 37.2 ≤
cp
T ≤ 59.7°C, 25.1 ≤ T
amb

30.2°C).

0
2
4
6
8
10
12
0 100 200 300 400 500 600
a
r
(
g
)
Q
out
(W)
Steady Operation
Quasi-steady Operation
Dry-out
63
Table 2.1. AFRL/RZPS design requirements.
Requirement Parameter
Thermal
Minimum Heat Load 500 W
Minimum Heat Flux 3 W/cm
2

Operating Temperature 200°C
Condenser Heat Sink Temperature 5 to 140°C
Tilt in One G ± 0 inches, horizontal
Conductance 50°C/W
Proof of Pressure Test 3102 psi (200°C)
Materials
Evaporator Envelope Material Titanium, CP Grade 2
Evaporator Wick Material Titanium, CP Grade 2
Transport Line Material Titanium, CP Grade 2
Working Fluid Water
LHP Dimensions
Evaporator Configuration 2.54 cm OD up to 25.4 cm long
Evaporator Footprint 20.32 × 10.16 cm
Condenser Footprint 30.48 × 28.58 cm
Transport Line Lengths Approx. 243.8 cm

64
Table 2.2. ACT LHP geometric design parameters.
Transport Lines
Vapor Line Length Approx. 243.8 cm
Vapor Line Diameter 0.9525 OD × 0.0889 cm wall
Liquid Line Length Approx. 335.3 cm
Liquid Line Diameter 0.6350 OD × 0.0889 cm wall
Condenser Line Length Approx. 279.4 cm
Condenser Line Diameter 0.9525 OD × 0.0889 cm wall
Compensation Chamber
Diameter 6.033 cm OD
Length 11.43 cm
Chamber Location Coaxial with evaporator
Wick Properties
Effective Pore Radius 9.1μm
Permeability 1.2×10
-12
m
2

Outside Diameter 2.286 cm
Length 20.32 cm
Inside Diameter 0.8001 cm
Grooves 6
Groove Depth 0.1524 cm
Groove Width 0.1524 cm

65
Table 2.3. Summary of LHP thermocouple locations
Thermocouple Location
TC04 Evaporator 1
TC05 Evaporator 2
TC06 Evaporator 3
TC07 Evaporator 4
TC08 Evaporator Outlet
TC09 Condenser Inlet
TC10 Condenser 1
TC11 Condenser 2
TC12 Condenser 3
TC13 Condenser Outlet
TC14 Bayonet Inlet
TC15 Compensation Chamber / Evaporator

66
Table 2.4. Summary of uncertainties.
Quantity Uncertainty
Measured Temperature ± (fixed value + confidence interval) °C
Acceleration + (0.01 + confidence interval + (a
r,oe
– a
r,cl
)) g
- (0.01 + confidence interval + (a
r,cl
– a
r,ie
)) g
Voltages ± (0.00025V + 0.005 + confidence interval) V
Constant Resistors ± 0.02%R Ω
Wick D, L ± 0.0000254 m
Calculated
cp
T
± max(ΔT
out
, ΔT
in
)
e
T
± max(Δ(TC04), Δ(TC05), Δ(TC06), Δ(TC07))
ΔT
sh
± 1.28°C
C
p,PAO

± ( ) ( ) ( ) ( )
cp p cp cp p cp p
* % 5 . 0 T C T T C T C − Δ + +
Q
out
Equation (2.8)
h

Equation (2.9)
R

Equation (2.10)


67
Table 2.5. Steady state operating characteristics for the stationary LHP (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T ≤ 71.6°C,
27.6 ≤ T
amb
≤ 38.7°C) showing effect of startup path.
Final Q
in

(W)
Path to Final
Q
in

e
T (°C) T
e/cc
(°C)
cp
T (°C)
Q
out
(W) h (W/m
2
-K) R (K/W)
100
0-100 66.6 ± 0.34 61.8 ± 0.86 39.5 ± 0.12 79.1 ± 4.3 1670 ± 390 0.342 ± 0.019
0-100 64.9 ± 0.34 59.9 ± 0.86 37.4 ± 0.12 78.4 ± 4.3 1660 ± 380 0.351 ± 0.020
0-100 64.8 ± 0.34 59.8 ± 0.86 37.6 ± 0.12 77.9 ± 4.3 1670 ± 390 0.348 ± 0.020
0-100 65.4 ± 0.34 60.1 ± 0.86 36.8 ± 0.12 79.0 ± 4.2 1540 ± 280 0.361 ± 0.020
0-100 65.4 ± 0.34 60.2 ± 0.86 38.2 ± 0.12 77.9 ± 4.3 1540 ± 300 0.350 ± 0.020
0-100-300-100 66.9 ± 0.34 62.2 ± 0.86 38.8 ± 0.12 79.1 ± 4.3 1730 ± 560 0.355 ± 0.020
0-100-500-100 67.4 ± 0.34 62.6 ± 0.86 39.0 ± 0.12 78.8 ± 4.3 1620 ± 500 0.361 ± 0.020
133
0-100-133 65.1 ± 0.34 59.0 ± 0.86 40.4 ± 0.12 114 ± 5.5 1770 ± 310 0.217 ± 0.011
166 0-100-133-166 65.6 ± 0.34 58.3 ± 0.86 43.0 ± 0.12 143 ± 6.5 1760 ± 250 0.158 ± 0.0076
200
0-200 66.3 ± 0.46 66.3 ± 0.86 46.1 ± 0.12 174 ± 7.7 1560 ± 230 0.116 ± 0.0058
0-200 66.4 ± 0.35 66.4 ± 0.86 45.7 ± 0.12 174 ± 7.7 1460 ± 140 0.119 ± 0.0057
0-200 65.5 ± 0.34 65.5 ± 0.86 45.4 ± 0.12 176 ± 7.8 1660 ± 250 0.114 ± 0.0055


68
Table 2.5, continued. Steady state operating characteristics for the stationary LHP (Q
cc
= 0 W, a
r
= 0 g,
cp
m&

= 0.0077 kg/s, 36.8 ≤
cp
T
≤ 71.6°C, 27.6 ≤ T
amb
≤ 38.7°C) showing effect of startup path.
Final Q
in

(W)
Path to Final
Q
in

e
T (°C) T
e/cc
(°C)
cp
T (°C)
Q
out
(W) h (W/m
2
-K) R (K/W)
300
0-100-300 77.1 ± 0.39 62.4 ± 0.86 51.6 ± 0.12 260 ± 11 1660 ± 160 0.0985 ± 0.0045
0-100-300 75.8 ± 0.40 60.9 ± 0.86 50.2 ± 0.12 259 ± 11 1670 ± 170 0.0988 ± 0.0045
0-200-300 78.1 ± 0.38 63.1 ± 0.86 52.0 ± 0.12 258 ± 11 1580 ± 140 0.101 ± 0.0046
0-300 79.3 ± 0.40 64.4 ± 0.86 52.8 ± 0.12 258 ± 11 1550 ± 160 0.103 ± 0.0047
400
0-200-400 94.0 ± 0.34 73.6 ± 0.86 57.8 ± 0.12 345 ± 14 1400 ± 85 0.105 ± 0.0045
0-300-400 94.3 ± 0.35 75.0 ± 0.86 58.4 ± 0.12 345 ± 14 1470 ± 98 0.104 ± 0.0045
0-400 97.4 ± 0.34 72.6 ± 0.86 56.5 ± 0.12 344 ± 14 1190 ± 60 0.119 ± 0.0051
500
0-100-300-500 116 ± 0.34 82.4 ± 0.86 61.7 ± 0.12 430 ± 18 1050 ± 45 0.126 ± 0.0053
0-100-500 119 ± 0.34 83.8 ± 0.86 63.1 ± 0.12 429 ± 18 999 ± 41 0.131 ± 0.0055
0-400-500 117 ± 0.34 83.6 ± 0.86 62.4 ± 0.12 431 ± 18 1050 ± 47 0.127 ± 0.0053
0-500 122 ± 0.34 85.3 ± 0.86 63.7 ± 0.12 432 ± 18 956 ± 42 0.135 ± 0.0057
600
0-200-400-600 145 ± 0.36 93.1 ± 0.86 67.8 ± 0.12 515 ± 21 803 ± 33 0.149 ± 0.0062
0-400-500-600 141 ± 0.34 93.3 ± 0.86 67.5 ± 0.12 515 ± 21 867 ± 37 0.142 ± 0.0060
0-600 148 ± 0.36 93.7 ± 0.86 67.7 ± 0.12 513 ± 21 757 ± 33 0.157 ± 0.0066


69
Table 2.6. Steady state operating characteristics for the stationary LHP showing effect of heat input to the compensation chamber (Q
in

= 500 W, a
r
= 0 g,
cp
m& = 0.0077 kg/s, 63.4 ≤
cp
T ≤ 64.8°C, 36.1 ≤ T
amb
≤ 38.1°C).
Q
cc
(W)
e
T (°C) T
e/cc
(°C)
cp
T (°C)
Q
out
(W) h (W/m
2
-K) R (K/W)
0 122 ± 0.34 85.3 ± 0.86 63.7 ± 0.12 432 ± 18 956 ± 42 0.135 ± 0.0057
5 115 ± 0.35 85.3 ± 0.86 63.6 ± 0.12 438 ± 18 1130 ± 58 0.118 ± 0.0050
10 109 ± 0.35 85.6 ± 0.86 63.8 ± 0.12 444 ± 18 1420 ± 110 0.103 ± 0.0044
15 107 ± 0.35 85.1 ± 0.86 63.4 ± 0.12 449 ± 19 1610 ± 130 0.0970 ± 0.0041
20 108 ± 0.34 86.1 ± 0.86 64.0 ± 0.12 456 ± 19 1610 ± 130 0.0972 ± 0.0041
25 110 ± 0.34 87.7 ± 0.86 64.2 ± 0.12 461 ± 19 1620 ± 130 0.0991 ± 0.0042
30 112 ± 0.34 90.0 ± 0.86 64.5 ± 0.12 466 ± 19 1610 ± 130 0.103 ± 0.0044
35 118 ± 0.35 95.5 ± 0.86 64.6 ± 0.12 468 ± 19 1580 ± 120 0.114 ± 0.0048
40 120 ± 0.34 97.4 ± 0.86 64.8 ± 0.12 473 ± 20 1580 ± 120 0.117 ± 0.0049
45 127 ± 0.34 104 ± 0.86 64.5 ± 0.12 475 ± 20 1530 ± 110 0.132 ± 0.0055
50 129 ± 0.35 105 ± 0.86 64.5 ± 0.12 477 ± 20 1510 ± 110 0.134 ± 0.0056


70
Table 2.7. The effect of compensation chamber temperature control on LHP operation (Q
in
= 500 W, a
r
= 0 g,
cp
m& = 0.0077 kg/s,
cp
T
= 52.5°C, T
amb
= 26.4°C)
Compensation Chamber
Conditions
Q
cc
(W)
e
T (°C)
cp
T (°C)
T
cc
(°C) Q
out
(W) h (W/m
2
-K) R (K/W)
Time to
Steady
State
Uninsulated, no
temperature control
0 115 52.4 59.3 442 ± 19 878 ± 37 0.142 ± 0.0060 300 min.
Insulated, no
temperature control
0 107 52.6 62.9 447 ± 19 1050 ± 51 0.122 ± 0.0052 375 min.
Insulated, temperature
controlled to T
cc
=
72.8°C, simultaneous
heat input startup
20 103 53.5 72.8 470 ± 20 1340 ± 77 0.106 ± 0.0045 310 min.
Insulated, temperature
controlled to T
cc
=
72.8°C, preconditioned
CC
100
decreased
to 20
103 53.4 72.8 467 ± 19 1350 ± 77 0.106 ± 0.0045 250 min.


71
Table 2.8. Steady state operating characteristics of the rotating LHP (Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 37.2 ≤
cp
T ≤ 59.7°C, 25.1 ≤ T
amb

≤ 30.2°C).
Q
in
(W) a
r
(g)
e
T (°C) T
e/cc
(°C)
cp
T (°C)
Q
out
(W) h (W/m
2
-K) R (K/W)
100 0.1 ± 0.012 64.9 ± 0.34 59.8 ± 0.86 38.0 ± 0.12 74.2 ± 4.2 1620 ± 340 0.362 ± 0.021
100 0.1 ± 0.012 64.6 ± 0.34 59.8 ± 0.86 38.1 ± 0.12 74.6 ± 4.2 1810 ± 510 0.355 ± 0.020
100 2.0 ± 0.050 66.4 ± 0.34 61.2 ± 0.86 37.5 ± 0.12 79.9 ± 4.5 1770 ± 420 0.361 ± 0.021
100 4.1 ± 0.091 71.2 ± 0.44 64.3 ± 0.86 37.2 ± 0.16 76.1 ± 4.8 1130 ± 150 0.446 ± 0.029
200 0.1 ± 0.012 70.8 ± 0.41 59.3 ± 0.86 42.0 ± 0.13 174 ± 7.8 1400 ± 110 0.165 ± 0.0078
200 0.1 ± 0.012 71.1 ± 0.43 59.5 ± 0.86 41.9 ± 0.13 172 ± 7.8 1380 ± 110 0.169 ± 0.0081
200 0.1 ± 0.012 71.5 ± 0.38 60.3 ± 0.86 42.0 ± 0.12 174 ± 7.8 1360 ± 110 0.170 ± 0.0079
200 2.0 ± 0.050 75.8 ± 0.65 61.6 ± 0.86 41.7 ± 0.21 188 ± 9.2 1220 ± 86 0.181 ± 0.0096
300 0.1 ± 0.012 76.4 ± 0.48 57.2 ± 0.86 46.0 ± 0.12 275 ± 12 1390 ± 100 0.111 ± 0.0051
300 2.0 ± 0.050 81.7 ± 0.76 61.2 ± 0.86 45.6 ± 0.16 285 ± 13 1280 ± 80 0.126 ± 0.0063
400 0.1 ± 0.012 86.4 ± 0.39 60.6 ± 0.86 49.3 ± 0.12 360 ± 15 1300 ± 82 0.103 ± 0.0045
400 0.1 ± 0.012 86.6 ± 0.39 61.3 ± 0.86 49.3 ± 0.12 361 ± 15 1300 ± 81 0.103 ± 0.0045
400 0.1 ± 0.012 86.1 ± 0.39 61.1 ± 0.86 49.3 ± 0.12 361 ± 15 1330 ± 84 0.102 ± 0.0044
400 0.1 ± 0.012 85.8 ± 0.42 61.1 ± 0.86 49.3 ± 0.12 361 ± 15 1330 ± 89 0.101 ± 0.0044
400 2.0 ± 0.050 87.6 ± 0.42 59.7 ± 0.86 49.3 ± 0.12 372 ± 16 1280 ± 74 0.103 ± 0.0045
400 4.0 ± 0.089 91.5 ± 0.43 59.8 ± 0.86 49.3 ± 0.12 376 ± 16 1160 ± 56 0.112 ± 0.0051
400 4.1 ± 0.091 93.0 ± 0.63 60.0 ± 0.86 49.3 ± 0.15 376 ± 17 1130 ± 59 0.116 ± 0.0055
400 6.0 ± 0.13 94.8 ± 0.34 60.7 ± 0.86 49.9 ± 0.12 369 ± 16 1030 ± 46 0.122 ± 0.0053


72
Table 2.8, continued. Steady state operating characteristics of the rotating LHP (Q
cc
= 0 W,
cp
m& = 0.0077 kg/s, 37.2 ≤
cp
T ≤ 59.7°C,
25.1 ≤ T
amb
≤ 30.2°C).
Q
in
(W) a
r
(g)
e
T (°C) T
e/cc
(°C)
cp
T (°C)
Q
out
(W) h (W/m
2
-K) R (K/W)
500 0.1 ± 0.012 110 ± 0.36 69.3 ± 0.86 52.9 ± 0.12 440 ± 18 960 ± 43 0.131 ± 0.0055
500 0.1 ± 0.012 107 ± 0.35 69.0 ± 0.86 52.6 ± 0.12 446 ± 19 1030 ± 48 0.123 ± 0.0052
500 0.1 ± 0.012 107 ± 0.41 69.2 ± 0.86 52.7 ± 0.12 447 ± 19 1030 ± 50 0.122 ± 0.0052
500 0.1 ± 0.012 109 ± 0.42 69.5 ± 0.86 52.9 ± 0.12 447 ± 19 1000 ± 46 0.125 ± 0.0053
500 0.1 ± 0.012 108 ± 0.35 68.9 ± 0.86 52.6 ± 0.12 448 ± 19 1010 ± 47 0.123 ± 0.0052
500 2.0 ± 0.050 114 ± 0.42 65.5 ± 0.86 52.5 ± 0.12 452 ± 19 887 ± 37 0.136 ± 0.0058
500 4.0 ± 0.090 110 ± 0.34 64.4 ± 0.86 53.1 ± 0.13 459 ± 19 982 ± 43 0.123 ± 0.0053
500 6.0 ± 0.13 111 ± 0.34 64.6 ± 0.86 54.0 ± 0.13 460 ± 20 934 ± 42 0.125 ± 0.0055
500 8.0 ± 0.17 115 ± 0.34 65.1 ± 0.86 54.8 ± 0.12 463 ± 20 865 ± 38 0.131 ± 0.0058
500 10 ± 0.21 117 ± 0.34 64.7 ± 0.86 55.2 ± 0.12 452 ± 20 824 ± 35 0.136 ± 0.0060
600 0.1 ± 0.012 137 ± 0.36 76.6 ± 0.86 56.0 ± 0.12 533 ± 22 759 ± 32 0.152 ± 0.0063
600 0.1 ± 0.012 141 ± 0.43 76.6 ± 0.86 55.9 ± 0.12 533 ± 22 707 ± 29 0.160 ± 0.0067
600 0.1 ± 0.012 141 ± 0.43 76.3 ± 0.86 55.8 ± 0.12 533 ± 22 703 ± 29 0.160 ± 0.0067
600 0.1 ± 0.012 140 ± 0.42 76.2 ± 0.86 55.8 ± 0.12 531 ± 22 707 ± 30 0.159 ± 0.0067
600 0.1 ± 0.012 133 ± 0.34 76.2 ± 0.86 56.0 ± 0.12 539 ± 23 821 ± 35 0.143 ± 0.0060
600 2.0 ± 0.050 140 ± 0.34 72.9 ± 0.86 56.3 ± 0.12 547 ± 24 726 ± 31 0.152 ± 0.0066
600 4.0 ± 0.090 152 ± 0.50 71.7 ± 0.86 57.1 ± 0.12 546 ± 23 621 ± 26 0.173 ± 0.0075
600 6.0 ± 0.13 148 ± 0.34 71.5 ± 0.86 57.8 ± 0.12 545 ± 24 651 ± 27 0.165 ± 0.0072
600 8.1 ± 0.17 147 ± 0.34 71.8 ± 0.86 58.8 ± 0.12 546 ± 24 662 ± 28 0.162 ± 0.0072
600 10 ± 0.21 145 ± 0.34 72.2 ± 0.86 59.7 ± 0.13 538 ± 24 691 ± 29 0.158 ± 0.0070

73
Table 2.9. Comparison of quasi-steady states for Q
in
= 200 W (Q
cc
= 0 W,
cp
m& = 0.0077
kg/s).
a
r
(g)
0 0.1 Δ
T
amb
(°C)
31.7 26.4 -5.3
T
cp,in
(°C)
41.1 36.9 -4.2
T
cp,out
(°C)
51.1 47.0 -4.1
cp
T (°C)
46.1 41.9 -4.2
e
T (°C)
66.3 71.1 4.8
T
bayonet inlet,max
(°C)
50.7 51.2 0.5
T
bayonet inlet,min
(°C)
42.6 38.9 -3.7
T
e/cc,max
(°C)
56.8 60.9 4.1
T
e/cc,min
(°C)
54.0 57.1 3.1
h (W/m
2
-K)
1560 1380 -180
R (W/K)
0.116 0.169 0.053
ΔT
sh
(°C)
11.2 11.6 0.4

74
REFERENCES
Anderson, J. Introduction to Flight. Boston: McGraw-Hill, 2000.
Boyer, R., Welsch, G., Collings, E. Materials Properties Handbook: Titanium Alloys.
Materials Park, OH: ASM International, 1994.
Cimbala, J., Brenizer, J., Chuang, A., Hanna, S., Conroy, C., El-Ganayni, A., Riley, D.
"Study of a Loop Heat Pipe Using Neutron Radiography." Applied Radiation and
Isotopes 61, 2004: 701-705.
Cloyd, J. "A Status of the United States Air Force's More Electric Aircraft Initiative."
Energy Conversion Engineering Conference. Honolulu, HI: IEEE, 1997. 681-686.
DOD. MIL-HDBK-310: Global Climatic Data for Developing Military Products. 1997.
Douglas, D., Ku, J., Kaya, T. "Testing of the Geoscience Laser Altimeter System (GLAS)
Prototype Loop Heat Pipe." 37th AIAA Aerospace Sciences Meeting and Exhibit.
Reno, NV: AIAA, 1999. AIAA 99-0473.
Faghri, A. Heat Pipe Science and Technology. Washington D.C.: Taylor and Francis,
1995.
Ghajar, A., Tang, W., Beam, J. "Comparison of Hydraulic and Thermal Performance of
PAO and Coolanol 25R." 6th AIAA/ASME Joint Thermophysics and Heat Transfer
Conference. Colorado Springs, CO: AIAA, 1994. AIAA-94-1965.
Hoang, T., Ku, J. "Transient Modeling of Loop Heat Pipes." 1st International Energy
Conversion Engineering Conference. Portsmouth, VA: AIAA, 2003. AIAA 2003-
6082.
Incropera, F., DeWitt, D. Fundamentals of Heat and Mass Transfer. New York City:
Wiley, 2002.
75
Kaya, T., Hoang, T., Ku, J., Cheung, M. "Mathematical Modeling of Loop Heat Pipes."
37th AIAA Aerospace Sciences Meeting and Exhibit. Reno, NV: AIAA, 1999. A99-
16362.
Kaya, T., Ku, J. "Experimental Investigation of Performance Characteristics of Small
Loop Heat Pipes." Proceedings of 41st Aerospace Sciences Meeting and Exhibit.
Reno, NV: AIAA, 2003. AIAA 2003-1038.
Ku, J. "Operating Characteristics of Loop Heat Pipes." 29th International Conference on
Environmental System. Denver, CO: SAE, 1999. 1999-01-2007.
Ku, J., Ottenstein, L., Kaya, T., Rogers, P., Hoff, C. "Testing of a Loop Heat Pipe
Subjected to Variable Accelerating Forces, Part 1: Start-Up." SAE, 2000a: 2000-01-
2488.
Ku, J., Ottenstein, L., Kaya, T., Rogers, P., Hoff, C. "Testing of a Loop Heat Pipe
Subjected to Variable Accelerating Forces, Part 2: Temperature Stability." SAE,
2000b: 2000-01-2489.
Maidanik, J. "Review: Loop Heat Pipes." Applied Thermal Engineering 25, 2005: 635-
657.
Maidanik, J., Vershinin, S., Kholodov, V., Dolgirev, J. Heat Transfer Apparatus. United
States Patent 4,515,209. 1985.
Montgomery, D., Runger, G. Applied Statistics and Probability for Engineers. New York
City: Wiley, 2003.
NACA. Report 1135: Equations, Tables, and Charts for Compressible Flow.
Washington, D.C.: U.S. Government Printing Office, 1953.
Ponnappan, R., Yerkes, K., Chang, W., Beam, J. "Analysis and Testing of Heat Pipe in
Accelerating Environment." Proceedings of the 8th International Heat Pipe
Conference. Beijing, China, 1992. B-19-1 - B-19-6.
Quigley, R. "More Electric Aircraft." Applied Power Electronics Conference and
Exposition. San Diego, CA: IEEE, 1993. 906-911.
76
Thomas, S., Klasing, K., Yerkes, K. "The Effects of Transverse Acceleration-Induced
Body Forces on the Capillary Limit of Helically Grooved Heat Pipes." ASME Journal
of Heat Transfer, 120, 1998: 441-451.
Thomas, S., Yerkes, K. "Quasi-Steady-State Performace of a Heat Pipe Subjected to
Transient Acceleration Loadings." AIAA Journal of Thermophysics, Vol. 11, No. 2,
1996: 306-309.
Vrable, D., Yerkes, K. "A Thermal Management Concept for More Electric Aircraft
Power System Applications." Aerospace Power Systems Conference Proceedings.
Williamsburg, VA: SAE, 1998. P-322.
White, F. Heat and Mass Transfer. New York City: Addison-Wesley, 1988.
Yerkes, K., Beam, J. "Arterial Heat Pipe Performance in a Transient Heat Flux and Body
Force Environment." SAE, 1992: 921944.
Zaghdoudi, M., Sarno, C. "Investigation of the Effects of Body Force Environment on
Flat Heat Pipes." AIAA Journal of Thermophysics and Heat Transfer, Vol. 15, No. 4,
2001: 384-394.

77
APPENDIX A. OPERATING PROCEDURES
A.1. Standard Operating Procedure
0. Ensure the main power breaker is in the “OFF” position. The breaker is located
on the 71B H-bay second floor landing. See Figure A.1 for picture. Ensure the control
panel box is in “Man” mode, potentiometer is set to zero (turned completely counter-
clockwise), and the E-stop button has been depressed. The control panel box is located in
Control Room 2. See Figure A.2 for picture of control panel box.
1. Mount test payload with sufficient factor of safety to ensure a reliable mounting
configuration consistent with the generated forces. The retaining method will be
designed for a payload mass subjected to a maximum table capacity of 12 g.
2. Mount the appropriate counter balance weight at the appropriate location to place
the centroid in the center of the table and level each spoke to minimize vibrational noise
and table runout.
3. Ensure all centrifuge maintenance has been completed.
4. Software Startup Procedure
a. Open the LabVIEW VI file needed to control the data acquisition and
table voltage.
b. It is up to the experimentalist to write the program used to control data
acquisition and table voltage. Sample programs that perform these tasks
are available. See Figure A.3 for sample.
c. Press the ‘SYSTEM ENABLED’ button on the front panel so that it is
illuminated.
d. Press the ‘Run’ button located in the top-left of the LabVIEW toolbar.
This will cause the program to become functional.
5. Prior to each set of experimental testing:
a. Check all bolts to ensure all experimental apparatus are tightened
properly.
78
b. Clear the table of foreign objects, including removing all tools and placing
in the appropriate CTK. Ensure all free standing equipment and furniture
are securely placed along the perimeter of the room.
c. Turn table power switch on. This provides power to the instrumentation
and devices on the table. See Figure A.4 for location.
d. Check out instrumentation for proper operation.
e. Turn on recirculating chiller located on the east wall. See Figure A.5 for
picture. Check chill bath coolant flows and flow rates of any intermediate
flow loops. A booster pump for the chill bath coolant is available. Follow
the chill bath plumbing schematic in Figure A.6 to enable operation. The
booster pump control is remotely located in Control Room 2. See Figure
A.7 for picture.
f. Lock inter-connect door.
g. Check camera operation.
h. Make final check on the table for tools or loose objects.
6. Unlock main power breaker and flip to ‘ON’ position. See Figure A.1 for
location.
a. Ensure the emergency stop button is activated and the potentiometer is
turned completely counter-clockwise on the control panel box prior to
proceeding with powering the centrifuge table motor.
b. Turn on warning beacon and evacuate personnel. Warning beacon
switches are located outside Test Cell 4 and on the west wall of control
room 1.
c. Engage table motor control switch on the north wall (cooling motor will
be operational). See Figure A.8 for location.
d. Secure outer doors.
e. Place “Test in Progress Do Not Enter” sign on the outer door.
7. Controlling the System
a. To control the voltage supplied to the table, turn the mode switch from
‘MAN’ to ‘AUTO’ on the control panel box and flip the ‘Table Voltage’
switch to the ‘ON’ position. While the ‘Table Voltage’ switch is ‘ON’,
79
the voltage can be adjusted using the slider bar on the left side called
‘Table Voltage’. Table voltage can be turned on and off as often as
desired while the system is engaged.
8. Conduct test procedure. Test procedures are experiment dependent and up to the
experimentalist to develop.
9. Shutdown
a. Slowly reduce the table voltage to zero using the slider bar. Flip the
‘Table Voltage’ switch to the ‘OFF’ position on the LabVIEW VI.
b. On the control panel box, turn the switch from ‘AUTO’ to ‘MAN’ and
press the ‘STOP’ button. Wait for the table to stop rotating then press the
‘E-STOP’ button.
c. Press the ‘SYSTEM ENABLED’ button on the LabVIEW VI so that it is
no longer illuminated. The program will stop.
10. Emergency Shutdown
a. If for any reason an emergency should occur press the ‘E-STOP’ on the
control panel box. Should the table “run away” or suddenly accelerate the
motor will automatically shutdown. Contact the appropriate personnel
prior to a restart after an emergency shutdown.
• Andrew Fleming 58942
• Larry Byrd 53238
• Travis Michalak 64429
A.2. Test Procedures
1. Stationary Operation
a. Set chiller to T
eg
= 35°C. Allow to come to steady state.
b. Turn high temperature coolant loop on and set to m& = 0.0077 kg/s.
Simultaneously, apply desired heat load to evaporator.
c. Allow LHP to achieve steady state operation by examining dT/dt plot of
TC00, TC01, TC04, TC05, TC06, TC07, TC08, and TC09. Steady state is
achieved when all of these thermocouples are bracketed by -0.01 ≤ dT/dt ≤
0.01 K/min.
80
d. Remove heat load from the evaporator and turn off the high temperature
coolant loop or adjust to next desired heat load and repeat (c).
2. Rotational Operation
a. Set chiller to T
eg
= 35°C. Allow to come to steady state.
b. Increase radial acceleration to a
r
= 0.1 g.
c. Turn high temperature coolant loop on and set to m& = 0.0077 kg/s.
Simultaneously, apply desired heat load to evaporator.
d. Allow LHP to achieve steady state operation by examining dT/dt plot of
TC00, TC01, TC04, TC05, TC06, TC07, TC08, and TC09. Steady state is
achieved when all of these thermocouples are bracketed by -0.01 ≤ dT/dt ≤
0.01 K/min.
e. Increase radial acceleration to desired level.
f. Again, allow LHP to achieve steady state operation by examining dT/dt
plot of TC00, TC01, TC04, TC05, TC06, TC07, TC08, and TC09. Steady
state is achieved when all of these thermocouples are bracketed by -0.01 ≤
dT/dt ≤ 0.01 K/min.
g. Decrease radial acceleration to a
r
= 0.1 g.
h. If shutting down, allow LHP to operate for 30 min., then remove heat load
from the evaporator and turn off the high temperature coolant. Decrease
radial acceleration to a
r
= 0.0 g. If continuing testing, allow LHP to
achieve steady state operation by examining dT/dt plot of TC00, TC01,
TC04, TC05, TC06, TC07, TC08, and TC09. Steady state is achieved
when all of these thermocouples are bracketed by -0.01 ≤ dT/dt ≤ 0.01
K/min. Repeat (e)-(h).
81

(a)

(b)
Figure A.1. Centrifuge table main power breaker: (a) Electrical panel MCC-6; (b)
Centrifuge table main power breaker.

82

Figure A.2. Centrifuge table control panel box.

83

Figure A.3. Sample LabVIEW control program.

84

Figure A.4. Centrifuge table power switch.

85

Figure A.5. Neslab recirculating chiller.

86

Figure A.6. Chill bath plumbing schematic.

Supply
Recirculating
Chiller
F
P
F/M
0-5LPM
F/M
0-25LPM
F/M
0-50LPM
To Centrifuge Table From Centrifuge Table
Return
Recirc
To Drain
Booster
Pump
Filter
1
2
3
4
5
6
87

Figure A.7. Booster pump control panel.

88

Figure A.8. Centrifuge table motor control power switch.

89
APPENDIX B. UNCERTAINTY ANALYSIS
The uncertainty analysis for this experiment was broken into the following three
different categories: measured, constant, and calculated quantities. For each steady state
condition, 151 data points were collected representing five minutes of data. Measured
values were averaged and uncertainties were calculated based on the fixed error of each
instrument plus the confidence interval for the average. The fixed error for
thermocouples and mass flow meter was determined via the calibration methods and can
be seen in Appendix C. The fixed error of the accelerometer was ±0.1 g and ±0.00025*V
+ 0.005 volts for voltage measurements. The confidence interval was based off a
statistical t-distribution with confidence level of 0.95 and was given by Montgomery
(Montgomery, 2003)

Confiuence Inteival = ¬t |
o
√n
1
(B.1)
where t is a tabulated value based on the confidence level and number of degrees of
freedom, σ is the sample standard deviation, and n is the number of data points in the
sample. Constant quantities included precision resistors used for current measurement
for heat input via evaporator, compensation chamber, and preheater with an uncertainty
of ±0.02% and wick diameter and length measurements, with an uncertainty of
±0.0000254 m. Calculated quantities used uncertainty methods that would be the most
conservative for the experiment. For averaged quantities including
cp
T ,
e
T , and h , the
largest uncertainty of the individual measurements was used as the uncertainty of the
average value. For the specific heat of PAO, the uncertainty was 0.5% of the total value
plus difference between the specific heat using the upper limit of the temperature
measurement and the average specific heat. The calorimetry of the cold plate, Q
out
, was
given by
0
out
= m C
p,PA0
(I
out
- I
in
) (B.2)
90
The uncertainty of Q
out
was determined to be

Δ0
out
= ||
o0
out
om
Δm 1
2
+ |
o0
out
oC
p,PA0
ΔC
p,PA0
+
2
+ |
o0
out
oI
out
ΔI
out
1
2
+ |
o0
out
oI
in
ΔI
in
1
2
|
1¡2

(B.3)
where

o0
out
om
= C
p,PA0
(I
out
- I
in
)
(B.4)

o0
out
oC
p,PA0
= m (I
out
- I
in
) (B.5)

o0
out
oI
out
= m C
p,PA0
(B.6)

o0
out
oI
in
= -m C
p,PA0
(B.7)
The average evaporative heat transfer coefficient, h , was given by
b

=
0
out
nÐI(I

e
- I
v
)
(B.8)
The uncertainty of h was given by

Δb

= ||
ob

o0
out
Δ0
out
+
2
+ |
ob


ΔÐ+
2
+ |
ob

oI
ΔI+
2
+|
ob

oI

e
ΔI

e
+
2
+ |
ob

oI
v
ΔI
v
+
2
|
1¡2

(B.9)
where

ob

o0
out
=
1
nÐI(I

e
-I
v
)
(B.10)
91

ob


=
-0
out

2
I(I

e
- I
v
)
(B.11)

ob

oI
=
-0
out
nÐI
2
(I

e
- I
v
)
(B.12)

ob

oI

e
=
-0
out
nÐI(I

e
-I
v
)
2
(B.13)

ob

oI
v
=
-0
out
nÐI(I

e
- I
v
)
2
(B.14)
The thermal resistance, R, was given by

R =
I

e
- I

cp
0
out

(B.15)
The uncertainty of the thermal resistance was given by

ΔR = ||
oR
o0
out
Δ0
out
1
2
+ |
oR
oI

e
ΔI

e
1
2
+ |
oR
oI

cp
ΔI

cp
+
2
|
1¡2

(B.16)
where

oR
o0
out
=
-(I

e
-I

cp
)
0
out
2

(B.17)

oR
oI

e
=
1
0
out
(B.18)

oR
oI

cp
=
-1
0
out
(B.19)

92
APPENDIX C. CALIBRATION OF THERMOCOUPLES AND FLOW METER
C.1. Thermocouple Calibration
The calibrations of the thermocouples used on the centrifuge table required new
programming since the data acquisition system was upgraded for the centrifuge table.
Control of a calibration bath and RTD were integrated with the centrifuge table’s data
acquisition unit. A LabVIEW program was generated to simultaneously control all three
devices.
The thermocouples were calibrated using a Hart Scientific 6330 Calibration Bath
and Hart Scientific 1502A NIST-Traceable platinum resistance temperature detector
(RTD). The bath was capable of producing steady state temperatures from 40 to 280°C.
The calibration bath used Dow Corning 200.50 silicon oil. There were several steps
required before a thorough calibration of the thermocouples could be determined. First,
LabVIEW software needed to be written to interface with the calibration bath, RTD, and
data acquisition system. Second, the characteristics of the calibration bath and RTD
needed to be determined with respect to response times and temperature fluctuations.
Finally, complete calibration curves for each of the thermocouples needed to be
developed.
In the loop heat pipe experimental setup, there were four grounded probe
thermocouples and twelve exposed tip type E thermocouples. The grounded probe
thermocouples were used in the calorimetry of the cold plate, temperature measurement
for the flow meter calibration, and the ambient temperature inside the box, where the
error needed to be reduced as much as possible. As a result, it was decided to only
calibrate these four over the anticipated temperature range, from 20 to 145°C, in 5°C
intervals. The twelve exposed tip thermocouples were mounted on the loop heat pipe in
various locations. They needed to be calibrated over the full 20 to 230°C temperature
range. Thus, the calibration of the thermocouples needed a second calibration bath that
was capable of achieving the temperature range of 20 to 40°C. The Hart Scientific 6330
Calibration Bath was not capable of maintaining a steady temperature below 40°C for
93
Dow Corning 200.50 bath fluid. The bath used for this portion of the calibration was a
Brinkmann Lauda RCS 20-D calibration bath using Brayco Micronic 889
(polyalphaolefin or PAO) as the bath fluid. The same RTD was used as before, and the
temperature set point for the calibration bath was set manually rather than through the
data acquisition interface. After some examination, it was noticed that due to the limited
capability of the Brinkmann bath to heat and cool, it was difficult to maintain a fairly
constant temperature in the PAO. As a result, an insulated copper tube with a closed base
was placed in the bath, and the copper tube was filled with PAO. This significantly
reduced the temperature fluctuation in the bath temperature.
The first step in the development of the thermocouple calibrations was to write
the LabVIEW software to interface with the calibration bath, RTD, and data acquisition.
Initially, sub-VI’s were developed to interface directly with the calibration bath for
reading the current temperature and setting the bath temperature set point as well as for
reading the RTD. Wire diagrams of the RTD read, calibration bath temperature set point,
and calibration bath current temperature read can be found in Figure C.1 through Figure
C.3, respectively. Once these sub-VI’s were developed, they needed to be incorporated
into a larger framework. The front panel and wire diagram for the automatic calibration
program can be seen in Figure C.4 while the front panel and wire diagram for the manual
calibration program can be seen in Figure C.5.
Due to electrical noise in the centrifuge table test cell and fluctuations in the
calibration bath temperature, a method was devised to determine that the calibration bath
had reached a steady state. The previous 100 RTD temperature values were recorded into
an array, and the standard deviation of the sampling was calculated. This number of
RTD readings was selected for a 95% probability and a confidence interval of 0.95
(Montgomery, 2003). When the standard deviation dropped below the specified
threshold of 0.005°C, the program would indicate that the system had reached steady
state and the thermocouples were then read. The bath temperature would then be
incremented or decremented as necessary. First, the full calibration process began by
placing the thermocouples in the Hart Scientific calibration bath for the temperature
range 40 to 230°C in 5°C increments ramping up and down with a standard deviation
threshold of 0.005°C. Then, the thermocouples and RTD were cleaned and placed in the
94
Brinkmann bath for the temperature range 20 to 35°C. The data from each process was
combined to produce one composite data set constituting the entire temperature range.
For data reduction, all 100 data points constituting one nominal temperature value
were averaged and the confidence interval was calculated for the RTD. After this
process, due to the increment and decrement of the calibration process, there were two
data points for each nominal temperature value, as shown in Figure C.6. Plots of RTD
versus each thermocouple were generated, and polynomial trend lines were fit for each
thermocouple. A sample plot of TC00 is given in Figure C.7. Also, in an effort to reduce
maximum deviation of the actual versus calculated RTD values, higher-order polynomial
trend lines, from first to fifth order, were implemented to evaluate maximum deviation.
A fifth-order polynomial was selected since it reduced the maximum deviation by
approximately a factor of 4 over a first-order trend line. Results from this analysis can be
seen in Table C.1 for TC00. Table C.2 shows the trend line equations for each
thermocouple in tabular form.
The uncertainty associated with each thermocouple was determined by accounting
for four sources of error: maximum measured uncertainty inherent to the RTD, the
maximum confidence interval of the RTD temperature over the 100 readings used in the
calibration, the maximum deviation of the calculated temperature from the measured
temperature, and the confidence interval associated with the 100 data points in the sample
of the thermocouples. The error inherent to the RTD was ±0.009°C. The maximum
confidence interval of the RTD from the 100 readings was ±0.0055°C over the entire
temperature range. The maximum deviations of the calculated temperature and the
measured temperature are thermocouple specific, and can be seen in Table C.3 along with
the total error of each thermocouple.
C.1.1. Calibration Procedure
1. Mount all of the thermocouples to the RTD probe with the thermocouple and
RTD tips as close to each other as possible.
2. Place the thermocouple and RTD bundle vertically into the Hart Scientific 6330
Calibration Bath, with the probes not touching any of the bath surfaces.
3. Turn on the RTD and calibration bath, setting the bath to 40°C.
4. Turn on the centrifuge table power.
95
5. In the control room, turn on LabVIEW and open the thermocouple calibration
program.
6. Set the lower temperature set point to 40°C and the upper temperature set point to
230°C, with a temperature increment of 5°C.
7. Set the standard deviation threshold to 0.005°C.
8. Run the VI.
9. Current temperatures can be examined while the calibration is in progress on the
main screen.
10. After this calibration cycle has been completed, remove and clean the
thermocouples and RTD, rebundle, and place in the copper tube that is located in
the Brinkmann Lauda RCS 20-D calibration bath.
11. Turn the bath on and set to 20°C.
12. Open the manual thermocouple calibration VI.
13. Ensure the “Keep Running” button is depressed, then run the program.
14. Wait for the RTD standard deviation to reach below 0.005, the press the ‘Proceed
to TC Read’ button.
15. Increment the bath temperature by 5°C up to 35°C and back down to 20°C,
recording temperature values at each location an overall two times.
16. Take the two data files and combine to make one composite data file.
C.2. Flow Meter Calibration
The calibration of the turbine flow meter for the high-temperature fluid loop
proved to be a difficult challenge. Due to the chemistry of PAO, there is a significant
difference in density and viscosity with respect to temperature. As a result, a calibration
surface that was dependent on flow meter output voltage and temperature was required to
determine the actual mass flow rate. An uncertainty analysis was also performed to fully
characterize the flow meter.
A calibration setup was developed using a Lauda RCS-20D calibration bath filled
with PAO, Tuthill pump from the high-temperature fluid loop, SS-56S6 Whitey inline
filter with a 140 micron filter, calibrated type-E thermocouple that was used for
measuring temperature entering the flow meter in the high temperature fluid loop, and an
Omega FTB-9506 turbine flow meter with FLSC-61 signal conditioner. A LabVIEW
96
code was developed to aid in the calibration of the flow meter (Figure C.8). This code
served as the stopwatch and data recording for the calibration. Temperature and flow
meter voltage were recorded, with raw data files generated. The pump was controlled via
a 0-10V input signal to a TECO FM50 motor controller. The flow meter required an
input voltage between 12-28VDC and output voltage between 0-10V. All data was
collected through the centrifuge table slip rings to the data acquisition to capture all error
inherent to the centrifuge table test bed. A calibrated Mettler PC4400 scale was used to
determine the mass collected during a given test run. Flow straightening sections
upstream and downstream were placed according to the manufacturer's instructions. A
three-way valve was installed after the flow meter, which allowed the entire flow system
to reach a steady temperature. A schematic of the setup is shown in Figure C.9. Once
the temperature was steady, the flow was diverted to a catch basin for a specified amount
of time. The voltage from the flow meter and the temperature from the thermocouple
were recorded during this time, and when the basin was full, the flow was again diverted
to recirculating the PAO back to the chiller. During each measurement, as many data
points as possible were collected across the time span with the limiting factor being the
iteration time on the LabVIEW software. The minimum number of data points collected
for any given run was 437. The voltages and temperatures were averaged and a
confidence interval of 0.95 was calculated based on a statistical t-distribution for each test
run. The test was repeated for a total of five averaged data points for each nominal
temperature and flow rate. These tests were completed over the range of T = 20 to 120°C
in intervals of 25°C and flow rates ranging from m& = 0.0064 to 0.025 kg/s in intervals of
approximately 0.002 kg/s. A 3-D paraboloid regression equation was generated using
SigmaPlot to relate temperature, flow meter voltage, and mass flow rate (Figure C.10).
The equation can be seen in Table C.4.
During the course of developing the calibration setup and testing, several issues
were encountered. Immediately from the start of working out the bugs in the system, it
was noticed that the flow meter would not output a voltage linearly as expected, but
rather responded in a quadratic fashion with a local maximum at approximately six volts.
It was discovered that the motor housing of the pump was not properly grounded, causing
electromagnetic interference to disrupt the operation of the flow meter. The output
97
voltage was extremely inconsistent, changing +/- 0.5 volts at apparently random times.
Initially an SS-56S6 Whitey inline filter with a 140 micron filter was placed in the flow
loop to catch any debris that may have been picked up by the pump. It was then noticed
that on the flow loop outlet, air bubbles were being rejected. All of the fittings were
retightened and the flow remained steady. At apparently random times the flow meter
would start outputting random voltage readings. It was discovered that screw terminal
was not tightened down on the wire connector, occasionally creating an open circuit loop.
During the course of data collection, the output voltage would develop a trend, as seen in
Figure C.11. When this was observed during data reduction, another data point would be
collected to replace it. This was likely due to fluctuations in the flow rate and was
observed more prevalently at higher flow rates.
With this type of calibration, it is critical to have a firm grasp on the uncertainty
associated with the mass flow rate. Three types of uncertainty were identified associated
with this calibration: error associated with the scale and time, error due to the voltage
confidence interval and thermocouple error and confidence interval, and deviation of the
fit equation from actual data. Each of these errors are described below.
The total error associated with this uncertainty analysis is given by
Δm
tot
= Δm
m¡t
+Δm
v¡T
+ Δm
uev
(C.1)
The error for
m/t
m& Δ was determined by
m =
m
t
(C.2)
which yields the uncertainty

Δm
m¡t
= ||
om
m¡t
om
Δm+
2
+|
om
m¡t
ot
Δt+
2
|
1¡2

(C.3)
where

om
m¡t
om
=
1
t

(C.4)
and
98

om
m¡t
ot
= -
m
t
2

(C.5)

The error for
V/T
m& Δ was determined by performing an error analysis on the fit equation
developed using SigmaPlot
m = y
o
+oI + bI +cI
2
+ JI
2
(C.6)
which yields the uncertainty

Δm
v¡T
= ||
om
v¡T
oI
ΔI+
2
+ |
om
v¡T
oI
ΔI+
2
|
1¡2

(C.7)
where

om
v¡T
oI
= o + 2cI
(C.8)
and

om
v¡T
oI
= b + 2JI
(C.9)
The error associated with the deviation of the fit equation from the actual data is given by
Δm
uev
= |m
a
- m
p
| (C.10)
Applying this uncertainty analysis to the data set, a maximum error of 4.0% was
calculated and imposed over the entire calibration range.
C.2.1. Calibration Procedure
1. Connect monitor, keyboard, and mouse to extensions in Test Cell 4.
2. Activate LabVIEW code entitled “Flow meter Calibration.”
3. Ensure three-way valve is in by-pass mode (flow diverting back to tank).
4. Turn on Lauda calibration bath and set to 20°C. Turn on the motor control unit.
5. Using the LabVIEW program, increase the motor control input voltage to the
desired setting (1.0).
6. Allow the flow meter to come to temperature using PAO from the calibration bath
that is by-passed back to the bath.
99
7. Take the empty 2 L flask and place on the scale. Tare the scale. Remove the
flask from the scale and place under the three-way valve.
8. When ready, quickly move the three-way valve into measurement mode while
simultaneously depressing the “Timer” button on the LabVIEW front panel.
9. Allow the flask to fill for 45 seconds for pump voltage settings one to five, and
for 30 seconds for pump voltages settings six to ten.
10. Quickly return the three-way valve to by-pass mode while simultaneously
depressing the “Timer” button on the LabVIEW front panel.
11. User will be prompted to enter the mass collected. Carefully place the filled flask
on the scale and record the mass in the program. Return the collected PAO to the
calibration bath.
12. Repeat steps 6 through 12, increasing the motor control input voltage by
increments of 1.0V up to 10.0V for a given temperature, then increasing the
temperature by 25°C up to 120°C.
100

Figure C.1. LabVIEW sub-VI wire diagram for RTD read.





Figure C.2. LabVIEW sub-VI wire diagram for calibration bath temperature set.
101

Figure C.3. LabVIEW sub-VI wire diagram for calibration bath temperature read.
102

(a)

(b)
Figure C.4. LabVIEW VI for controlling the automatic thermocouple calibration: (a)
Front panel; (b) Wire diagram
103

(a)

(b)
Figure C.5. LabVIEW VI for manual thermocouple calibration: (a) Front panel; (b)
Wire diagram.
104

Figure C.6. RTD temperature vs. time from the thermocouple calibration procedure.
0
50
100
150
200
250
0 20 40 60 80
Time (hrs)
R
T
D

T
e
m
p
e
r
a
t
u
r
e

(
°
C
)
105

Figure C.7. Sample RTD vs. thermocouple plot for TC00.
0
20
40
60
80
100
120
140
160
0 50 100 150
R
T
D
TC00
106

(a)

(b)
Figure C.8. LabVIEW VI for flow meter calibration program: (a) Front panel; (b) Wire
diagram.

107

Figure C.9. Schematic of flow meter calibration loop

108

Figure C.10. Temperature and flow meter voltage versus mass flow rate calibration
curve for the high-temperature fluid loop flow meter.

2.8
4.4
6
7.6
9.2
0
0.005
0.01
0.015
0.02
0.025
0.03
20
45
70
95
120
V (V)
(
k
g
/
s
)
T (C)
109

(a)

(b)
Figure C.11. Sample data collected during one time run for the flow meter calibration.
(a) “Shotgun Blast” good data set; (b) “Trend” bad data set.


110
Table C.1. Maximum deviation of calculated RTD and experimental RTD corresponding
to each order of polynomial for thermocouple TC00.

Polynomial Order First Second Third Fourth Fifth
Maximum Deviation (°C) 0.48 0.45 0.16 0.15 0.11
111
Table C.2. Coefficients for the trend line of each thermocouple.
5
5
4
4
3
3
2
2 1 0 TC TC TC TC TC actual
T a T a T a T a T a a T + + + + + =


T
h
e
r
m
o
c
o
u
p
l
e
a
0
a
1
a
2
a
3
a
4
a
5
T
C
0
0

(
C
P

I
n
)
4
.
5
7
0
6
0
4
4
5
4
2
8
7
2
9
E
-
0
1
1
.
0
5
6
5
2
8
3
5
0
5
4
6
0
4
E
+
0
0
-
3
.
3
8
1
9
0
5
2
0
8
7
5
3
4
8
E
-
0
3
6
.
0
2
3
0
2
2
0
3
9
1
2
0
9
7
E
-
0
5
-
4
.
2
2
2
5
0
8
6
0
1
1
4
0
7
3
E
-
0
7
1
.
0
1
7
1
9
3
7
0
7
5
7
3
7
1
E
-
0
9
T
C
0
1

(
C
P

O
u
t
)
-
6
.
7
1
7
3
1
6
7
0
2
4
8
6
7
5
E
-
0
2
1
.
0
9
1
6
2
4
8
1
7
2
3
3
4
4
E
+
0
0
-
4
.
4
5
6
9
0
7
2
2
9
8
0
9
8
8
E
-
0
3
7
.
4
5
5
9
6
9
5
8
2
9
7
9
6
4
E
-
0
5
-
5
.
1
2
7
5
3
8
7
0
6
1
8
8
4
0
E
-
0
7
1
.
2
3
5
3
2
2
2
7
9
4
1
9
3
4
E
-
0
9
T
C
0
2

(
B
o
x

A
m
b
i
e
n
t
)
1
.
6
7
6
0
8
8
5
1
8
5
7
4
9
5
E
-
0
1
1
.
0
8
3
9
5
5
8
8
4
9
5
7
3
2
E
+
0
0
-
4
.
1
8
1
3
2
5
7
4
7
9
8
6
0
7
E
-
0
3
7
.
0
7
1
1
1
1
0
4
3
9
3
3
8
5
E
-
0
5
-
4
.
8
8
3
4
4
3
9
9
3
9
5
7
5
6
E
-
0
7
1
.
1
7
8
4
1
7
6
6
4
7
7
2
6
8
E
-
0
9
T
C
0
3

(
F
l
o
w

M
e
t
e
r

I
n
)
3
.
4
1
5
7
1
3
4
8
8
2
4
2
3
6
E
-
0
2
1
.
1
1
2
7
1
5
8
3
4
3
0
9
7
9
E
+
0
0
-
5
.
0
2
4
0
4
1
8
3
0
3
4
7
9
7
E
-
0
3
8
.
0
8
0
8
2
5
8
5
1
0
7
7
4
9
E
-
0
5
-
5
.
4
3
8
5
8
6
3
3
6
2
8
8
2
3
E
-
0
7
1
.
2
9
5
1
8
2
9
4
7
9
7
8
9
3
E
-
0
9
T
C
0
4

(
E
v
a
p

1
)
3
.
8
8
5
3
8
9
7
7
2
2
1
3
8
8
E
+
0
0
7
.
9
6
3
0
0
8
5
0
2
4
3
8
1
9
E
-
0
1
4
.
2
4
5
4
7
6
0
6
7
2
1
5
5
2
E
-
0
3
-
3
.
8
8
6
7
7
4
6
1
3
0
7
0
5
8
E
-
0
5
1
.
5
8
2
5
5
4
1
6
0
1
5
7
2
1
E
-
0
7
-
2
.
3
4
3
8
2
5
4
8
0
3
1
1
7
4
E
-
1
0
T
C
0
5

(
E
v
a
p

2
)
3
.
8
4
6
6
6
3
9
9
1
5
7
7
8
2
E
+
0
0
7
.
7
9
3
8
9
2
9
5
9
1
2
6
4
9
E
-
0
1
4
.
5
8
7
8
1
1
5
8
4
2
7
0
4
7
E
-
0
3
-
4
.
1
9
1
4
7
4
8
3
3
6
8
6
8
3
E
-
0
5
1
.
7
0
6
0
7
3
6
9
1
4
9
7
0
0
E
-
0
7
-
2
.
5
3
2
7
2
1
5
5
2
6
2
5
4
5
E
-
1
0
T
C
0
6

(
E
v
a
p

3
)
3
.
4
3
1
4
4
5
9
6
5
2
0
9
7
6
E
+
0
0
7
.
8
0
1
3
5
7
9
6
6
3
1
3
5
8
E
-
0
1
4
.
5
9
3
2
9
6
5
7
8
4
4
7
3
2
E
-
0
3
-
4
.
2
0
5
1
9
2
2
9
3
1
0
9
7
1
E
-
0
5
1
.
7
1
4
7
8
4
2
4
6
7
0
1
1
2
E
-
0
7
-
2
.
5
4
9
3
3
1
9
9
7
0
6
3
5
4
E
-
1
0
T
C
0
7

(
E
v
a
p

4
)
3
.
1
3
9
7
9
0
2
7
6
3
2
6
9
9
E
+
0
0
7
.
9
0
8
3
6
9
0
9
1
0
5
5
6
8
E
-
0
1
4
.
2
8
5
5
6
6
5
2
7
2
3
2
6
3
E
-
0
3
-
3
.
8
7
5
7
3
1
8
3
1
7
1
7
3
1
E
-
0
5
1
.
5
6
2
7
1
5
2
5
7
0
0
2
9
4
E
-
0
7
-
2
.
2
9
5
8
1
1
2
1
0
6
1
4
0
1
E
-
1
0
T
C
0
8

(
E
v
a
p

O
u
t
)
3
.
6
9
9
1
5
2
8
8
5
1
7
7
1
6
E
+
0
0
7
.
8
2
1
7
9
3
8
2
1
4
8
3
2
2
E
-
0
1
4
.
5
7
9
2
8
2
8
6
2
3
2
7
1
3
E
-
0
3
-
4
.
2
1
6
7
8
3
4
3
2
2
8
0
0
3
E
-
0
5
1
.
7
2
7
8
6
9
0
5
0
6
7
9
6
3
E
-
0
7
-
2
.
5
8
0
2
6
0
5
7
0
4
3
2
3
8
E
-
1
0
T
C
0
9

(
C
o
n
d

I
n
)
3
.
4
9
0
9
9
2
1
9
2
7
7
0
4
8
E
+
0
0
7
.
7
0
7
8
0
1
1
8
6
0
8
1
8
0
E
-
0
1
4
.
8
5
1
0
9
6
0
0
1
1
8
2
9
2
E
-
0
3
-
4
.
4
6
6
9
3
7
4
2
7
5
3
0
9
4
E
-
0
5
1
.
8
2
7
3
2
2
8
4
3
6
2
7
9
4
E
-
0
7
-
2
.
7
2
4
1
1
0
9
7
5
0
4
2
9
1
E
-
1
0
T
C
1
0

(
C
o
n
d

1
)
3
.
5
8
9
7
4
9
6
2
5
3
9
8
7
3
E
+
0
0
7
.
8
3
4
7
2
1
4
9
0
1
6
7
4
3
E
-
0
1
4
.
5
9
6
3
0
3
1
8
0
2
5
5
5
2
E
-
0
3
-
4
.
2
5
0
6
7
8
5
9
7
5
5
0
8
8
E
-
0
5
1
.
7
4
4
3
8
7
3
2
8
2
0
0
7
8
E
-
0
7
-
2
.
6
0
5
0
0
1
6
7
9
7
2
6
6
0
E
-
1
0
T
C
1
1

(
C
o
n
d

2
)
3
.
4
8
4
7
3
3
2
5
8
3
3
0
6
2
E
+
0
0
7
.
8
1
2
2
2
0
7
9
0
6
8
2
0
5
E
-
0
1
4
.
5
2
7
0
2
3
1
7
8
7
2
8
4
3
E
-
0
3
-
4
.
1
2
5
5
2
0
6
2
7
2
2
0
1
7
E
-
0
5
1
.
6
7
4
8
7
6
0
8
2
9
0
5
6
7
E
-
0
7
-
2
.
4
7
8
2
7
0
3
3
7
3
6
0
2
3
E
-
1
0
T
C
1
2

(
C
o
n
d

3
)
3
.
5
2
1
2
1
5
6
0
4
8
5
0
2
6
E
+
0
0
7
.
6
5
7
7
2
9
8
9
7
1
9
9
6
4
E
-
0
1
4
.
9
8
6
0
1
4
7
7
9
1
7
8
7
0
E
-
0
3
-
4
.
6
0
6
0
4
8
5
7
8
3
0
8
7
5
E
-
0
5
1
.
8
8
9
8
1
1
9
4
0
0
7
8
7
8
E
-
0
7
-
2
.
8
2
5
2
5
3
0
5
1
7
3
1
0
7
E
-
1
0
T
C
1
3

(
C
o
n
d

O
u
t
)
3
.
3
7
7
5
8
5
2
8
4
0
0
6
9
1
E
+
0
0
7
.
7
5
6
9
0
8
3
3
7
2
5
9
0
1
E
-
0
1
4
.
6
3
1
7
4
3
4
6
0
3
9
3
0
7
E
-
0
3
-
4
.
1
9
5
0
8
1
3
3
6
6
5
5
3
7
E
-
0
5
1
.
6
9
0
6
9
4
1
4
3
6
5
5
1
6
E
-
0
7
-
2
.
4
8
3
4
2
0
5
8
1
8
0
8
2
7
E
-
1
0
T
C
1
4

(
B
a
y
o
n
e
t

I
n
)
3
.
6
0
2
3
3
6
4
7
7
1
0
0
3
8
E
+
0
0
7
.
7
5
8
4
3
0
5
2
0
6
6
8
5
7
E
-
0
1
4
.
6
4
6
8
8
8
5
0
2
1
2
1
1
6
E
-
0
3
-
4
.
2
4
2
9
9
7
0
4
0
1
2
7
9
3
E
-
0
5
1
.
7
2
5
3
9
8
1
3
4
5
3
3
3
9
E
-
0
7
-
2
.
5
5
8
4
0
7
7
1
1
5
5
4
4
6
E
-
1
0
T
C
1
5

(
C
C
/
E
v
a
p
)
3
.
4
8
3
4
6
9
2
1
5
1
1
4
4
7
E
+
0
0
7
.
7
4
2
8
4
3
6
0
4
0
1
1
7
2
E
-
0
1
4
.
6
6
4
9
8
6
3
7
7
3
8
9
6
4
E
-
0
3
-
4
.
2
4
4
1
4
3
3
0
9
5
9
7
6
1
E
-
0
5
1
.
7
2
0
8
7
4
0
2
5
9
5
7
2
4
E
-
0
7
-
2
.
5
4
5
0
8
8
6
9
5
5
9
3
2
5
E
-
1
0
112
Table C.3. Maximum deviation and total error for each thermocouple.
Thermocouple
Maximum
Deviation ±Total Error
TC00 (CP In) 0.106414 0.122388
TC01 (CP Out) 0.098267 0.111181
TC02 (Box Ambient) 0.097236 0.111085
TC03 (Flow Meter In) 0.098736 0.111083
TC04 (Evap 1) 0.295217 0.307238
TC05 (Evap 2) 0.280751 0.293427
TC06 (Evap 3) 0.280739 0.292962
TC07 (Evap 4) 0.322196 0.334702
TC08 (Evap Out) 0.315632 0.328080
TC09 (Cond In) 0.274506 0.288721
TC10 (Cond 1) 0.278507 0.291299
TC11 (Cond 2) 0.278103 0.290722
TC12 (Cond 3) 0.256994 0.269226
TC13 (Cond Out) 0.291867 0.304493
TC14 (Bayonet In) 0.285927 0.298477
TC15 (CC/Evap) 0.285186 0.297719


113
Table C.4. 3-D paraboloid regression equation for high-temperature fluid loop flow
meter.

2 2
0
dV cT bV aT y m + + + + = &

y
0
2.07704738E-03
a -4.69012732E-05
b 2.35000226E-03
c 1.91650117E-07
d 1.49811559E-05

114
APPENDIX D. LOOP HEAT PIPE MOUNTING
The loop heat pipe was mounted onto the centrifuge table such that the centerline
of the tubing coincided with the outer table radius as much as possible. A small deviation
existed since the condenser section and the evaporator/compensation chamber were both
straight. This induced a non-uniform radial acceleration field over the lengths of these
sections that needed to be quantified. As shown in Figure D.1, the straight condenser
section, with length L
c
, was geometrically aligned on the table so that the centerline was
as close to the table radius, R
ct
, as much as possible over its length. The radius of the
condenser midpoint, R
cm
, was found as shown in Figure D.1 as well. The condenser
endpoints were first set to coincide with the centrifuge table radius. The half-angle is
given by
0 = sin
-1
|
I
c
2R
ct
1 (D.1)
The length from the center of the centrifuge table to the midpoint of the condenser is
I
cm
= R
ct
cos 0 (D.2)
One-half of the change in radius from this point to the centrifuge table radius is used to
determine the radius of the midway point of the condenser section.

R
cm
= I
i
+
1
2
(R
ct
- I
cm
)
(D.3)
The evaporator section and the compensation chamber were also straight and the method
to locate these components in relation to the outer table radius is similar to that described
above for the condenser section. However, the evaporator section is directly attached to
the compensation chamber and the length of the evaporator is different than that of the
compensation chamber. Therefore, further care was taken in determining the maximum
deviation of the centerline radius of these two components from the radius of the
centrifuge table.
115
Stands were designed using G-7 phenolic to mount the loop heat pipe vertically
with support at the compensation chamber, evaporator, condenser, and two on the
transport lines (Figure D.2). The tops of these stands were anchored to the table to
reduce deflection when the table was rotating. A survey was taken at 22 locations on the
loop heat pipe to determine how far various portions of the loop heat pipe were from the
centerline radius and at what distance that location was from the origin at point 1 (Figure
2.6(d)). The loop heat pipe had a minimum radius of 119.2 cm at locations 5 and 6 and
maximum radius of 123.3 cm at location 15 at the outside edge of the tubing at each
location. The entire loop heat pipe fitted within 4.6 cm for a percent radial difference of
3.7%. Complete survey data can be seen in Table D.1.

116

Figure D.1. Mounting of LHP to minimize acceleration gradient.

c
L
2
1
cm
R
R
cm
L
1
L - R r = Δ
ct
R
117

Figure D.2. LHP survey locations.


1
2
3
4 5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
118
Table D.1. Loop heat pipe mounting survey data.
Survey Location s (cm) r (cm) (tube centerline)
1 0.0 123.0
2 15.9 122.9
3 26.0 122.2
4 35.9 121.5
5 35.9 120.5
6 35.9 120.5
7 47.6 120.5
8 59.7 121.1
9 59.7 122.1
10 59.7 120.8
11 59.7 121.1
12 64.1 123.0
13 76.5 121.1
14 76.5 121.1
15 92.4 123.4
16 93.3 121.1
17 93.3 121.1
18 93.3 122.7
19 97.2 122.7
20 101.6 122.1
21 119.1 121.1
22 136.8 122.1


119
APPENDIX E. BRAYCO MICRONIC 889 TECHNICAL DATA
Technical data for various properties against temperature for Brayco Micronic
889, including density, specific heat at constant pressure, thermal conductivity, and
kinematic viscosity were provided by Ghajar et al (1994). Brayco Micronic 889 technical
data from Castrol was compiled and the data was curve-fitted using a least squares
approach and varying order polynomials. The equations are only valid on the
temperature range -54 ≤ T ≤ 135ۤ°C. The authors report that for ρ, C
p
, and k a maximum
deviation of less than ±0.5% from the reported data for the property equations. For ν,
they reported a maximum deviation of +3.4%. The equations for these properties are
given as

µ = 1.S6 · 1u
3
- 4.S6I + u.u1S7I
2
- u.28u · 1u
-4
I
3
+u.174 · 1u
-7
I
4

(E.1)
k = u.1S4 - S.88 · 1u
-5
I (E.2)
C
p
= 1.u22 + S.77 · 1u
-3
I (E.3)
For all of these equations, T is in Kelvin, ρ is in kg/m
3
, C
p
is in kJ/kg-K, and k is in W/m-
K.
120



(c)
Figure E.1. Brayco Micronic 889 properties vs. temperature. (a) ρ vs. T; (b) k vs. T; (c)
C
p
vs. T.
0
100
200
300
400
500
600
700
800
900
1000
ρ
(
k
g
/
m
3
)
(a)
0.128
0.13
0.132
0.134
0.136
0.138
0.14
0.142
k

(
W
/
m
-
k
)
(b)
0
500
1000
1500
2000
2500
3000
-100 -50 0 50 100 150
C
p
(
J
/
k
g
-
k
)
T (°C)
(c)
121
APPENDIX F. CENTRIFUGE TABLE UPGRADES
The previous centrifuge table data acquisition system dated back to the early
1990’s. The original computer was a Pentium 386 running ViewDAC for data collection
and reduction. At the start of this project, it was determined that the data acquisition
system needed to be upgraded, including the data acquisition unit and computer system.
Since all of this equipment was going to be upgraded, it was decided to completely
evaluate the existing data acquisition wiring and document this information.
The first step to updating the data acquisition system was to record the original
data acquisition wiring. All of the wiring on the centrifuge table was rewired and
documented so that it would be easier to trace wiring back to the centrifuge table control
room. The wiring scheme on the centrifuge table can be seen in Figure F.1. The terminal
strip at the bottom of this photo is located on the rotating table, while the terminal strips
at the top are on the stationary support. These are connected via a 40-ring slip ring. A
new wiring panel was developed for the wiring coming from the stationary terminal strips
above the centrifuge table back to the control room, as shown in Figure F.2. Wiring on
the centrifuge table terminal strip now matches the wiring coming into the centrifuge
table control room. This information was completely documented for future reference.
Each circuit from the centrifuge table to the slip ring wiring panel in the control room
was checked for continuity by hand.
After verifying the wiring configuration was in proper operating condition,
documentation for the new data acquisition hardware was reviewed so that the upgrades
could be started. Initially, a new wiring panel was developed for the new data acquisition
interface as shown in Figure F.3. This panel was designed to accommodate 64 channels
(three wires per channel) for data acquisition, as well as 16 channels (four wires per
channel) for voltage and current control. Each circuit from the data acquisition interface
panel to the two new data acquisition modules was checked for continuity by hand.
Jumper cables were created to transfer signals coming in from the slip rings to the data
acquisition wiring panel.
122
The new data acquisition system from Agilent Technologies has a mainframe
(E8408A) with four slots into which the following cards were installed: a command
module (E1406A), an 8/16 channel D/A converter (E1418A), a 5½-digit multimeter
(E1411B), and a sixty-four channel, 3-wire multiplexer (E1476A). The command
module serves as the main source of communication between all of the cards in the
mainframe. The command module also exchanges data and commands between the
computer and data acquisition system. The D/A converter is a control type module,
allowing the user to request a certain output voltage or current from up to sixteen
channels. The multimeter reads the voltages that the multiplexer collects from
thermocouples, pressure transducers, accelerometers, etc., as well as any externally
applied voltages, currents, and resistances. Communication between the data acquisition
system and the computer takes place via the general purpose interface bus (GPIB) or
IEEE-488 protocol. Essentially, text commands are sent from the computer to the data
acquisition system. Then, if the command requests a control signal, the proper output is
processed. If the command were for data acquisition, then ASCII data is returned to the
user for processing.
With the new data acquisition system and computer assembled, documentation for
the data acquisition system needed to be reviewed to determine the proper commands
necessary to use the computer to communicate with the data acquisition system. Initially,
single text lines were sent from the computer using Agilent’s VISA Assistant software.
Commands were sent to read thermocouple temperatures on one of the channels of the
multiplexer.
The software used for writing the data acquisition code was LabVIEW, a visual
computer language. Virtual instruments (VIs) for communicating with the data
acquisition system started fairly crudely. Virtual instruments are subprograms that are
written with certain inputs and calculated outputs, which can greatly simplify a
complicated code. First, the task of reading several voltage channels and outputting the
data to the screen was accomplished. Second, the reading of several voltage channels
was placed inside a timed loop such that data would be recorded at regular intervals and
written to a file that Microsoft Excel could read. Next, the proper conversions for voltage
123
to temperature, acceleration, and pressure were written into the code so that actual data
was recorded to a file.
The next task to be accomplished with this software was control communication.
What was desired was a system where a certain voltage could be applied to increase the
angular velocity of the centrifuge table, and to control a variety of other devices, such as
pumps, heaters, etc. Following a similar process to the development of the data
acquisition software, a virtual instrument was written that could control the output
voltage of several channels along with the capability of turning them on and off at any
time. This VI was then merged with the data acquisition program with appropriate
Boolean commands for control. This VI was tested and verified when data was collected
for a liquid-vapor separator experiment on the centrifuge table. A program was tailored
for this experiment, including the appropriate flow meter, pressure transducer,
accelerometer conversions, and data recording.
After control of the centrifuge table was accomplished using a voltage from the
D/A converter to control the angular velocity, it was decided to control the acceleration
directly. A relation was then developed between voltage and acceleration. Centrifugal
acceleration is given by
o
i
+
=

2
g
(F.1)
Voltage is related to angular velocity by
I = Bæ (F.2)
where B is an experimental constant to be determined. Substituting this relation into the
expression for centrifugal acceleration yields

o
i
+
=
r
g
|
I
B
1
2

(F.3)
Solving equation (F.3) for voltage yields the relation used for deriving a corresponding
voltage for a chosen centrifugal acceleration.

I = B

o
i
+
g
r

(F.4)
124
Thus, voltage is linearly related to the square root of centrifugal acceleration. This
relation can be extended to relate voltage and the acceleration magnitude. The magnitude
of acceleration is given by

o
+
=
¹
o
θ
+
2
+ o
i
+
2
+ o
z
+
2

(F.5)
where
+
θ
a ,
+
r
a , and
+
z
a are the accelerations normalized by gravity in the azimuthal,
radial, and axial directions on the centrifuge table. Solving for the radial acceleration
gives

o
i
+
2
= o
+
2
- o
θ
+
2
- o
z
+
2

(F.6)
When the centrifuge table is rotating with a constant velocity,
+
θ
a = 0 and
+
z
a = −1. After
substituting these values in for equation (F.6) and solving for the square root of
centrifugal acceleration
¹o
i
+
2
= (o
+
2
- 1)
1¡4

(F.7)
Substituting equation (F.7) into (F.4) yields the relation between table voltage and the
magnitude of the acceleration.

I
ct
= B¹
g
r
(o
+
2
- 1)
1¡4

(F.8)
To calculate B, experimental data relating table voltage with centrifugal acceleration was
collected. Then, a plot of voltage versus the square root of centrifugal acceleration was
generated and a linear best fit regression was derived with the voltage intercept forced
through the origin. The corresponding slope is then r g B / . A sample plot can be seen
in Figure F.4. After the slope for this equation was found, an acceleration control slide
bar was added to the data acquisition program by deriving a corresponding voltage
output. It is important to note that the slope is experiment specific, and if the location of
the accelerometer is changed, a new slope needs to be found.
It was desired to have the capability of reading higher temperatures on the
centrifuge table. The current thermocouple amplifier on the centrifuge table is for Type
T thermocouples, which have an operating temperature range between -250 to 350°C. A
125
new Type E thermocouple amplifier (Omega OM7-47-E-07-2-C) has been installed on
the underside of the centrifuge table opposite of the existing thermocouple amplifier so
that either one can be used, depending on the experimental requirements specified,
providing operating temperatures between -200 to 900°C.
126

Figure F.1. Updated wiring on the centrifuge table.
127

Figure F.2. Wiring panel from centrifuge table to the centrifuge table control room.
128

Figure F.3. Wiring panel for the new data acquisition system.
129

Figure F.4. Centrifuge table voltage versus
+
r
a .

0
0.5
1
1.5
2
2.5
3
3.5
0 1 2 3 4
O
u
t
p
u
t

T
a
b
l
e

V
o
l
t
a
g
e

(
V
)
130
APPENDIX G. LABVIEW PROGRAMS

131

(a)

(b)
Figure G.1. LabVIEW VI for the LHP experiment: (a) Front panel; (b) Wire diagram.

132

(a)

(b)
Figure G.2. LabVIEW sub-VI wire diagram for voltage output control: (a) Output on;
(b) Output off.


133

Figure G.3. LabVIEW sub-VI wire diagram for data acquisition communication.

134

Figure G.4. LabVIEW sub-VI wire diagram for data analyzing.
135

Figure G.5. LabVIEW sub-VI wire diagram for data recording.

136
APPENDIX H. CENTRIFUGE WIRING TABLES

137
Table H.1. E1418A 8/16-CH D/A Converter wiring.



Module
Terminal
Number
Terminal
Letter
Wire
Bundle
Number
Wire
Color
Terminal
Strip
Number
Terminal
Number
Function Slip Ring Wire Color
HS red 01
HI white 02 Centrifuge Table Voltage (+) Green-Blue Shield
LO blue 03 Centrifuge Table Voltage (-) Black-Blue Shield
LS yellow 04
HS green 05
HI brown 06 Heater Power Voltage (+) Red-Red Shield
LO purple 07 Heater Power Voltage (-) Black-Red Shield
LS orange 08
HS red 09
HI white 10 Pump Voltage (+) 43 Red-Red Shield
LO blue 11 Pump Voltage (-) 44 Black-Red Shield
LS yellow 12
HS green 13
HI brown 14
LO purple 15
LS orange 16
HS red 01
HI white 02
LO blue 03
LS yellow 04
HS green 05
HI brown 06
LO purple 07
LS orange 08
HS red 09
HI white 10
LO blue 11
LS yellow 12
HS green 13
HI brown 14
LO purple 15
LS orange 16
HS red 01
HI white 02
LO blue 03
LS yellow 04
HS green 05
HI brown 06
LO purple 07
LS orange 08
HS red 09
HI white 10
LO blue 11
LS yellow 12
HS green 13
HI brown 14
LO purple 15
LS orange 16
HS red 01
HI white 02
LO blue 03
LS yellow 04
HS green 05
HI brown 06
LO purple 07
LS orange 08
HS red 09
HI white 10
LO blue 11
LS yellow 12
HS green 13
HI brown 14
LO purple 15
LS orange 16
GND red 01
EXT TRIGn white 02
GND blue 03
GND yellow 04
CAL HS green 05
CAL HI brown 06
CAL LO purple 07
CAL LS orange 08
3
4
24
1
2
23
7
8
9
26
6
25
5
31
12
13
28
10
11
27
23
24
25
26
27
14
15
29
16
30
138
Table H.2. Data acqusition terminal board wiring.
E1476A
Module
Terminal
Number
E1476A
Terminal
Letter
Main
Terminal
Strip
Number
Main
Terminal
Number
Function
Slip Ring
Number
Wire Color
Board
Terminal
Number
Location
H 01 TC00 3 red 01 CP In
L 02 TC Ground 19 black GND
G 03
H 04 TC01 4 white 02 CP Out
L 05 TC Ground 19 black GND
G 06
H 07 TC02 5 blue 03 Box Ambient
L 08 TC Ground 19 black GND
G 09
H 01 TC03 6 yellow 04 Flow meter In
L 02 TC Ground 19 black GND
G 03
H 04 TC04 7 green 05 Evap 1
L 05 TC Ground 19 black GND
G 06
H 07 TC05 8 brown 06 Evap 2
L 08 TC Ground 19 black GND
G 09
H 01 TC06 9 purple 07 Evap 3
L 02 TC Ground 19 black GND
G 03
H 04 TC07 10 orange 08 Evap 4
L 05 TC Ground 19 black GND
G 06
H 07 TC08 11 red 09 Evap Out
L 08 TC Ground 19 black GND
G 09
H 01 TC09 12 white 10 Cond In
L 02 TC Ground 19 black GND
G 03
H 04 TC10 13 blue 11 Cond 1
L 05 TC Ground 19 black GND
G 06
H 07 TC11 14 yellow 12 Cond 2
L 08 TC Ground 19 black GND
G 09
H 01 TC12 15 green 13 Cond 3
L 02 TC Ground 19 black GND
G 03
H 04 TC13 16 brown 14 Cond 4
L 05 TC Ground 19 black GND
G 06
H 07 TC14 17 purple 15 Bayonet In
L 08 TC Ground 19 black GND
G 09
H 01 TC15 18 orange 16 Evap/CC
L 02 TC Ground 19 black GND
G 03
H 04 Accel x-axis 21 red 2
L 05 Accel GND 20 black 1
G 06
H 07 Accel y-axis 22 white 3
L 08 Accel GND 20 black 1
G 09
H 01 Accel z-axis 23 blue 4
L 02 Accel GND 20 black 1
G 03
H 04 Flowmeter 24 green
L 05 Flowmeter GND 25 black
G 06
H 07 Evap Heater Resistor Voltage (+) 26 red
L 08 Evap Heater Resistor Voltage (-) 27 orange
G 09
H 01 Evap Heater Voltage (+) 28 white
L 02 Evap Heater Voltage (-) 29 brown
G 03
H 04 Preheater Resistor Voltage (+) 30 yellow
L 05 Preheater Resistor Voltage (-) 31 blue
G 06
H 07 Preheater Voltage (+) 32 red
L 08 Preheater Voltage (-) 33 black
G 09
H 01 CC Heater Resistor Voltage (+) 34 gray
L 02 CC Heater Resistor Voltage (-) 35 purple
G 03
H 04 CC Heater Voltage (+) 36 pink
L 05 CC Heater Voltage (-) 37 tan
G 06
H 07
L 08
G 09
03
02 04
05
00
01 01
02
09
04 10
11
06
03 07
08
15
06 16
17
12
05 13
14
21
08 22
23
18
07 19
20
24
09 25
26
139
APPENDIX I. SAMPLE CALCULATIONS
Example Calculation of Average Heat Transfer Coefficient for Flat Plate Flow
Given:

Altitude: H = 5 km
Mach number: Ma

= 0.8
Wall temperature: T
w
= 135 °C = 408.15 K
Plate length: L = 1.0 m

Calculations:

Freestream temperature:
T

= (7.7664E-4)H
4
– (2.8994E-2)H
3
+ (5.3483E-1)H
2
– (9.5033)H + (4.8507E+1)
T

= 284.37 K

Freestream density:
ρ

= (-4.9336E-6)H
3
+ (2.0898E-3)H
2
– (8.9917E-2)H + 1.0868
ρ

= 0.6870 kg/m
3


Film temperature:
T
*
= T

(0.5 + 0.039 Ma

2
) + 0.5T
w
= (284.37 K)(0.5+0.039(0. 8)
2
)+0.5(408.15 K)
T
*
= 353.36 K

Air density at the film temperature:
ρ
*
= ρ

/(T

/T
*
) = (0.6870 kg/m
3
)(284.37 K)/(353.36 K) = 0.5528 kg/m
3



Freestream speed of sound:
a

= √(γRT

) = √((1.4)(286.9 m
2
/s
2
·K)( 284.37 K)) = 337.97 m/s

Freestream velocity:
U

= Ma

a

= (0.8)(337.97 m/s) = 270.37 m/s

Freestream absolute viscosity (Reference values from Incropera and DeWitt, 2002):
μ

= μ
R
(T

/T
R
)
0.76
= (184.6 × 10
-7
N·s/m
2
)(284.37 K / 300 K)
0.76

μ

= 1.772 × 10
-5
N·s/m
2




Reynolds number:
Re
L
= (ρ

U

L)/μ

= (0.6870 kg/m
3
)(270.37 m/s)(1 m) / (1.772 × 10
-5
N· s/m
2
)
Re
L
= 1.05 × 10
7

[TURBULENT]

140

Prandtl number at the film temperature:
Pr
*
= (-1.2593E-9)T
3
+ (1.7778E-6)T
2
– (9.4177E-4)T + (8.6418E-1)
Pr
*
= 0.6978

Specific heat at the film temperature:
C
p
*
= (4.4444E-7)T
3
– (3.3333E-5)T
2
– (6.9921E-2)T + (1.0187E+3)
C
p
*
= 1009.44 J/(kg·K)

Recovery factor for turbulent flow:
r = Pr
*1/3
= (0.6978)
1/3
= 0.8870

Adiabatic wall temperature:
T
aw
= T

[1+r((γ-1)/2)Ma

2
] = (284.37 K)(1+(0.8870)(0.4/2)(0.8)
2
) = 316.66 K

Absolute viscosity at the film temperature:
μ
*
= μ
R
(T
*
/T
R
)
0.76
= (184.6 × 10
-7
N· s/m
2
)(353.36 K / 300 K)
0.76
= 2.091 × 10
-5
N·s/m
2


Local skin friction coefficient at the end of the plate:
C
f,L
*
= 0.455/(ln
2
(0.06ρ
*
U

L/ μ
*
))
= 0.455/(ln
2
(0.06(0.5528 kg/m
3
)(270.37 m/s)(1 m) / (2.091 × 10
-5
N· s/m
2
))
= 0.002705

Local Stanton number at the end of the plate:
St
L
*
= (C
f,L
*
/2) / (1+12.7(Pr
*2/3
-1)(C
f,L
*
/2)
1/2

= (0.002705/2) / (1+12.7((0.6978)
2/3
-1)(0.002705/2)
1/2
) = 0.001502

Local heat transfer coefficient at the end of the plate:
h
L
= St
L
*
ρ
*
U

C
p
*
= (0.001502)(0.5528 kg/m
3
)(270.37 m/s)(1009.44 J/kg·K)
= 226.7 W/m
2
·K

Average heat transfer coefficient over the length of the plate:
h = 1.15 i
L
= 1.15(226.7 W/m
2
·K) = 260.7 W/m
2
·K

Average heat flux dissipated over the plate:
q
w
= h (T
w
- T
aw
) = (260.7 W/m
2
·K)(408.15 K - 316.66 K) = 23847.7 W/m
2

WRIGHT STATE UNIVERSITY SCHOOL OF GRADUATE STUDIES March 20, 2009 I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY Andrew James Fleming ENTITLED Aircraft Thermal Management Using Loop Heat Pipes BE ACCEPTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of Science in Engineering. ____________________________________ Scott K. Thomas, Ph.D. Thesis Director ____________________________________ George P.G. Huang, P.E., Ph.D. Department Chair Committee on Final Examination ____________________________________ Scott K. Thomas, Ph.D. ____________________________________ Kirk L. Yerkes, Ph.D. ____________________________________ J. Mitch Wolff, Ph.D. ____________________________________ James A. Menart, Ph.D. ____________________________________ Joseph F. Thomas, Jr., Ph.D. Dean, School of Graduate Studies

ABSTRACT Fleming, Andrew James. M.S.Egr., Department of Mechanical and Materials Engineering, Wright State University, 2009. Aircraft Thermal Management using Loop Heat Pipes.

The objective of this thesis was to determine the feasibility of using loop heat pipes to dissipate waste heat from power electronics to the skin of a fighter aircraft and examine the performance characteristics of a titanium-water loop heat pipe under stationary and elevated acceleration fields. In the past, it has been found that the boundary condition at the condenser can be a controlling factor in the overall performance of this type of thermal management scheme. Therefore, the heat transfer removed from the aircraft skin has been determined by modeling the wing as a flat plate at zero-incidence as a function of the following parameters: airspeed: 0.8 ≤ Ma∞ ≤ 1.4; altitude: 0 ≤ H ≤ 22 km; wall temperature: 105 ≤ Tw ≤ 135°C. In addition, the effects of the variable properties of air have been taken into account. Heat transfer due to thermal radiation has been neglected in this analysis due to the low skin temperatures and high airspeeds up to Ma∞ = 1.4. It was observed that flight speed and altitude have a significant effect on the heat transfer abilities from the skin to ambient, with heat rejection becoming more difficult with increasing Mach number or decreasing altitude. An experiment has been developed to examine operating characteristics of a titanium-water loop heat pipe (LHP) under stationary and elevated acceleration fields. The LHP was mounted on a 2.44 m diameter centrifuge table on edge with heat applied to the evaporator via a mica heater and heat rejected using a high-temperature polyalphaolefin coolant loop. The LHP was tested under the following parametric ranges: heat load at the evaporator: 100 ≤ Qin ≤ 600 W; heat load at the compensation chamber: 0 ≤ Qcc ≤ 50 W; radial acceleration: 0 ≤ ar ≤ 10 g. For stationary operation (az = 1.0 g, ar = 0 g), the LHP evaporative heat transfer coefficient decreased monotonically, iii

which was likely due to vapor bubble formation in the primary wick. Heat input to the compensation chamber was found to increase the evaporative heat transfer coefficient and decrease thermal resistance for Qin = 500 W. Operating the LHP in an elevated acceleration environment (az = 1.0 g. Flow reversal in the LHP was found for some cases.thermal resistance decreased to a minimum then increased. and wall superheat increased monotonically. iv . However. Evaporative heat transfer coefficient and thermal resistance was found not to be significantly dependent on radial acceleration. wall superheat was found to increase slightly with radial acceleration. ar > 0 g) revealed dry-out conditions from Qin = 100 to 400 W and varying accelerations and the ability for the LHP to reprime after an acceleration event that induced dry-out.

.............. 27  2.....................................3........ Mathematical Model ............................................ Operating Procedures ................................TABLE OF CONTENTS 1..................... Abstract ..............................5.4.........................................2.............................................. 14  2..........2................................. 39  2............................................................................................................................ Uncertainty Analysis .... 3  1........1.......................................... 19  2............................ Future Work .............1................................... 77  A......................................................................................................... Experimental Setup ...................  Titanium-Water Loop Heat Pipe Characteristics Under Stationary and Elevated Acceleration Fields ........... Test Procedures ...............................................  Convective Heat Transfer from High-Speed Aircraft Skin ........................................... Introduction ................................ 1  1....................... 1  1................................... Abstract ... 14  2...........5................ 89  Appendix C...................................... 1  1.......................................... 5  1................................ Introduction ................................6...................3....... 6  2......................................................................................................2......... 14  2... 40  References ................ Conclusions .... Results and Discussion ...................... Results and Discussion ........................... Standard Operating Procedure . 92  C...................4..................... 74  Appendix A....................................... Thermocouple Calibration....................................................................................................................................................................... Calibration of Thermocouples and Flow Meter ................................. Conclusions ................................ 92  v .............................................................................................. 79  Appendix B.................................1.................. 77  A........................................1.

.................................................... LabVIEW Programs ............................................. Loop Heat Pipe Mounting ................................................. 121  Appendix G............... 95  Appendix D...........................C................................ 130  Appendix H. 114  Appendix E.... Brayco Micronic 889 Technical Data.....................................2............................................................................. 119  Appendix F.... 136  Appendix I......... 139  vi ..... Flow Meter Calibration ........................................................ Centrifuge Table Upgrades .............................................. Sample Calculations . Centrifuge Wiring Tables ...............................................

.. (b) H = 10 km............................. Temperature difference (Tw − Taw ) versus altitude for various Mach numbers (Tw = 135ºC......... ...... Average heat flux dissipated over the plate versus altitude for various atmospheric conditions (Tw = 135ºC......... 1997.. 1% hot day)...............2......... Comparison of atmospheric properties versus altitude: (a) Temperature... 11  Figure 1............. 2000)........98) (DOD. 45  Figure 2........... Adapted and reprinted with permission from AIAA (Hoang and Ku.. 1% hot day). Thermocouple locations on the LHP: (a) Locations of thermocouples TC04 through TC15 across the LHP..................... L = 1... 1% hot day): (a) H = 0 km.......................................6.............................................. 2003)............................................... (b) Cross-sectional view.........................10.......................4..............LIST OF FIGURES Figure 1.... Ma∞ = 0...... ...... ....... 7  Figure 1.................................. ...1........ Anderson..... 43  Figure 2.... ..................1. Titanium-water loop heat pipe test article as delivered.....5.....4.. (b) Density (DOD........... Adiabatic wall temperature versus altitude for various Mach numbers (1% hot day)... 44  Figure 2.......... Average heat flux dissipated over the plate versus altitude for various Mach numbers (Tw = 135ºC.......98..................... 42  Figure 2........................5................... Maximum Mach number before heat is transferred from the air to the skin versus altitude for various wall temperatures (1% hot day).................................... 9  Figure 1........... 2003)...... Local heat flux dissipated over the plate versus plate length for various Mach numbers (Tw = 135ºC.......... 8  Figure 1........ 8  Figure 1.......... L = 1................................... L = 1..... Evaporator schematic: (a) Side view....0 m..................................................3.. 10  Figure 1..... Average convective heat transfer coefficient versus altitude for various Mach numbers (Tw = 135ºC....... L = 1..... 1% hot day)....... Temperature difference (Taw − T∞ ) versus altitude for various Mach numbers (1% hot day)..........0 m.. 2000)......... 9  Figure 1....................8.................... .. 46  vii ......... 10  Figure 1................. (b) Locations of TC04 through TC07 within the evaporator........................ .......2.... Adapted and reprinted with permission from AIAA (Hoang and Ku. 1% hot day)........... .0 m.0 m.3......... Loop heat pipe operation.. 1997.......................... 12  Figure 1................ ..... 12  Figure 2............ Average heat flux dissipation versus altitude for various wall temperatures (Ma∞ = 0...........9... (c) H = 20 km........ ......................... Anderson........................................7.. Schematic of Centrifuge Table Test Bed..............

..... 63... mcp = 0........ (d) Qin = 400 W. mcp = 0.......7°C): (a) Evaporative heat transfer coefficient............1°C): (a) Qcc = 25 W........ 31.. 36... 63. (b) Condenser section........ Transient temperature profiles in the condenser and bayonet tube of & the stationary LHP (Qcc = 0 W. ar = & 0 g... Tamb = 31..1°C): (a) Qin = 100 W...... Tcp = 67... flowmeter..........6 ≤ Tamb ≤ 38... ....... Qcc = 0 W..1 ≤ Tamb ≤ 38..... mcp = 0. 36.... (b) Condenser temperatures... (b) Reservoir.. ..0077 kg/s..7.. TC03. mcp = 0...... and liquid/liquid heat exchanger....7 ≤ Tamb ≤ 38.... Transient temperature profiles of the stationary LHP for Qin = 200 & W (Qcc = 0 W.7°C): (a) Transient temperature profiles... Qcc = 0 W......1°C...8 ≤ Tcp ≤ 67...... Transient temperature profiles in the condenser and bayonet tube of & the stationary LHP for Qcc = 25 to 50 W (Qin = 500 W. ar = 0 g..6............ Steady state temperature distribution versus transported heat for the & stationary LHP (Qcc = 0 W........................... ar = 0 g. (b) Qin = 200 W....8°C... (f) Qcc = 50 W.. and TC01....1°C): (a) Evaporator section..... (e) Qin = 500 W....7°C.........8.......... 47  Figure 2... 49  Figure 2.. 27... Use of a cold-start test to determine when steady state occurred for & the stationary LHP (Qin = 600 W.4 ≤ Tcp ≤ 64... ar = 0 g..... 51  Figure 2...... 36...8°C.7°C....... Tcp = 67.... mcp = 0............ (b) Thermal resistance.........1°C): (a) Evaporator temperatures. Steady state performance characteristics of the stationary LHP & versus compensation chamber heat input (Qin = 500 W........ 53  Figure 2................ front and top views: (a) Evaporator and compensation chamber: (b) Transport lines.........0077 kg/s... filter...8 ≤ Tcp ≤ 71. Tamb = 38....... (c) Evaporative heat transfer coefficient and thermal resistance...... Mounting of loop heat pipe to centrifuge table. (c) Transient thermal resistance and evaporative heat transfer coefficient.........0077 kg/s......... ar = 0 g. 36..9...6°C......7 ≤ Tamb ≤ 38....Figure 2. 48  Figure 2........ ..... 36.... (b) Qcc = 30 W.......... (c) Cold plate..... (b) Transient rate of change of temperatures.........7°C..... (c) Qin = 300 W.............................. 54  Figure 2....0077 kg/s.... mcp = 0.............0077 kg/s..12.11. (c) Wall superheat................. (d) Complete loop heat pipe... Tcp = 46. (c) Qcc = 35 W. ......... ........ 31.... (e) Qcc = 45 W..........6°C........ (f) Qin = 600 W.. (c) Condenser with cold plate.........1 ≤ Tamb ≤ 38.......... ar = 0 g........ 56  viii ....... mcp = 0.....8 ≤ Tcp ≤ 71. (d) Qcc = 40 W.. 55  Figure 2...... pump.0077 kg/s....... 52  Figure 2..1°C): (a) Initial startup................. ar = 0 g..... (b) Complete startup until steady state....... High temperature fluid loop: (a) Schematic......10. Steady state performance characteristics of the stationary LHP & versus transported heat (Qcc = 0 W.....4 ≤ Tcp ≤ 64......14... Tamb = 38. 50  Figure 2.....0077 kg/s... (b) 2φ-1φ point oscillation in the condenser.15.0077 kg/s.1°C): (a) Transient temperature traces... mcp = 0..... ..............13..... ar = 0 g. TC00............ Transient startup of the stationary LHP (Qin = 600 W..

........ mcp = 0.20................ ..7°C........4...8................ 83  Figure A.0 g.........7°C............ and condenser under elevated acceleration (to scale....1 g startup phase..Figure 2...........5........ ........0077 kg/s....... (b) Transient temperature trace at ar = 4... .......... 61  Figure 2..0 g and t = 31240 s........0077 kg/s...... mcp = 0........................................................... (c) Te. Centrifuge table motor control power switch..............1 ≤ Tamb ≤ 30. Quasi-steady state temperature traces of the LHP and cold plate at & elevated acceleration for Qin = 200 W (Qcc = 0 W.....................2 ≤ Tcp ≤ 59. compensation chamber...1 ≤ Tamb ≤ 38.....16..... 57  Figure 2....... . Centrifuge table power switch....................... LabVIEW sub-VI wire diagram for calibration bath temperature read......... 88  Figure C....19. top view)..... 87  Figure A.... Sample LabVIEW control program............ 58  Figure 2............................. Chill bath plumbing schematic................. 86  Figure A....................1....4°C): (a) Transient temperature trace at ar = 0........................ 100  Figure C.................................9°C............... (b) Te..... 37....7°C): (a) Evaporative heat transfer coefficient.................. Tamb = 28..2 ≤ Tcp ≤ 59......7.... Qcc = 0 W............ . 82  Figure A........ 62  Figure A...... Effect of resultant acceleration vector direction on fluid distribution within the LHP: (a) Resultant acceleration vector orientation versus radial acceleration...... ... (c) Transient rate of change of temperatures.max = 150°C... (b) Liquid pooling in the evaporator.1.... Centrifuge table main power breaker: (a) Electrical panel MCC-6.. Steady state performance map of the LHP relating radial & acceleration and heat transported (Qcc = 0 W...0°C): (a) Te.... LabVIEW sub-VI wire diagram for RTD read................ (b) Centrifuge table main power breaker............. Tamb = 26. LabVIEW sub-VI wire diagram for calibration bath temperature set........... 101  ix ......................17........... .. (b) Transition to and steady state at ar = 10........ .....3..........1°C): (a) ar = 0....... 27.......... Tcp = 41.......................7°C...................................2.... Transient temperature traces of the LHP at elevated acceleration & showing dry-out behavior (Qin = 400 W...3.............7°C.9 ≤ Tamb ≤ 30....18.... . 37.....0077 kg/s.......................... ... (c) Liquid pooling in the condenser bends. Qcc = 0 W..... 85  Figure A.................2...... .... mcp = 0..................6...... ....21.... Steady state performance characteristics of the LHP versus & transported heat at stationary and elevated acceleration (Qcc = 0 W................. 55..................... .... Centrifuge table control panel box...... 81  Figure A................... ....max = 175°C.. mcp = 0. 25............. mcp = 0..0077 kg/s............. 37............ 25.........................2 ≤ Tcp ≤ 67.. (c) Wall superheat.......0077 kg/s.................. ..2°C)... Neslab recirculating chiller.. 60  Figure 2.max = 200°C... 100  Figure C. ... (b) Thermal resistance. Transient temperature traces of the LHP at elevated acceleration (Qin & = 600 W. ........2 ≤ Tcp ≤ 59...........1 g and t = 13834 s. 59  Figure 2.. Booster pump control panel...... 84  Figure A..

....... ........... ..... 105  Figure C........................... 108  Figure C... 109  Figure D..... Temperature and flow meter voltage versus mass flow rate calibration curve for the high-temperature fluid loop flow meter......... (b) k vs. LabVIEW VI for flow meter calibration program: (a) Front panel.................. Sample data collected during one time run for the flow meter calibration............. 120  Figure F.... 133  Figure G.....4.............1........................ Centrifuge table voltage versus a r+ ........................10..........4...... 135  x ......... 127  Figure F........................................... (a) “Shotgun Blast” good data set................ ... T.... 132  Figure G................ (b) Wire diagram........................ Brayco Micronic 889 properties vs........... T...... ...........2................. Mounting of LHP to minimize acceleration gradient.......... .......................... 126  Figure F...................... ............... LabVIEW sub-VI wire diagram for voltage output control: (a) Output on.........................................7...... ....................................................11.............. LabVIEW sub-VI wire diagram for data acquisition communication............... . LabVIEW sub-VI wire diagram for data recording.... 106  Figure C.... 131  Figure G.... ...... 128  Figure F.......................8...........1.. ................. LabVIEW VI for manual thermocouple calibration: (a) Front panel.... Wiring panel from centrifuge table to the centrifuge table control room... ............................................. LabVIEW VI for the LHP experiment: (a) Front panel.. (a) ρ vs.............................3................ 107  Figure C........... thermocouple plot for TC00......... (b) Output off..........................................................................Figure C... LabVIEW VI for controlling the automatic thermocouple calibration: (a) Front panel.........1......5.............................. LabVIEW sub-VI wire diagram for data analyzing...... time from the thermocouple calibration procedure......2............................................................ 134  Figure G........... Updated wiring on the centrifuge table............. 104  Figure C........................5.............2................................ RTD temperature vs............. .............. Sample RTD vs.... .. LHP survey locations............ Wiring panel for the new data acquisition system..............9....... 116  Figure D........................ ....................................................... (b) “Trend” bad data set.............................. T................................................. 102  Figure C......... (b) Wire diagram.......1........6.... (c) Cp vs.. (b) Wire diagram.......... (b) Wire diagram ...................4.....3.. ................. 117  Figure E................ temperature..... 103  Figure C.. Schematic of flow meter calibration loop .................. 129  Figure G..

.. Tcp = 52...................................... ...................... .3.................... ACT LHP geometric design parameters...4.............. 118  Table H.... ..4......... 1997).0077 kg/s.. Steady state operating characteristics for the stationary LHP showing & effect of heat input to the compensation chamber (Qin = 500 W.7°C..................................... 111  Table C.. ............................... 69  Table 2.............. 25.3................4 ≤ Tcp ≤ 64.............. 112  Table C.0077 kg/s........................ mcp = 0.....6.....9............... Tamb = 26............. ar = 0 g...8 ≤ Tcp ≤ 71.......5........1............................... 13  Table 2.........0077 kg/s....................... 37... ............................................2....4°C) .... 137  xi .............6°C.............. 65  Table 2...0077 kg/s).................. 63..... 73  Table C............... 66  Table 2........ Regression equations for air properties versus temperature (Incropera and DeWitt..0077 kg/s................. Maximum deviation of calculated RTD and experimental RTD corresponding to each order of polynomial for thermocouple TC00.......................... Comparison of quasi-steady states for Qin = 200 W (Qcc = 0 W............. Loop heat pipe mounting survey data.................... Summary of uncertainties................... 63  Table 2..........................2 ≤ Tcp ≤ 59...... .................. ............2.... Coefficients for the trend line of each thermocouple. mcp = 0..... E1418A 8/16-CH D/A Converter wiring. 3-D paraboloid regression equation for high-temperature fluid loop flow meter...... 67  Table 2..................8................................. ........7°C) showing effect of startup path.1.. 13  Table 1................. Summary of LHP thermocouple locations...1.........................2................................................................8°C........................................... ......................................... 27.. & mcp = 0............ 70  Table 2.......1....... 64  Table 2.. ............1..........1 ≤ Tamb ≤ 30... Steady state operating characteristics of the rotating LHP (Qcc = 0 W.......... 36.............. Maximum deviation and total error for each thermocouple....6 ≤ Tamb ≤ 38.................. mcp = 0.....LIST OF TABLES Table 1.....2°C).........7................................5°C....... .. 113  Table D. ...... The effect of compensation chamber temperature control on LHP & operation (Qin = 500 W..1°C)....... AFRL/RZPS design requirements...... ar = 0 g.. 2002)....... ar = 0 g.. Steady state operating characteristics for the stationary LHP (Qcc = 0 & W...........1 ≤ Tamb ≤ 38......... 110  Table C........ 71  & Table 2.......... 36..................... mcp = 0........ Regression equations for air properties versus altitude for 1% hot (DOD.. ...................................

........... Data acqusition terminal board wiring......Table H.........2.. 138  xii ............................. ...

radial coordinate. flow meter calibration constant Flow meter calibration constant Experimental constant related to Eq. 9. acceleration. m/s.81 m/s2 Heat transfer coefficient. (F. U / a Number of data points Nusselt number. kg Mach number. W/(m2-K) Altitude. m Frequency. W Heat flux. thermal resistance. m Particular gas constant.2) Flow meter calibration constant Skin friction coefficient. K/W Coefficient of determination Rayleigh number. m Mass. J/(kg-K) Flow meter calibration constant Diameter. Hz Acceleration due to standard gravity. m/s2. 2 / Specific heat. m2/(s2-K). / Heat transfer rate.NOMENCLATURE a b B c Speed of sound. Prandtl number. / / Cf Cp d D f g h H k L m Ma n Nu Pr Q q r R R2 Ra xiii . W/m2 Recovery factor. W/(m-K) Length. m Thermal conductivity.

(N-s)/m2 Kinematic viscosity. degrees Absolute or dynamic viscosity. K-1 Ratio of specific heats Temperature difference. Stanton number. 5. volume. t-distribution Temperature. / Time. m2/s Density. L / Flow meter calibration constant Greek Letters α β γ ΔT ε θ μ ν ρ σ φ ω Thermal diffusivity. m/s Voltage. K Velocity. Stefan-Boltzmann constant.Re St t T U V y0 Reynolds number. kg/m3 Standard deviation. s. V. m2/s Inverse temperature.67×10-8 W/(m2-K4) Fluid phase Angular velocity. K Emissivity Angle. rad/s Superscripts * + 1φ 2φ Subscripts 1 2 Heat into vaporization of fluid Heat leak to the compensation chamber xiv Film condition Normalized Single-phase Two-phase .

∞ a amb aw c cc cl cm conv cp ct dev D e eg e/cc ie in L max min m/t oe out p PAO r R rad s sh Freestream condition Actual Ambient Adiabatic wall Condenser Compensation chamber Centerline Condenser midpoint Convection Cold plate Centrifuge table Deviation Diameter Evaporator Ethylene glycol Evaporator/compensation chamber junction Inner edge Cold plate inlet. heat out of condenser Predicted Polyalphaolefin Radial Reference condition Radiation Surface Superheat xv . heat in to evaporator Local Maximum value Minimum value Mass/time Outer edge Cold plate outlet.

surr TC tot v V/T w z θ Surroundings Thermocouple Total Vapor Voltage/temperature Wall Axial Azimuthal xvi .

Ed and Judy. xvii . You are truly worth all my praise. wisdom. thank you to my Lord and Savior Jesus Christ for all the blessings and gifts You have bestowed on me. Bekah Puterbaugh. Travis Michalak. Dave Courson for all of the technical assistance and knowledge you have given to me throughout the duration of experimentation. Kirk Yerkes. Last. and friendship: Mr. Energy/Power/Thermal Division. Propulsion Directorate.ACKNOWLEDGEMENTS This research effort was conducted as a part of the in-house program at the Air Force Research Laboratory. Jennifer. and guidance in helping me be everything I am. Cindy Obringer. I would like to thank Dr. and Dr. Thank you to my parents. support. Levi Elston. for your unfailing love and support through our years together. Dr. for your love. Larry Byrd. AFRL/RZPS. Thank you to Mr. Thermal and Electrochemical Branch. Thank you to my wife and best friend. OH. Mr. Ms. Quinn Leland for the mentoring. Thank you to the thermal crew for your guidance. Wright-Patterson Air Force Base. Scott Thomas. help. Dayton. Dr. but most certainly not least. and knowledge you have given me over the past several years. and Ms.

DEDICATION To my beautiful wife, Jennifer, and son, Ethan. I love you.

xviii

1. CONVECTIVE HEAT TRANSFER FROM HIGH-SPEED AIRCRAFT SKIN 1.1. Abstract The objective of the present analysis was to determine the feasibility of using loop heat pipes to dissipate waste heat from power electronics to the skin of a fighter aircraft. In the past, it has been found that the boundary condition at the condenser can be a controlling factor in the overall performance of this type of thermal management scheme. Therefore, the heat transfer removed from the aircraft skin has been determined by modeling the wing as a flat plate at zero-incidence as a function of the following parameters: airspeed: 0.8 ≤ Ma∞ ≤ 1.4; altitude: 0 ≤ H ≤ 22 km; wall temperature: 105 ≤ Tw ≤ 135°C. In addition, the effects of the variable properties of air have been taken into account. Heat transfer due to thermal radiation has been neglected in this analysis due to the low skin temperatures and high airspeeds up to Ma∞ = 1.4. It was observed that flight speed and altitude have a significant effect on the heat transfer abilities from the skin to ambient, with heat rejection becoming more difficult with increasing Mach number or decreasing altitude. 1.2. Introduction The More Electric Aircraft initiative (MEA) is the concept for future aircraft including warfighter, transport, helicopters, and commercial aircraft. This approach has been adopted by the United States Air Force since the early 1990’s with the purpose of reducing or removing as many of the hydraulic, mechanical, and pneumatic power components and replacing them with electrically driven devices. This approach to aircraft design was first envisioned during World War II. However, at that time, the power generation capability and power conditioning equipment required was not feasible due to volume requirements. As a result, hydraulic, pneumatic, and mechanical systems became the norm for aircraft until this initiative. Under the MEA paradigm, power for systems such as flight control actuation, anti-ice, braking, environmental control, engine starting, and fuel pumping will be provided by a starter/generator driven by the gas 1

generator spool of the aircraft engine (Quigley, 1993). The MEA initiative has been analytically proven to improve aircraft reliability, maintainability, support, and operations cost as well as reduce weight, volume, and enhance battle damage reconfigurability (Cloyd, 1997). While the reduction of hydraulic, pneumatic, and mechanical systems in favor of electrical systems is beneficial, it presents a problem in terms of thermal management. Replacing the centralized hydraulic system with an electrical based system removes a primary method of transporting and removing waste heat (Vrable and Yerkes, 1998). A separate cooling fluid system for thermal management would be contrary to the goals of the MEA initiative. Therefore, thermal management would need to be distributed over the entire aircraft. As a result, a new approach to thermal management involves handling heat loads on a local level. This means taking individual components in the aircraft and locally handling their heat rejection requirements. The operating envelope for military aircraft places stringent limitations on any proposed thermal management system. The on-board electrical flight control actuation system operates at altitudes from sea level to above 12 km, airspeeds from stationary to supersonic speeds, transient body forces up to 9 g due to maneuvering, and ambient temperatures from -68 to 58ºC. MEA has resulted in the development of hightemperature, high-efficiency, and high-density power electronic component technologies. The next-generation power electronics will be capable of operating at cold plate temperature excursions up to 200ºC, which presents an opportunity to reject heat through the aircraft skin to the ambient using passive cooling. In addition, the actuation system rejects heat continuously at a rate of Q = 500 W (q = 3 W/cm2) and has transient heat rejection rates of Q = 5000 W over a period of one second. Possible thermal management scenarios include direct connection of the electronics package to the skin, high-thermal conductivity graphite straps, or the use of a loop heat pipe between the package and the skin to provide mounting flexibility. The objective of this analysis is to determine the external heat transfer possibilities of the aircraft skin. The heat flux and heat transfer coefficient have been found as functions of the skin and ambient temperatures, the altitude, and airspeed.

2

1997) and data for the “standard atmosphere” (Anderson. R (1. ReL ∞ ∞ ∞ (1. 1953) .039Ma 0.5 w (1. 2000). In order to be conservative in the calculation of heat transfer coefficients.3.5 0.4) The absolute viscosity of air is given by the following relation (NACA.2) The freestream speed of sound is ∞ ∞ (1.3) The freestream velocity is ∞ Ma∞ ∞ (1. The Reynolds number for a plate of length L is determined by evaluating the properties of air at the freestream condition.1 and Table 1. 1997) as shown in Figure 1.6) 3 .1) The air density at the film temperature and at altitude was evaluated using the perfect gas law ∞ ∞ (1.5) R where μR is a reference viscosity evaluated at a known reference temperature TR. 1% (DOD.1.1. Mathematical Model The temperature and density of air vary considerably with altitude and also vary day-to-day depending on weather conditions. The film temperature was used as the reference temperature to evaluate the air properties (White. 1988) ∞ Also presented are data for the lowest temperature recorded with a frequency-of-occurrence of 0. data for the highest temperature recorded with a frequency-of-occurrence of 1% were used to generate equations for temperature and density versus altitude (DOD.

000 were turbulent.8) For the purposes of this analysis.332ReL Pr / / (1. The adiabatic wall temperature is (White. as shown in Table 1.L ln 0. The local skin friction coefficient at the end of the plate was found by evaluating the air properties at the film temperature. greater than 500.664 f.Regression equations for the specific heat and Prandtl number were determined as functions of temperature using data from Incropera and DeWitt (2002).13) 4 .12) The local heat transfer coefficient at the end of the plate is L St L ∞ p (1.06 ∞ (1. 1988) 0.11) 1 12.L ⁄2 /   (1. the skin friction coefficient is given by (White.10) The local Stanton number at the end of the plate for laminar flow is given by (White. Reynolds numbers less than 500.7 Pr / 1 f. For laminar flow. 1988) aw ∞ 1 1 2 Ma∞ (1.9) and for turbulent flow f.L ⁄2 0.455 0.2. 1988) St L and for turbulent flow St L L ∞ p f.L ∞ ⁄ (1.7) where the recovery factor is Pr 1/2 for laminar flow Pr 1/3 for turbulent flow (1.000 were considered to be laminar.

The maximum Mach number increases with altitude and wall temperature up to a maximum at approximately 18 km. 1988) 1. The overall trend of the adiabatic wall temperature with altitude follows the freestream air temperature in Figure 1. The maximum Mach number achievable before heat is transferred from the air to the skin is given by Ma∞. In general. The average heat flux dissipated from the plate is shown in Figure 1. which indicates that heat is 5 .5 over a range of wall temperatures. For low Mach numbers.4.The local heat transfer coefficient was calculated using the appropriate skin friction coefficient and Stanton number based on laminar or turbulent flow.7. the heat flux is positive for all values of altitude.4.T∞). the convective heat transfer coefficient increases with Mach number.15) Thermal radiation was neglected in this analysis as it contributed less than 1. Figure 1.6% to the total heat rejected from the plate surface.max 1 w ∞ 1 2 1 / (1. the average convective heat transfer coefficient decreases monotonically with altitude due to the continual decrease in the air density. 1. which indicates that heat is transferred from the air to the aircraft skin. ΔT = (Taw .6. 1988) w w aw (1.2 as a function of altitude and Mach number. versus altitude.14) The heat flux dissipated over the plate. In Figure 1.15 L (1. Results and Discussion The adiabatic wall temperature is shown in Figure 1. as expected. The average heat transfer coefficient over the length of the plate is approximated by (White.3 presents the temperature difference. Of interest is the portion of the curves in which this difference is negative. is defined in terms of the adiabatic wall temperature (White. This temperature difference demonstrates the increase in the adiabatic wall temperature over the freestream due to aerodynamic heating.16) and is plotted in Figure 1. The temperature difference ΔT = (Tw Taw) is given in Figure 1. both local and average.1 and increases with Mach number as expected.

which is directly influenced by the actuator thermal management system. This means that the adiabatic wall temperature is higher than the skin temperature due to aerodynamic heating effects.10. 6 . has a strong effect on the heat dissipation rate.4.9 shows the average heat flux dissipated over the plate versus altitude for the 1% hot day.5. which in turn affected the overall heat transfer coefficient. It was found that the altitude and speed of the aircraft significantly affected the amount of heat that could be rejected from the skin. At low altitudes. It was also shown that the assumption of a “standard atmosphere” could result in significant errors in the prediction of the heat dissipation as compared to the data for the 1% hot day or the 1% cold day. In general. The heat flux increases dramatically with altitude and wall temperature for low altitudes. where hL is high at the leading edge and at the beginning of turbulent flow and decreases as the boundary layer grows.1. the average heat flux follows the behavior of the local heat transfer coefficient. the 1% cold day. The analysis showed that the aircraft skin temperature. and then increases as the flow transitions to turbulent where it once again decreases. heat transfer from the skin to the air went to zero. 10. and 20 km is shown in Figure 1. The altitude of the aircraft affected the freestream temperature and density. however. and if the Mach number was high enough. At high Mach numbers.8. The effect of wall temperature on average heat flux for a given airspeed is shown in Figure 1. A performance map of this phenomenon was provided. and the standard atmosphere data as presented in Figure 1. especially at low altitudes. The effect of heated plate length on the local heat flux for H = 0. the heat flux is negative at low altitudes due to the negative ΔT as shown in Figure 1. 1. The local heat flux starts low and decreases in the laminar region of the plate. q w is significantly higher for the 1% cold day due to the combined effects of the lower atmospheric temperature and the higher air density. One item to note is that Figure 1. Aerodynamic heating of the skin reduced the heat transfer.transferred from the aircraft skin to the air. Conclusions An analysis of the heat transfer from a heated plate has provided important insights for the possible use of the aircraft skin to reject heat from electric actuator systems.

Comparison of atmospheric properties versus altitude: (a) Temperature. Anderson.2 1 0.4 0. ρ∞ (kg/m3) 7 .1. (b) Density (DOD.6 1. 2000). 1997.6 0.8 0.4 1.8 1.2 0 0 5 10 H (km) 15 20 (b) Figure 1.340 320 300 280 1% Hot Day (a) Standard Atmosphere 1% Cold Day T∞ (K) 260 240 220 200 180 1.

2 Ma = 1.4 ΔT = [Taw-T∞] 20 Figure 1.98 Ma = 1. 8 .450 400 350 Taw (K) 300 250 200 150 100 50 0 0 5 10 H (km) 15 Ma = 0.4 20 Figure 1.8 Ma = 0.2 Ma = 1. Temperature difference (Taw − T∞ ) versus altitude for various Mach numbers (1% hot day).98 Ma = 1. Adiabatic wall temperature versus altitude for various Mach numbers (1% hot day).2.8 Ma = 0. 140 120 (K) 100 80 60 40 20 0 0 5 10 H (km) 15 Ma = 0.3.

9 .4. Temperature difference (Tw − Taw ) versus altitude for various Mach numbers (Tw = 135ºC.98 Ma = 1. Maximum Mach number before heat is transferred from the air to the skin versus altitude for various wall temperatures (1% hot day).5 0 0 5 10 H (km) 15 20 Tw = 105 ºC Tw = 115 ºC Tw = 125 ºC Tw = 135 ºC Figure 1.4 ΔT = [Taw-T∞] 20 Figure 1.5 1 0.5 2 Ma∞ . 2.2 Ma = 1. 1% hot day).5.140 120 (K) 100 80 60 40 20 0 0 5 10 H (km) 15 Ma = 0. max 1.8 Ma = 0.

1% hot day). 1% hot day).800 700 600 h (W/m2-K) 500 400 300 200 100 0 0 5 10 H (km) 15 Ma = 0. L = 1.5 0 -1 -1.8 Ma = 0. 10 .5 0 -0.0 m. Average convective heat transfer coefficient versus altitude for various Mach numbers (Tw = 135ºC.4 qw (W/cm2) 1 0.5 Ma = 0.2 Ma = 1. 3 2.4 20 Figure 1.7.98 Ma = 1. L = 1.8 Ma = 0.5 -2 H (km) 5 10 15 20 Figure 1.0 m.2 Ma = 1.98 Ma = 1.5 2 1.6. Average heat flux dissipated over the plate versus altitude for various Mach numbers (Tw = 135ºC.

2 Ma = 1.4 (a) qw (W/cm2) 1 0 -1 -2 -3 -4 3 2. Local heat flux dissipated over the plate versus plate length for various Mach numbers (Tw = 135ºC.5 (b) qw (W/cm2) 2 1.5 0 1.5 2 Figure 1. 11 .8 0.4 3 2 Ma = 0. (b) H = 10 km.8.2 1 (c) qw (W/cm2) 0.5 1 L (m) 1.6 0.2 0 0 0.8 Ma = 0. 1% hot day): (a) H = 0 km.4 0.98 Ma = 1.5 1 0. (c) H = 20 km.

5 1 0. Average heat flux dissipation versus altitude for various wall temperatures (Ma∞ = 0. L = 1.5 0 0 5 10 15 H (km) 20 Figure 1. 2. 2000). 12 .0 m.10. Anderson. 1997.98) (DOD. Ma∞ = 0.0 m. L = 1. Average heat flux dissipated over the plate versus altitude for various atmospheric conditions (Tw = 135ºC.98.9.8 7 6 1% Hot Day Standard Atmosphere 1% Cold Day qw (W/cm2) 5 4 3 2 1 0 0 5 10 H (km) 15 20 Figure 1.5 2 qw (W/cm2) Tw = 105 ºC Tw = 115 ºC Tw = 125 ºC Tw = 135 ºC 1. 1% hot day).

99725 13 .99954 Table 1.Table 1.1.0868E+0 -8.6418E-1 a1 -6.7664E-4 1.4177E-4 a2 -3.9336E-6 — R2 0.99916 0.3333E-5 1.4444E-7 -1. Regression equations for air properties versus altitude for 1% hot (DOD.99779 0.8994E-2 7.2. 2002).8507E+1 -9. 1997). Regression equations for air properties versus temperature (Incropera and DeWitt.0898E-3 -4.9921E-2 -9. y = a0 + a1H + a2H2 + a3H3 + a4H4 (H in km) Property T∞ (ºC) ρ∞ (kg/m3) a0 a1 a2 a3 a4 4.5033E+0 5. y = a0 + a1T + a2T2 + a3T3 (T in K) Property cp (J/kg-K) Pr a0 1.0187E+3 8.3483E-1 -2.2593E-9 R2 0.7778E-6 A3 4.9917E-2 2.

0 g. radial acceleration: 0 ≤ ar ≤ 10 g.2.44 m diameter centrifuge table on edge with heat applied to the evaporator via a mica heater and heat rejected using a high-temperature polyalphaolefin coolant loop. 1985). Evaporative heat transfer coefficient and thermal resistance was found not to be significantly dependent on radial acceleration. Abstract An experiment has been developed to examine operating characteristics of a titanium-water loop heat pipe (LHP) under stationary and elevated acceleration fields. For stationary operation (az = 1. Flow reversal in the LHP was found for some cases. Operating the LHP in an elevated acceleration environment revealed dry-out conditions from Qin = 100 to 400 W and varying accelerations and the ability for the LHP to reprime after an acceleration event that induced dry-out. The LHP was invented in 1972 by Gerasimov and Maidanik (Maidanik. 2. thermal resistance decreased to a minimum then increased. TITANIUM-WATER LOOP HEAT PIPE CHARACTERISTICS UNDER STATIONARY AND ELEVATED ACCELERATION FIELDS 2.1. compensation chamber. and wall superheat increased monotonically. which was likely due to vapor bubble formation in the primary wick. However. wall superheat was found to increase slightly with radial acceleration. Heat input to the compensation chamber was found to increase the evaporative heat transfer coefficient and decrease thermal resistance for Qin = 500 W. The LHP consists of an evaporator..2. the LHP evaporative heat transfer coefficient decreased monotonically. heat load at the compensation chamber: 0 ≤ Qcc ≤ 50 W. The LHP was mounted on a 2. liquid and vapor 14 . and was later patented in the United States (Maidanik et al. ar = 0 g). 2005) in the former Soviet Union. The LHP was tested under the following parametric ranges: heat load at the evaporator: 100 ≤ Qin ≤ 600 W. Introduction Loop heat pipes (LHP's) are two-phase thermal transport devices that operate passively using the latent heat of vaporization to transport heat from one location to another.

This heat transfer to the bayonet tube raises the temperature of the subcooled liquid entering the compensation chamber to the saturation temperature as it travels to the end of the evaporator. which is an inverted meniscus wick in direct contact with the exterior evaporator wall. a two-phase mixture. The vapor is captured in the axial vapor grooves in the primary wick and is directed via a manifold at the end of the evaporator to the vapor line due to the increased pressure within the evaporator. After exiting the condenser section. menisci are developed in the primary wick which establishes a capillary pressure head that returns liquid to the evaporator from the condenser. The majority of the input heat is used to vaporize the working fluid within the primary wick structure. The condensate in the compensation chamber is 15 . After sufficient heat is rejected. and. The working fluid enters the condenser as a superheated vapor. which delivers the liquid to the end of the evaporator where the vapor manifold resides. The vapor from the evaporator section travels via the vapor line to the condenser section.transport lines made of smooth tubing. it may or may not become a subcooled liquid. the vapor becomes a saturated vapor. This capillary head must be greater than the total system pressure drop in order for the LHP to continue to operate without drying out. The rest of the heat is transferred by conduction through the primary wick. most of the evaporator heat input evaporates liquid in the primary wick. Part of this vapor stream condenses onto the secondary wick. the liquid will continue to lose heat due to convection and/or thermal radiation to the ambient. which often has a circular crosssection. The location of the point at which the working fluid becomes a subcooled liquid (2φ-1φ) is dependent on the heat input at the evaporator.2).1. depending on the amount of heat rejection. The subcooled liquid returns to the evaporator via the bayonet tube. and a condenser as shown in Figure 2. This latent heat is then rejected from the compensation chamber to the ambient. where liquid is evaporated into vapor channels leading to the compensation chamber (Figure 2. The rest of the vapor condenses onto the wick lining the compensation chamber. Due to evaporation. the heat rejection at the condenser. and the saturation temperature in the compensation chamber. which is in intimate contact with the bayonet tube. Heat is applied directly to the exterior wall of the evaporator. which is also made of smooth tubing. As stated previously. Heat is rejected from the condenser to the ultimate heat sink. a saturated liquid.

there was essentially no overshoot. or a change in the operating conditions.8 g for 30 seconds followed by ar = 0. For smaller heat loads.8 g. (3) vertically with evaporator above the compensation chamber with no radial acceleration.8 g. (2000a) performed experiments on a miniature aluminum/anhydrous ammonia LHP by using a spin table to examine the effects of varying acceleration on start-up. The compensation chamber allows the LHP to automatically regulate itself during transient situations like startup.0 g for 300 seconds periodically. Controlling the heat transfer through the shell of the compensation chamber can adjust the saturation point in the condenser.0 g for 300 seconds periodically.2 and 4. (2) horizontally with the evaporator and vapor line outboard on the table. defined as the difference between the evaporator and compensation chamber wall temperatures. and combinations of ar =1. such as at Qin = 5 W. as well as varying heat load inputs up to Qin = 100 W. combination of constant ar = 1. The compensation chamber can also be used to control the location of the 2φ-1φ point in the condenser. and (4) vertically with evaporator below the compensation chamber with no radial acceleration. Their experimental results indicated that the wall superheat.drawn back to the evaporator section through the secondary wick by capillary action. for heat loads greater than Qin = 50 W.2 and 4. where the majority of the condenser section is free of subcooled liquid. shutdown. When temperature overshoot in the evaporator was examined. but 16 .0 g for 300 seconds periodically. a temperature overshoot of a few degrees was always observed. appeared to be independent of input heat load and acceleration. Several acceleration profiles were examined. including ar = 0. including LHP startup before acceleration was applied and vice versa. Several different experiments were conducted.2 g. There has been limited experimentation on the acceleration effects on loop heat pipes and heat pipes. The compensation chamber provides for storage of excess liquid when the evaporator heat input is high. Four mounting configurations were examined: (1) horizontally with the compensation chamber and liquid line outboard on the table.2 g for 30 seconds followed by ar = 0. constant ar = 4.8 g followed by ar = 0. In this way. constant ar = 1. thereby changing the amount of subcooling of the liquid returning to the evaporator. the secondary wick and the compensation chamber behave similar to a conventional heat pipe. constant ar = 4.0 g. Ku et al. constant ar = 1.

Under cyclic transverse acceleration. Transport capacity of the heat pipe dropped from Qout = 138 W at radial accelerations of ar = 1. In all of the experiments the LHP continued to operate without problems. The heat pipe was able to reprime after dry-out events with subsequent reduction of transverse acceleration. the LHP started successfully.8 g peak-to-peak and evaporator heat inputs up to Qin = 83 W. Similar research has been conducted to examine body force effects on heat pipes.0 g were investigated. Yerkes and Beam (1992) examined the same flexible copper-water arterial wick heat pipe as Ponnappan et al. The temperature difference between the evaporator and condenser remained fairly constant up to ar = 4 g then decreased from ar = 4 to 10 g. significant fluid slosh was thought to create a cyclic variation in heat pipe temperature.0 g. under transient transverse and axial acceleration forces with periodic and burst transverse accelerations from f = 0. (2000b). This in turn changed the LHP operating temperature. It was observed that pooling of excess fluid had a significant effect on the heat transport of the heat pipe at steady state transverse acceleration.01 to 0. in an extension of the previous experimental study. Temperature 17 . (1992) examined a flexible copper-water arterial wick heat pipe subjected to transverse acceleration using a centrifuge table. In every experiment.0 g to Qout = 60 W at ar = 10. This decrease was due to a more uniform distribution of fluid within the wick at the higher radial acceleration. condenser.03 Hz and magnitudes from ar = 1. Periodic acceleration led to a quasi-steady operating temperature. With sufficient time. in the sense that all the operating conditions needed to be taken into account. Their experimental results showed that the radial acceleration caused a redistribution of fluid in the evaporator.1 to 9. the temperature overshoot ranged from 0 to 45°C. Heat transport potential decreased with increasing transverse acceleration causing partial dry-out of the artery and pooling in the condenser.at Qin = 25 W. Temperature hysteresis could also be caused by the radial acceleration. Evaporator heat loads up to Qin = 150 W and steady state radial accelerations up to ar = 10. examined the temperature stability of the same miniature LHP under varying heat loads and acceleration levels. Ponnappan et al. and compensation chamber. Ku et al. The effect was not universal. constant acceleration could either increase or decrease the LHP operating temperature.

It was discovered that the thermal resistance increased and then decreased with respect to increasing acceleration frequency. At Qin = 20 W the heat pipe did not experience any dry-out conditions when the radial acceleration was increased and then decreased stepwise from ar = 0 to 10 g. condenser temperatures of Tc = 3. 20. the heat pipe experienced dry-out conditions at ar = 0 and 2 g. a step increase from ar = 0 to 10 to 0 g with a 10 second stabilization at each step. The thermal resistance of the heat pipe was noted to decrease then increase with increasing heat transported when dry-out started.15. The amplitude of the radial acceleration ranged from ar = 1. 40. Zaghdoudi and Sarno (2001) examined the body force effects on a flat copperwater heat pipe via a centrifuge setup. The heat pipe was mounted such that the accelerating forces were opposite to the liquid flow. At Qin = 50 W. 0. but quickly reprimed at the higher radial accelerations. Heat loads ranging from Qin = 20 to 250 W were applied to the evaporator.rise was lower at the onset of dry-out conditions when compared to steady state transverse acceleration. with evaporator heat loads from Qin = 75 to 150 W.1 to 9. (1998) examined a helically grooved copper-ethanol heat pipe as a function of evaporator heat input and transverse radial acceleration. 18 .8 g. Frequency of the steady periodic burst transverse acceleration had no effect on the heat pipe temperature and tended to delay the onset of dry-out. The thermal resistance also increased with increasing evaporator heat loads. and sinusoidal acceleration frequencies of f = 0. 0.01. and increasing then decreasing the acceleration from ar = 0 to 10 g after thermal stabilization. Thomas and Yerkes (1996) examined the same flexible copper-water arterial wick heat pipe as Ponnappan et al. Thomas et al. The previous dry-out history adversely affected the thermal resistance of the heat pipe when dry-out occurred prior to increasing the acceleration frequency.05. Heat loads of Qin = 20. 0. and 0. The effects of the previous dry-out history of the heat pipe were also examined. 0. and 35°C. This indicated the elevated body forces actually aided the performance of the heat pipe by increasing the capillary limit due to the forces generated from acceleration gradients down the length of the helical groove. Three types of accelerations were performed in this study: A parabolic profile from ar = 0 to 10 to 0 g with a 5 second stabilization at ar = 10 g.1. or in an “unfavorable” mounting condition.2 Hz.

and electrical signals for analog or digital control. This suggested that the heat pipe quickly reprimed after the acceleration event. electrical signals were . radial acceleration: 0 ≤ ar ≤ 10 g. The test bed consisted of a 2. A schematic of this test bed can be seen in Figure 2. and radial acceleration field. quickly returning to normal after the acceleration burst. 2. heat input at the compensation chamber. For the first two types of acceleration profile. the transient behavior during startup and steady operation has been examined. These tests demonstrated the importance of prior operation history when the heat pipe was subjected to elevated body forces. Experimental Setup The Centrifuge Table Test Bed at Wright-Patterson Air Force Base (AFRL/RZPS) was used to determine the heat transfer characteristics of the titanium-water LHP under stationary and elevated acceleration fields. The test bed was able to deliver the following to devices mounted to the rotating table: Conditioned DC electrical power through three separate power supplies. Transient temperature distributions.3. it was observed there was a delayed increase in evaporator temperature and decrease in condenser temperature. In addition. quickly returning to normal in the absence of the accelerating force.3. Thermal resistance also experienced a delayed increase in onset and remained elevated even in the absence of an accelerating force. A performance map has been developed that relates dry-out to the heat load and radial acceleration for the experimental conditions described. heat load at the compensation chamber: 0 ≤ Qcc ≤ 50 W. Thermal resistance had a similar trend. The objective of the present experiment was to determine the operating characteristics of a titanium-water loop heat pipe subjected to varying heat loads and accelerations. For the third type of acceleration profile. there was a much more gradual increase in evaporator temperature and nearly negligible decrease in condenser temperature.44 m diameter horizontal rotating table driven by a 20 hp DC electric motor. and the thermal resistance have been found in terms of the heat input at the evaporator. This was likely due to the pooling of fluid in the condenser. temperature-controlled ethylene glycol coolant. The experimental parametric ranges were as follows: heat load at the evaporator: 100 ≤ Qin ≤ 600 W. 120 VAC power. 19 In addition. the evaporative heat transfer coefficient.and 60 W were applied to examine the effect on evaporator and condenser temperature as well as thermal resistance.

collected from instruments on the table and stored in a data acquisition computer. and voltages were all measured using a data acquisition mainframe (Agilent VXI E8408A) with a command module (Agilent E1406A). accelerations. The rotational speed of the centrifuge table. heater power. Kepco ATE150-3. The volumetric flow rate of the coolant mixture was controlled using a high-pressure booster pump.4 L/min. 20 . Throughout experimentation the flow The uncertainty in this & rate was held constant at Veg = 2. 5½ digit multimeter module (Agilent E1411B). The & radial acceleration could exceed ar = 12 g. or controlled digitally using a signal generator in the data acquisition system. The temperature of the coolant was maintained at a constant setting by a recirculating chiller (Neslab HX-300). and other low voltage control devices on the table were controlled using an 8/16-channel D/A converter module (Agilent E1418A). The acceleration field was measured using an orthogonal triaxial accelerometer (Columbia SA-307HPTX) with an uncertainty of ± 0. These slip rings were separated from the instrumentation slip rings to reduce electrical noise. mass flow rates. with a maximum onset of approximately ar = 10 g/s. and HP 6290A) through power slip rings.01 g. and a 64-channel 3-wire multiplexer module (Agilent E1476A). measurement was less than 2. The current was determined from the voltage drop across a precision resistor in series with the heater. inducing a tangential acceleration.5M. Instrumentation signals generated on the table were acquired through a custombuilt forty-channel instrumentation slip ring using a data acquisition system. which aided the low-pressure pump in the recirculating chiller.0% Heat was rejected from the centrifuge table using an ethylene-glycol/water mixture that was delivered to the rotating centrifuge table via a double-pass hydraulic rotary coupling (Deublin 1690-000-115). This type of measurement was required due to the voltage drop between the control room and the table. Power was supplied to heaters on the table by three precision power supplies (Kepco ATE150-7M. Communication between the data acquisition unit and the computer was established using a general purpose interface bus (GPIB) coupled with a custom-designed LabVIEW virtual instrument. The heater power was calculated by multiplying the voltage drop across the heater by the current. The acceleration field could be varied manually using a potentiometer. Temperatures.

A summary of the requested design parameters can be seen in Table 2..1. A summary of the loop heat pipe specifications can be seen in Table 2. in Lancaster.4. The minimum transport line length was 2 m to simulate relevant aircraft geometries.5. An evaporator operating temperature of 200°C and condenser operating temperature between 5 and 140°C were selected to match relevant acquisition and rejection temperatures aboard aircraft. depending on the materials of the thermocouple wires. evaporator. A summary of their locations can be seen in Table 2. Another problem that is present when slip rings are used is electrical noise. when the thermocouple wires are connected to the wires leading to a slip ring.48 × 28.32 × 10. under contract FA8601-06-P-0076. Small deviations existed since the condenser section and the evaporator/compensation chamber were both straight. a titanium-water loop heat pipe.3. To avoid this problem. This converted the millivoltage signals from the thermocouples to 0 to 10 V signals without the creation of extra junctions.2. to match commercial off-the-shelf heaters and cold plates. Inc. Initial design parameters set by AFRL/RZPS were to develop a loop heat pipe capable of a minimum heat load of 500 W and minimum heat flux of 3 W/cm2. a Type E thermocouple amplifier was installed on the centrifuge table (Omega OM7-47-E-07-2-C) with internal cold junction compensation.16 cm and 30. This induced a non-uniform radial acceleration field over the lengths of these sections that needed to be quantified.6). and transport lines (Figure 2. PA.56 cm. After several design iterations. ACT delivered the loop heat pipe shown in Figure 2. at least one extra junction is created.Gathering temperature data from rotating machinery using slip rings presents unique problems. Stands were designed using G-7 phenolic to mount the loop heat pipe with support at the compensation chamber. condenser. The loop heat pipe was mounted onto the centrifuge table such that the centerline of the tubing coincided with the outer table radius as much as possible. The tops of these stands were anchored to the table to reduce 21 . The LHP was instrumented with twelve type E exposed tip thermocouples as seen in Figure 2. was developed for AFRL/RZPS by Advanced Cooling Technologies (ACT). The evaporator and condenser dimensions were selected to be 20. The test article. respectively. This problem was reduced (not eliminated) by the use of a low-pass filter for each of the thermocouple signals coming from the table before the data acquisition system. First.

Inc. After 22 . rather than controlling the temperature of the evaporator by controlling the compensation chamber temperature. the compensation chamber is not insulated and the temperature is closely controlled during operation. as shown in Figure 2. the LHP compensation chamber was allowed to “float” into equilibrium with the evaporator and condenser. Complete survey data can be seen in Appendix D. A flexible electric heat tape (Thermolyne) was wound around the compensation chamber and surrounded by Kaowool insulation and aluminum foil to minimize heat losses.2 cm and a maximum radius to centerline of 123. so a high-temperature fluid loop was constructed and mounted to the centrifuge table to act as an interface between the LHP and the low-temperature fluid loop. it was desired to have the option of operating the LHP condenser section at elevated temperatures.6 cm for a percent acceleration difference of 3.deflection when the table was rotating. The loop heat pipe had a minimum radius to centerline of 119. followed by the evaporator stand. The entire loop heat pipe fitted within 4. A mica heater (Minco) was located between the evaporator body and a ceramic fiber insulative layer. This would minimize parasitic heat gain. heat was applied to the LHP at the evaporator while the heat transfer to the compensation chamber was independently controlled. As a result. The high-temperature working fluid (Brayco Micronic 889 polyalphaolefin or PAO oil) flowed from the custom-made copper reservoir into a positive displacement gear pump (Tuthill). In normal operation. the centrifuge table was equipped with an on-board fluid loop for dissipating heat from sources on the table. A survey was taken at 22 locations on the loop heat pipe to determine how far various portions of the loop heat pipe were from the centerline radius. For these experiments. During operation.8%. which used ethylene glycol as its working fluid. To minimize heat loss to the environment. As previously mentioned. and reduce the use of external heaters or coolers on the compensation chamber. In the present experiment. including the compensation chamber.) for structural support and enclosed with sheet metal sides to minimize convective heat losses. insulating the LHP.7. the entire assembly was thoroughly insulated using Kaowool blankets and aluminum foil. was selected to mimic a typical configuration of a LHP in an aircraft environment where bay temperatures could be higher than the LHP temperatures. The assembly was placed inside an aluminum frame (80/20.3 cm.

For repeatability. To ensure that the bath had reached steady state at a given temperature. and another was placed prior to the flow meter (TC03). the bath temperature was first incremented from the lowest temperature to the highest temperature. When the standard deviation of 100 readings dropped below the specified threshold of 0. After the PAO exited the three heat exchangers on the condenser. Four grounded probe thermocouples for the high temperature loop and twelve exposed tip type E thermocouples mounted on the LHP were used in the experiment. and the bath temperature was changed. Type E thermocouple probes were installed at the inlet and outlet of the three heat exchangers for calorimetry (TC00 and TC01). where the error needed to be minimized. The twelve exposed tip thermocouples were mounted on the LHP in various locations and needed to be calibrated over the full range of 20 to 230°C in 5°C increments. stored in an array. The PAO then returned to the reservoir. and then decremented from highest to lowest. it flowed to a liquid/liquid heat exchanger that transferred heat from the high-temperature coolant loop to the low-temperature ethylene glycol loop. coolant flow meter calibration and the measurement of the ambient temperature. Hart Scientific 6330. The temperature readings from the sixteen thermocouples were compared to a NIST-traceable platinum resistance temperature detector (Hart Scientific RTD 1502A) with a resolution of ± 0. T = 20 to 140°C. The grounded probe thermocouples were used for calorimetry. These four thermocouples were calibrated over the anticipated range of 20 to 145°C in 5°C intervals. Thermocouple calibrations were conducted over two temperature ranges depending on the anticipated operating temperatures. The calibration procedure consisted of using two separate recirculating chiller baths (Brinkmann Lauda RCS 20-D.005°C. and the two sets of 100 data 23 .009°C.passing through a filter and a flow-straightening section. An electrical tape heater was mounted to the copper tubing after the flow meter to allow for preheating the PAO prior to reaching the calorimeter on the condenser section. 100 readings from the thermocouples were sampled. the PAO was directed through the turbine flow meter (Omega FTB-9506). T = 40 to 230°C) with PAO as the working fluid to achieve the required temperature range. the RTD temperature was continuously monitored. This was needed due to the dependence of the viscosity of PAO on temperature. which consisted of three heat exchangers plumbed in series and mounted to the condenser section.

changes significantly with temperature. A three-way valve was installed after the flow meter.95. Since the viscosity of the PAO. All of the data was collected through the instrumentation slip rings on the centrifuge table to the data acquisition system to capture all errors inherent to the centrifuge table test bed. the flow was again diverted to recirculating the PAO back to the chiller bath. Plots of the RTD temperature versus each thermocouple temperature were generated. A fifth-order polynomial was selected since it reduced the maximum deviation from the data by approximately a factor of four over a first-order trend line.points collected for each thermocouple at a given temperature were used to determine two average readings. the confidence interval of the thermocouple average reading at a confidence level of 0. the confidence interval of the RTD average reading at a confidence level of 0. the flow was diverted to a catch basin for a specified amount of time. and polynomial trend lines were fitted for each thermocouple as can be seen in Appendix C. from the high-temperature fluid loop. which allowed the entire flow system to reach a steady temperature. A lab 24 . The gear pump. used in the high-temperature fluid loop. and the maximum deviation of the temperature calculated using the polynomial curve fit from the actual measured temperature. The uncertainty associated with each thermocouple was determined by accounting for four sources of error: the stated uncertainty of the RTD. inline filter (Whitey SS56S6 140 micron) and a calibrated grounded thermocouple probe. were installed in a line from the bath to the turbine flow meter (Omega FTB-9506) and signal conditioner (Omega FLSC-61). a “calibration surface” was generated that related the output voltage of the flow meter and the temperature of the PAO at the entrance of the flow meter to the mass flow rate. and when the basin was full. The turbine flow meter used in the high-temperature fluid loop was calibrated to achieve accurate results for the amount of heat extracted from the LHP. Flow straightening sections upstream and downstream were placed according to the manufacturer's instructions. The voltage from the flow meter and the temperature from the thermocouple were recorded during this time. This was critical for the calculation of the evaporative heat transfer coefficient and the thermal resistance of the LHP.95. The calibration setup consisted of a recirculating chiller bath (Brinkmann Lauda RCS 20-D) filled with PAO from the same source as used in the high-temperature fluid loop. Once the temperature was steady.

The general root-sumsquare uncertainty equation used for all uncertainties was given by / Δ Δ Δ (2.scale (Mettler PC4400) was used to determine the mass collected during a given test run to within ± 0. x2. The minimum number of data points collected for any given run was 437.95 for each test run.002 kg/s. A 3-D paraboloid regression equation was generated using SigmaPlot to relate temperature. and mass flow rate.3) and the root-sum-square error associated with the temperature and voltage measurements given by Δ V/T 2 Δ 2 Δ /   (2. The voltages and temperatures were averaged and a confidence interval was calculated based on a confidence level of 0. the root-sum-square total error associated with the scale and stopwatch given by Δ 1 m/t / Δ Δ (2. b. These tests were completed over the range of T = 20 to 120°C in intervals of 25°C and flow & rates ranging from m = 0. c. The uncertainty of the mass flow rate measurement was affected by the maximum deviation of the regression equation from the actual data. a. The test was repeated for a total of five averaged data points for each nominal temperature and flow rate. flow meter voltage. and d are calibration constants (Appendix C).3 gm. the confidence interval for the temperature and flow meter voltage measurements.4) 25 . During each measurement. as many data points as possible were collected across the time span with the limiting factor being the iteration time on the LabVIEW software.1) where y0. …).025 kg/s in intervals of approximately 0.0064 to 0.2) where y = f(x1. and was given by cp (2.

L is the length of the evaporator.7) where Tcp is the average cold plate temperature.0077 kg/s. The average evaporative heat transfer coefficient was defined as out e v (2. was determined using the average evaporator temperature and the average temperature of the cold plate.8) 26 . (1994) and used in equation (2.The percent error on the mass flow rate decreased with increasing flow rate. h .6) where D is the inside diameter of the evaporator shell. R.5) (Appendix E). Qout. The heat transferred from the LHP condenser to the cold plate. Te is the average evaporator temperature measured by the four thermocouples embedded in the wall between the heater and the wick (Figure 2. was selected as it was the best estimate of heat actually transported by the LHP.PAO / (2.PAO Δ in p. and R are given by Δ out p.PAO Δ out cp p. the uncertainty associated with that setting was 4.0%. The thermal resistance of the loop heat pipe.5(b)). The heat rejected to the cold plate. was defined as out cp p.PAO as a function of temperature was developed by Ghajar et al. The root-sum-square uncertainty of Qout. Since the & mass flow rate was kept constant at mcp = 0. and was defined as e out cp (2.PAO out in (2.5) A linear fit equation for C p. and Tv is the external temperature of the vapor line at the outlet of the evaporator.PAO out in Δ cp cp out in Δ cp p. Qout.

Δ

1
e v

Δ

out out e out e v v

Δ
e v out / e v

Δ Δ
v

Δ

e

(2.9)

Δ

e out

cp

Δ

1
out out

Δ

1
e out

/

Δ

cp

(2.10)

The uncertainty of C p,PAO was estimated by Ghajar et al. to be 0.5% of the value. For each steady state condition, 151 data points were collected from each sensing device representing five minutes of data. Measured values were averaged and uncertainties were calculated based on the fixed error of each instrument and the confidence interval for the average at a confidence level of 0.95. A summary of the uncertainties for this experiment can be found in Table 2.4. Details of the uncertainty analysis can be found in Appendix B. 2.4. Results and Discussion The purpose of this series of experiments was to determine the operating characteristics of a titanium-water loop heat pipe subjected to changes in evaporator heat input, compensation chamber heat input, and radial acceleration. Steady state and transient temperature data were collected which provided insight into the fluid-thermal behavior of the LHP. The raw data was reduced to obtain the evaporative heat transfer coefficient, thermal resistance, and evaporator wall superheat in terms of the heat transported and radial acceleration level. Quasi-steady phenomena and dry-out of the LHP were observed and quantified in a performance map. Figure 2.8 presents a typical stationary (az = 1.0 g, ar = 0.0 g) cold-start test of the LHP, which consisted of the following: With the LHP at ambient conditions, the recirculating chiller in the low-temperature fluid loop was set to Teg = 35°C. Heat was applied as a step function to the evaporator section (in this case, Qin = 600 W) while the
& pump for the high-temperature fluid loop was simultaneously turned on ( mcp = 0.0077

27

kg/s). The mass flow rate of the high-temperature fluid loop was maintained constant at this value throughout this series of experiments to minimize the uncertainty associated with the calorimetry of the cold plate. Figure 2.8(a) shows the transient temperature response of the evaporator, vapor line, and calorimeter inlet and outlet. The temperatures appear to become steady after approximately 6000 s. However, in order to determine when steady state occurred the time rate of change of the temperatures was averaged over 15 min. intervals and plotted with respect to time as shown in Figure 2.8(b). It was observed that dT/dt approached zero shortly after 6000 s, but for times greater than 6000 s, significant oscillations occurred. The oscillations in dT/dt were not apparent in the raw temperature traces, but steady state was found to occur at approximately 18,000 s. This was further demonstrated by calculating the thermal resistance and heat transfer coefficient for this test at different times, as shown in Figure 2.8(c). This methodology was used throughout testing to ensure that a repeatable steady state was reached. Figure 2.9 also shows transient temperature traces during the Qin = 600 W test described in the previous paragraph. In Figure 2.9(a), the evaporator temperature After (TC04) increased very quickly while the rest of the LHP did not react. evaporator (TC08) suddenly increased.

approximately 60 s, the thermocouple located on the vapor line nearest to the exit of the This was followed in turn by increases in temperature reflected by the thermocouples located throughout the condenser section. This shows the progression of the saturated vapor clearing the condenser section of liquid, which was subsequently displaced into the evaporator section and the compensation chamber via the bayonet tube. Figure 2.9(b) shows that the evaporator temperature was significantly higher than the condenser temperatures, which led to a relatively high value of thermal resistance, which will be discussed in detail below. Figure 2.10 shows temperature traces in the condenser (TC09 through TC13) and at the bayonet inlet (TC14). Each figure shows the transient temperature after the stationary LHP reached steady state conditions at heat inputs ranging from 100 ≤ Qin ≤ 600 W. In Figure 2.10(a), with Qin = 100 W, the liquid entering the bayonet tube was highly subcooled at approximately 40°C. At this heat input level, the majority of the condenser was flooded with subcooled liquid. In fact, only TC09 (condenser inlet) indicated two-phase flow. Figure 2.10(b), with Qin = 200 W, was a unique case that is 28

described further in the following paragraph (Figure 2.11). Figure 2.10(c) to Figure 2.10(f) shows that the 2φ-1φ point progressed through the liquid line as heat input increased until it reached the bayonet inlet. If the heat input at the evaporator is high enough, saturated vapor will pass through the bayonet tube and reach the evaporator section. This point represents a performance limit to the LHP operation because if vapor enters the evaporator, the wick will dry out and the LHP will overheat. Figure 2.11 shows the oscillatory behavior of the LHP for the heat input of Qin = 200 W. Initially, at t = 0, the evaporator temperatures (TC04, TC05, TC06, and TC07) ranged from 66 to 68°C. The evaporator temperature nearest to the bayonet tube outlet (TC07) was the lowest, which indicated that the subcooled liquid that entered the evaporator tended to reduce the evaporator temperature at this point. The vapor line and condenser temperatures (TC08 through TC13) ranged from 46 to 58°C. The vapor line (TC08) was the highest, with the first three thermocouples in the condenser (TC09, TC10, TC11) decreasing slightly. The vapor became saturated within the condenser, and condensation formed on the interior walls of the tubing. From the point at which the quality of the working fluid was x = 1 (saturated vapor) to where it reached x = 0 (saturated liquid), the temperature should have been constant, except for the fact that the pressure dropped slightly due to viscous losses. This drop in the saturation pressure in turn decreased the saturation temperature. Past TC11, the other condenser temperatures (TC12, TC13) dropped significantly. This showed that the 2φ-1φ point, where x = 0, occurred between TC11 and TC12. The working fluid after this point became a subcooled liquid, where the temperature drop was due to sensible heat extraction by the cold plates. Interestingly, at t = 0, the temperature at the bayonet inlet (TC14) was higher than the outlet of the condenser. Under typical operation, this was not the case due to convective losses from the liquid lines. As time progressed from t = 0 (Figure 2.11(a)), several things occurred nearly simultaneously. The evaporator thermocouple nearest to the vapor manifold (TC07) suddenly decreased, which indicated movement of subcooled liquid from the exit of the bayonet tube into the evaporator. The junction between the evaporator and the compensation chamber (TC15) increased and then decreased in temperature over a relatively short period. This was due to warm liquid in the evaporator section being 29

This behavior indicated that the liquid slug had reversed direction. flow reversal was found at some operating points. This movement of liquid out of the evaporator may be due to the sudden appearance of a vapor bubble within the wick structure of the evaporator section which would tend to drive the heated liquid in the evaporator in the opposite direction. the temperatures in the evaporator and the bayonet tube inlet (TC14) started to increase. The inlet of the bayonet tube (TC14) decreased. Douglas et al.11(a).11(b). (1999) discussed flow reversal in LHPs as a phenomenon that occurred during startup and continued until the capillary pressure in the secondary wick could no longer maintain the system pressure drop.e. and then between TC12 and TC13 as shown schematically in Figure 2. followed by cooler liquid from the bayonet tube exit. in the present experiment. However. Discussion of flow reversal within LHPs in the literature was limited to startup and shutdown operation. and the two thermocouples measuring the subcooled liquid in the condenser increased (TC12 and TC13). but is actually closer to the behavior of a pulsating heat pipe. across the TC12 location. As can be seen in Figure 2. convective and radiative heat transfer from the LHP to the ambient caused the flow reversal. The dramatic increase in the condenser section (TC12) shows that the 2φ-1φ point moved from between TC11 and TC12. It was concluded that with no heat input. TC12 increased to the saturated vapor temperature existing within the first half of the condenser. flow reversal was not discussed as part of normal operation. This type of percolation is not typical of a fully operational LHP. (2004) used neutron radiography to visualize LHP operation and observed flow reversal only occurred when the heat input was reduced to Qin = 0 W. the 2φ1φ point re-crossed thermocouple location TC12 in the condenser. this was indicative of movement of the slug of liquid that existed from the 2φ-1φ point in the condenser to the meniscus within the grooves of the secondary wick inside the evaporator section. i. At approximately t = 80 s. 30 . Cimbala et al. The significant rise in the bayonet inlet temperature TC14 shows that warm liquid originally in the evaporator was now flooding back through the bayonet tube into the liquid line. as shown in Figure 2. In general.pushed through the grooves into the compensation chamber. while the condenser temperatures TC12 and TC13 decreased. In fact. the period of the oscillation was approximately 150 s. Again.2(a).

The four evaporator temperatures in Figure 2.6) increased more rapidly than Qout. the relatively large temperature drop defined by equation (2.12 shows the various steady state LHP temperatures versus transported heat for the stationary case. The behavior of the condenser temperatures with heat transported was slightly different. Figure 2. decreased monotonically with transported heat. drives the thermal resistance to a high value. as shown in Figure 2. the evaporator outlet temperature increased at a faster rate. As dry-out was approached. The evaporative heat transfer coefficient. the condenser temperatures continued to rise. a significant temperature drop was present between the inlet of the condenser (TC09) and the thermocouples within the condenser. The thermal resistance of the stationary LHP versus heat transported is presented in Figure 2. This behavior was controlled by the slope of the average evaporator temperature versus that of the evaporator outlet. However. which is indicative of an increased superheat penalty. but diverged from the vapor outlet temperature. which tended to increase the evaporator temperature.13(b). This temperature drop was a result of the fact that most of the condenser 31 . as shown in Figure 2. more of the wick in the evaporator section was depleted of liquid. where it is seen to decrease. Figure 2. the temperatures measured at TC10 and TC11 rose to match that at TC09.12(a) increased monotonically with heat transported.Figure 2.13 shows the thermal performance of the stationary LHP for heat inputs ranging from Qin = 100 to 600 W. As the heat input increased to Qin = 200 W. reach a minimum.7). which resulted in an overall decrease in h . the 2φ-1φ point traveled past the end of the condenser into the liquid lines such that all of the condenser temperatures matched the evaporator outlet temperature (TC08). At low power inputs. The temperature difference (Te − Tv ) defined in equation (2. which means that the time averaged location of the 2φ-1φ point moved farther into the condenser (between TC11 and TC12). This condition also shows that the liquid returning to the evaporator section (TC14) was highly subcooled.12(b). This shows that the 2φ-1φ point resided between TC09 and TC10.13(a).12(a). As the heat input increased. At a heat input of Qin = 300 W. (Te − Tcp ) . At the lowest heat input value (Qin = 100 W). and then increase. which means that very little of the available condenser was being used for two-phase condensation.

and wall superheat. With respect to the evaporative heat transfer coefficient. defined as the difference between the average evaporator temperature and the temperature of the evaporator/compensation chamber junction. where the vapor exiting the evaporator was slightly superheated and the 2φ-1φ point was out of the condenser.5. which resulted in the thermal resistance increasing with transported heat. For this particular test. The minimum R corresponds to the point in Figure 2. after which the compensation chamber heat input was incremented in steps of 5 W. the evaporator temperatures were relatively uniform. no notable difference was observed between starting the LHP while the unit was at ambient temperature versus a step change in the evaporator heat input from a lower to higher value or a higher to lower value.14 shows the operating characteristics and performance of the stationary LHP for an evaporator heat input of Qin = 500 W while varying the compensation chamber heat input from Qcc = 0 to 50 W. the temperature drop decreased with transported heat. Figure 2. the LHP was allowed to achieve steady state conditions for the given evaporator heat input. h increased by 68% with an increase in the overall heat input of only 3%. Past this point. As the 2φ-1φ point moved through and then exited the condenser. was found to monotonically increase with an increasing amount of transported heat. and the condenser outlet temperature dropped below the saturation temperature. the evaporator temperatures and the evaporator exit temperature both decreased while the condenser temperatures remained constant. which decreased the thermal resistance. The decrease in the average evaporator temperature significantly affected the evaporative heat transfer coefficient and the thermal resistance.section was flooded by subcooled liquid which was close to the cold plate temperature. In Figure 2.14(b). The drop in the condenser outlet temperature indicated that the 32 . This trend continued until approximately Qcc = 15 W. In fact. When a small amount of heat was input to the system through the compensation chamber (Qcc = 5 W). the evaporator section increased in temperature more rapidly than the condenser section. The wall superheat. as shown in Figure 2. for Qcc = 0 W.14(c). thermal resistance.14(a) and Figure 2.12(b) where the 2φ-1φ point just exited the condenser. the evaporator exit temperature decreased to the saturation temperature within the condenser. at which point the evaporator temperature leveled off. A summary of the stationary steady state data points and the path to reach steady state can be seen in Table 2.

Increasing the heat input to the compensation chamber moved the 2φ-1φ interface to the condenser outlet at Qcc = 15 W. In the present experiment. Figure 2.15(c) through Figure 2. this conclusion held true for Qcc ≥ 20 W.15 shows the transient temperature traces of the condenser. this subcooling is necessary to balance the additional heat input and results in underutilizing the condenser and a degradation of the thermal conductance. Also of interest is the temperature increase at the bayonet inlet (TC14) starting when Qcc = 35 W seen in Figure 2. The liquid-vapor meniscus moved backward due to the increased pressure within the compensation chamber until a point at which the pressure was balanced. bayonet tube. When the LHP operated at Qin = 500 W. the 2φ-1φ interface was located in the liquid return line. In addition. which was different than that seen at Qin = 200 W. subcooled liquid moved through the bayonet inlet as seen in typical operation.15(a) and Figure 2.15(f). with Qcc = 25 and 30 W. For Qcc ≥ 20 W. In Figure 2. the temperature of the evaporator/compensation chamber interface (TC15) did not vary appreciably. Operation at this point maximized the amount of heat transfer due to condensation with the added benefit of cooler liquid in the compensation chamber and evaporator which decreased the thermal resistance and increased the evaporative heat transfer coefficient. as the amount of subcooling increased the thermal resistance and decreased the evaporative heat transfer coefficient by way of an increased superheat penalty. This did not hold true for Qcc < 20 W. Forward flow then resumed and heat was lost through the liquid 33 . and evaporator/compensation chamber junction for Qcc = 25 to 50 W. with Qcc = 35 through 50 W.14(b). the evaporator temperatures increased and the 2φ-1φ point continued to move toward the evaporator which resulted in an increase in the thermal resistance. Ku (1999) indicated that operating the compensation chamber at a higher temperature by using an external heater in effect increases the amount of subcooling in the condenser and liquid return line.2φ-1φ point moved from the liquid line into the condenser section. According to Ku (1999). This was similar to the oscillating phenomena described for Qin = 200 W except that the liquid-vapor meniscus in the secondary wick was driven backward by the elevated vapor pressure within the compensation chamber. a sudden increase in temperature at the bayonet inlet (TC14) showed that flow reversal occurred in the evaporator section.15(b). which was due to the heat input at the shell of the compensation chamber. In Figure 2.

To estimate the amount of heat loss to the ambient in the uninsulated case. and preconditioning the temperature to Tcc = 72. the exposed compensation chamber was modeled as a cylinder in free convection with radiation. The average Nusselt number for free convection was given by (Incropera and DeWitt. It was observed that the average evaporator temperature increased.line.6.13) 34 . The average heat transfer coefficient was given by NuD The total heat loss per unit length from the compensation chamber was given by (2.12) 0.7 shows the effect of operating the stationary LHP for Qin = 500 W with the compensation chamber uninsulated.8°C via simultaneous heat input to the compensation chamber (Qcc = 20 W) and evaporator. The uninsulated compensation chamber operated at a temperature 3. shown by the slow decrease in temperature at the bayonet inlet (TC14) to the temperature of the subcooled liquid at the condenser outlet (TC13). Table 2.14) (2. thermocouple TC15 was relocated to the top side of the compensation chamber to directly monitor its operating temperature. To further explore compensation chamber heat input and heat loss to the ambient. insulated.387RaD 0.559/Pr / / / 1 (2. A summary of the steady state data points for Qcc = 0 to 50 W can be seen in Table 2. For this series of tests.6°C lower than the insulated case as expected due to free convection and radiative heat loss. 2002) NuD where RaD (2. and the thermal resistance increased in the uninsulated state when compared to the insulated case.8°C via heat input (Qcc = 100 W reduced to Qcc = 20 W) prior to heat input to the evaporator. the evaporative heat transfer coefficient decreased.60 0.11) with air properties evaluated at the average temperature of the freestream and the surface. temperature controlled to Tcc = 72.

Qin = 600 W) while simultaneously starting the pump for the high& temperature loop ( mcp = 0. For controlling the temperature of the compensation chamber. then the acceleration was reduced back to ar = 0. preconditioning the compensation chamber required approximately one hour less time to to ar = 0. ar > 0 g). which moved the 2φ-1φ point away from the condenser. ar = 10.16(b)). for this particular case. and operating temperatures were nearly identical between simultaneous compensation chamber and evaporator heat input startup and compensation chamber temperature preconditioning. The LHP was allowed to achieve steady state conditions at ar = 0. which was a nominally small value to prevent damage to the power slip rings (Figure 2. Figure 2.01 K/min. In addition. the recirculating chiller in the low temperature loop was set to Teg = 35°C.3 (Boyer et al.0 g). With the LHP at ambient conditions.14(c). reach steady state conditions over the simultaneous heat input startup. Heat was applied as a step function (in this case.1 g.0 g.2 W.16(c)) decreasing to below the threshold of 0.16(a)). When the uninsulated case was included with the previous compensation chamber heat input data. thermal resistance. the radial acceleration was increased However. As a result. indicated by dT/dt (Figure 2. demonstrating that the startup procedure had no impact on steady state conditions. The heat loss from the compensation chamber for the uninsulated case was found to be Qcc = -6.16 presents transient LHP temperatures for a typical test at elevated acceleration (az = 1.0077 kg/s). it was found that evaporative heat transfer coefficient and thermal resistance followed the trends shown in Figure 2.The emissivity of grade 2 titanium used in this calculation was ε = 0. it was advantageous to operate the LHP compensation chamber insulated for improved performance. then the acceleration was increased to the next desired radial acceleration value (in this case. The LHP was again allowed to achieve steady state conditions at the given acceleration (Figure 2. The average evaporator and cold plate temperatures were significantly different which was likely due a 10°C higher ambient temperature during the test involving the variation of compensation chamber heat input from Qcc = 0 to 50 W.1 g was reached before increasing the acceleration level. steady state at ar = 0. When the 35 ..1 g. 1994). If another elevated acceleration was desired. These results were expected since removing the insulation from the compensation chamber in effect provided cooling.1 g for a minimum of thirty min. the evaporative heat transfer coefficient.

indicated by the small oscillations in temperature at the TC13.17(a)) which influenced the distribution of fluid in the LHP.16(b).18(a). was found to decrease to a minimum. this pooling could either open or close the passage to vapor flow (Figure 2. the resultant acceleration vector magnitude and direction changed (Figure 2. the average evaporator temperature increased by 11°C. The evaporative heat transfer coefficient. pooling occurred in the bends of the condenser coil. Figure 2. These phenomena may be due in part to fluid redistribution in the LHP and is discussed in the following paragraph. Subcooled liquid entering the primary wick of the evaporator was forced to the outboard side of the evaporator body. The thermal resistance of the LHP (Figure 2. with significantly larger radii during turns.000 s in Figure 2. evaporative heat transfer coefficients. In the condenser. again due to the acceleration gradient. Due to the short radius.17(b)). again similar to the stationary test results shown in Figure 2.1 to 10. opposite of the heat source. Figure 2.17(c)).18 shows the thermal performance of the LHP for radial accelerations ranging from ar = 0.18(b)). All of these phenomena are a result of centrifuge testing. and perhaps leading to a partial dry-out of the wick (Figure 2.13(b). In fact. regardless of the radial acceleration. The elevated acceleration also hindered the ability of the secondary wick in the compensation chamber to supply the evaporator with liquid due to pooling.0 g and heat inputs ranging from Qin = 100 to 600 W. Depending on the acceleration vector direction. The 2φ-1φ point moved to the condenser outlet from the liquid line with increasing acceleration. strong acceleration gradients occur that could have advantageous or adverse effects on the LHP operation. Operation in an aircraft environment. then increase. As the rotational velocity of the centrifuge increased. The amount of subcooling increased overall as indicated by the decrease in temperature at the bayonet inlet (TC14).13(a) for the stationary LHP. will provide a more uniform acceleration gradient across the LHP and potentially yield different temperature profiles. again decreased with transported heat. it was found that the evaporative heat transfer coefficient and thermal resistance data were in close agreement with each other.0 g at t = 15.acceleration was increased to ar = 10. This indicated that bench top testing of the LHP was a reliable method for determining the evaporative heat transfer coefficient and thermal 36 . similar to the trend in Figure 2. and thermal resistances. when combining the stationary and elevated acceleration test data.

18(c)) was higher at elevated accelerations when compared to ar = 0 g. was independent of heat input and acceleration during startup. Dry-out was indicated by a steady increase in the evaporator temperature and a decrease in the heat extracted by the calorimeter Qout. (2000a and 2000b) observed that radial acceleration changed the fluid distribution throughout the LHP which changed operating temperatures and that acceleration could either increase or decrease LHP operating temperatures. The wall superheat (Figure 2. Figure 2. However. the position of the 2φ1φ point in the condenser moved toward the evaporator as indicated by a sequential decrease in the condenser temperatures.resistance of a LHP in an elevated acceleration environment.19 shows the transient response of the LHP during a series of dry-out events. causing the operating temperatures to change. In the present elevated acceleration tests. At this time. After the evaporator temperature TC06 reached Te.8.0 g at t = 300 s.19(a). A summary of all elevated acceleration steady state data points can be seen in Table 2. it will be shown that this was not true with respect to finding the dry-out limit. In fact. in all instances. the evaporator temperature continued to increase. the LHP reached steady state while rotating slowly at ar = 0. it was observed that the accelerating force changed the fluid distribution within the LHP. the radial acceleration was reduced back to ar = 0.1 g. which returned to within 4°C of the previous steady state. This larger temperature difference in TC13 was attributable to a slight change in the location of the 37 .1 g. but then leveled off and then decreased back to nearly the same temperature as the previous steady state. Ku et al.1 g and Qin = 400 W. defined as the temperature difference between the evaporator and compensation chamber. temperature overshoots were observed for mid-range heat inputs and the wall superheat. In addition. This was possibly due to fluid redistribution in the evaporator forcing liquid away from the heater. it was observed that elevated acceleration forces increased operating temperatures over those at ar = 0. This occurred because the evaporator no longer generated a sufficient flow of vapor which changed the operating point of the LHP. all of the LHP temperatures returned to within 1°C of the original steady state except for TC13 (condenser outlet). In addition. The rotational speed of the centrifuge table was increased until the radial acceleration reached ar = 8. However. In Figure 2. Significant temperature overshoots were not observed in any of the elevated acceleration tests.max = 150°C.

2φ-1φ point in the condenser.19(b).20(a) were quite close to the case shown in Figure 2. In addition. with a maximum evaporator temperature at deceleration of Te.20(a). however.max = 175°C before decreasing the radial acceleration to ar = 0. Following the previously mentioned startup procedures.1 and ar = 4. the LHP may not have recovered.0 g. the intensity of dry-out did seem to have an impact on the ability of the LHP to reprime.1 g.0 g with the same heat input (Qin = 400 W).19(b) and Figure 2.19(a) and Figure 2. the 2φ-1φ point was near TC10. showing that the heat pipe was operating during reversals in the liquid flow due to the liquid-vapor meniscus in the secondary wick moving back and forth. This recovery behavior shows that the LHP was capable of repriming at the end of an acceleration burst even if the heat input remained constant. The linear distance between these two points 38 .19(c). In each instance. Similar to the stationary case at this heat input. the evaporator temperature TC06 was allowed to reach Te. the maximum evaporator temperatures reached a peak and then monotonically decreased. However. the period of the oscillation of the ar = 0. The only significant differences in the independent variables between the two tests were the ambient temperature (ΔTamb = 5. the evaporator temperature continued to increase.9.19(c).19(c). the radial acceleration was again increased from ar = 0. This indicated that if the evaporator temperature were much higher than 200°C. and in Figure 2. whereas in Figure 2. reached a maximum.max = 200°C before decelerating.1 g. In Figure 2. Figure 2. the evaporator temperature TC06 reached Te. as shown in Figure 2. was the location of the 2φ-1φ point in the condenser: For ar = 0 g. all of the experiments presented in Figure 2. the LHP temperatures oscillated. In Figure 2. the LHP reached a quasi-steady state while the centrifuge table rotated slowly for ar = 0. the evaporator temperature reached two maximums before finally decreasing back to the previous steady state.max = 200°C.1 g case was nearly identical to the ar = 0 g case (approximately 175 s).19(b). In fact.19 were performed sequentially. and then decreased to the original steady state. In Figure 2. whereas for the case in which ar = 0. Overall. the temperatures shown in Figure 2.1 g. Of note.1 to 8.20 shows the temperature traces associated with the test at Qin = 200 W and ar = 0. this point resided close to TC12. which would have required that the heat input be reduced to zero. and the relatively small value of the radial acceleration. as presented in Table 2.11(b).3°C).

1 g. it was obvious that the location had changed significantly between the two cases. and radial acceleration on a titanium-water loop heat pipe were investigated for Qin = 100 to 600 W. which was nearly steady in the ar = 0. Qcc = 0 to 50 W. compensation chamber heat input. and wall superheat monotonically increased.0 g and Qin = 100 to 600 W.1 g. Oscillations were again seen at this acceleration level. A transient temperature rate of change method was developed to ensure steady state had been achieved. the radial acceleration was increased to ar = 4.0 g. It was believed that this small value of the radial acceleration resulted in a significant change in location of the 2φ-1φ point due to the pooling of liquid. and the LHP again reached a quasi-steady state. It was observed that dry-out conditions occurred at varying radial accelerations for Qin = 100 to 400 W. This may be due to the distance that the meniscus travelled within the evaporator. 39 . then increased over the same range.was approximately 143 cm.1 g.5. The average evaporator temperature increased by more than 30°C. Figure 2. For evaporator heat input Qin = 100 to 600 W.0 g at Qin = 500 and 600 W. which resulted in wider swings in the evaporator temperatures and significant oscillations of the cold plate outlet temperature. While it was impossible to know the exact location of the 2φ-1φ point due to the coarse resolution of the thermocouples in the condenser. Quasi-steady state conditions were observed at Qin = 200 W and ar = 4. This demonstrated that bench-top testing cannot be used to determine the dry-out limit with respect to elevated acceleration. and ar = 0. but the period of the oscillations increased to approximately 350 s. as shown in Figure 2.21 shows the steady state performance map for the LHP relating radial acceleration and heat transported for ar = 2. which was a much larger range than that for ar = 0.0 g.20(b). the evaporative heat transfer coefficient decreased and the thermal resistance increased from ar = 0 to 0.0 g. thermal resistance decreased to a minimum. Conclusions The effect of changes in evaporator heat input. it was observed that the evaporative heat transfer coefficient decreased monotonically.1 g case. 2. and the temperature in the bayonet inlet ranged from 37 ≤ Tbayonet inlet ≤ 70°C. In addition. Dry-out conditions were not observed through ar = 10. Flow reversal was observed at Qin = 200 W due to vapor bubble generation in the evaporator.0 to 10. After achieving the quasi-steady state at ar = 0.0 to 10.

2. These results differ from Ku (1999) in that an improvement was observed for compensation chamber heat input up to the point where subcooling was occurring in the condenser. Typical operation of LHPs is not along regimented schedules but in transient environments where heat sources. As such. it was found that the average evaporator temperatures dropped by 15°C and evaporative heat transfer coefficient improved by 68% with only a 3% increase in heat load. Flow reversal was observed starting at Qcc = 35 W due to the increased pressure in the compensation chamber driving the liquid/vapor meniscus backwards. The LHP was found to be able to reprime after an acceleration event that caused dry-out without the heat input being reduced to zero.When examining the effect of compensation chamber heat input for Qin = 500 W. Future Work Experimentation in this thesis has been conducted using strict regimented procedures for repeatability and to allow comparisons across data sets.6. This led to conclusion that bench top testing of the LHP is a reliable method for determining the evaporative heat transfer coefficient and thermal resistance of a LHP in an elevated acceleration environment induced by a centrifuge table. but is not sufficient for determining the wall superheat and dry-out limit. it was found that dry-out conditions occurred more readily at lower heat inputs (Qin = 100 to 400 W) than at higher heat inputs (Qin = 500 to 600 W). heat sinks. and accelerating forces are varying with respect to time. These results may or may not actually occur in an aircraft environment as centrifuge operation can induce artifacts in the data due to the short radius of operation. experimentation should be conducted using transient profiles to more closely mimic actual aircraft environments. It was also observed that radial acceleration had little effect on the evaporative heat transfer coefficient and thermal resistance of the LHP. Experimentation should also be conducted with tighter 40 . Operating the LHP compensation chamber uninsulated at Qin = 500 W was found to degrade the LHP performance for this particular case and preconditioning the compensation chamber temperature prior to evaporator heat input shortened the time to steady state. When examining the effect of radial acceleration. Wall superheat was found to be higher at steady state elevated accelerations when compared to ar = 0 g.

41 .control over the compensation chamber temperature to enhance repeatability and tailor operation to specific heat sources and sinks.

Adapted and reprinted with permission from AIAA (Hoang and Ku. 42 .1. 2003). Loop heat pipe operation.Qcc Qin Liquid Q2 Q1 Evaporator Compensation Chamber Bayonet Tube Primary Wick Heat Exchanger Liquid Line Vapor Line 1φ Qout Liquid-Phase (Subcooler) 2φ Qout Vapor-Phase (Condenser) Figure 2.

43 .2. 2003). Adapted and reprinted with permission from AIAA (Hoang and Ku.Meniscus Liquid from Compensation Chamber Vapor to Compensation Chamber Liquid from Liquid Line Non-wick Flow Path Q1 Primary Wick Bayonet Tube Bayonet Tube Secondary Wick Primary Wick Q1 Q2 Liquid Flow (a) Secondary Wick Q2 Non-wick Flow Path (b) Figure 2. Evaporator schematic: (a) Side view. (b) Cross-sectional view.

Schematic of Centrifuge Table Test Bed. 44 .Triaxial Accelerometer TV Camera Hydraulic Rotary Test Article Coupling Instrumentation Slip Rings Counterbalance Weights TV Monitor Thermocouple Signal Conditioner Cool Bath Centrifuge Table Gear Box Power Slip Rings 20 HP DC Motor 0 .1 kW DC Power Supplies Data Acquisition & Control Isolation Transformer Motor Controller 120 VAC Computer w/ LabVIEW Control Room Centrifuge Table Room Figure 2.3.

Figure 2.4. Titanium-water loop heat pipe test article as delivered. 45 .

(b) Locations of TC04 through TC07 within the evaporator.5. 46 .14 15 4 5 6 7 8 9 10 11 12 13 (a) 4 5 6 7 Qin (b) Figure 2. Thermocouple locations on the LHP: (a) Locations of thermocouples TC04 through TC15 across the LHP.

(a) (b) (c) (d) Figure 2. Mounting of loop heat pipe to centrifuge table. 47 . (d) Complete loop heat pipe.6. (c) Condenser with cold plate. front and top views: (a) Evaporator and compensation chamber: (b) Transport lines.

High temperature fluid loop: (a) Schematic.7. TC00. flowmeter. pump. and TC01. 48 . (c) Cold plate. TC03.LHP Condenser Cold Plate TC00 TC01 LHP Evaporator Electric Heater PAO Heater Flow meter TC03 Filter Gear Pump Reservoir Liquid-Liquid Heat Exchanger Recirculating chiller (a) Reservoir L/L HX Filter Pump TC03 (b) Flow meter TC00 Cold plate TC01 (c) Figure 2. (b) Reservoir. and liquid/liquid heat exchanger. filter.

160 140 120 (TC08) (TC09) (TC04) (a) 100 T (C) 80 60 40 20 0 3.Cond In (b) TC00 .Evap 2 TC06 .3 0. Qcc = 0 W. 49 .Evap 1 TC05 .5 1 0. (b) Transient rate of change of temperatures.Evap 1 TC08 .1 R (K/W) h (W/m2-K) 600 500 400 300 200 100 0 0 5000 h R 0. Tamb = 38.CP In TC01 .CP Out TC04 .05 0 10000 15000 20000 25000 30000 t (s) Figure 2.1°C): (a) Transient temperature traces.5 900 800 700 (c) 0.5 0 -0. Tcp = 67.Evap Out TC09 .Evap 4 TC08 .CP Out TC04 .2 0. (c) Transient thermal resistance and evaporative heat transfer coefficient.5 (TC00) (TC01) TC00 . mcp = 0.25 0.15 0.7°C. Use of a cold-start test to determine when steady state occurred for the & stationary LHP (Qin = 600 W.Evap 3 TC07 .8.CP In TC01 .Cond In dT/dt (K/min) 2 1. ar = 0 g.5 3 2.Evap Out TC09 .0077 kg/s.

1°C): (a) Initial startup.Evap 1 TC08 . Tcp = 67.0077 kg/s.9.Cond 2 TC12 .Evap Out TC09 .120 (a) 100 80 T (C) (TC04) (TC13) (TC14) (TC11) (TC12) (TC15) 60 40 20 0 0 160 140 120 100 200 400 600 t (s) 800 (TC08) (TC09) (TC10) TC04 .Cond In TC10 . (b) Complete startup until steady state.Cond Out TC14 .Cond 1 TC11 . & mcp = 0. ar = 0 g. Transient startup of the stationary LHP (Qin = 600 W.Evap/CC 1000 1200 (b) T (C) 80 60 40 20 0 0 5000 10000 15000 t (s) 20000 25000 30000 (TC04) (TC08) (TC09-TC15) Figure 2.7°C.Bayonet In TC15 . 50 .Cond 3 TC13 . Qcc = 0 W. Tamb = 38.

TC13) (TC14) (d) T (C) (TC09. Transient temperature profiles in the condenser and bayonet tube of the & stationary LHP (Qcc = 0 W. 51 .TC13) (TC14) (f) T(C) (TC09. 36.8 ≤ Tcp ≤ 71. (c) Qin = 300 W.Cond 3 TC14 .1°C): (a) Qin = 100 W. (b) Qin = 200 W.65 60 55 50 45 40 35 65 60 55 50 45 40 35 75 70 65 60 55 50 45 85 80 75 70 65 60 55 100 95 90 85 80 75 70 110 105 100 95 90 85 80 (TC09) T (C) (TC10) TC09 .Cond 1 TC12 .Bayonet In (a) (TC09. (f) Qin = 600 W. (e) Qin = 500 W.6°C.TC13) (TC14) (e) T (C) (TC09.Cond In TC11 .7 ≤ Tamb ≤ 38.TC11) (b) T (C) (TC12) (TC13) (TC14) (c) T (C) (TC09.TC13) (TC14) Figure 2. mcp = 0. ar = 0 g.Cond 2 TC13 . 31.0077 kg/s.10.Cond 4 (TC11) (TC12-TC14) TC10 . (d) Qin = 400 W.

Tcp = 46.7°C): (a) Transient temperature profiles.0077 kg/s. (b) 2φ-1φ point oscillation in the condenser.70 65 60 T (C) (TC04) (TC05) (TC06) (TC07) (a) Evaporator Vapor Line Evap / CC Condenser Liquid Line Cold Plate (TC08) (TC09) (TC10) (TC11) 55 50 45 (TC12) (TC13) (TC14) (TC15) (TC00) (TC01) 40 0 50 100 150 t (s) t = 80 200 250 300 0 ≤ t ≤ 30 13 1φ 12 Condenser 30 ≤ t ≤ 80 80 ≤ t ≤ 160 t = 160 (b) 11 2φ 10 9 Figure 2. mcp = 0. Transient temperature profiles of the stationary LHP for Qin = 200 W (Qcc = & 0 W. ar = 0 g. 52 .1°C.11. Tamb = 31.

12. 36.0077 kg/s. (b) Condenser section.6°C.7 ≤ Tamb ≤ 38. mcp = 0. 31. 53 .160 140 120 100 T (C) (a) 80 60 40 20 0 120 100 80 TC04 Evap 1 TC05 Evap 2 TC06 Evap 3 TC07 Evap 4 TC08 Evap Out TC15 Evap/CC (b) T (C) 60 40 20 0 0 100 200 300 Qout (W) 400 TC08 Evap Out TC09 Cond In TC10 Cond 1 TC11 Cond 2 TC12 Cond 3 TC13 Cond Out TC14 Bayonet In 500 600 Figure 2. Steady state temperature distribution versus transported heat for the & stationary LHP (Qcc = 0 W.1°C): (a) Evaporator section.8 ≤ Tcp ≤ 71. ar = 0 g.

(c) Wall superheat 54 .2 0.8 ≤ Tcp ≤ 67.05 0 60 50 40 (c) ΔTsh (K) 30 20 10 0 0 100 200 300 Qout (W) 400 500 600 Figure 2.13.3 (b) R (K/W) 0.25 0. mcp = 0.2500 2000 h (W/m2-K) (a) 1500 1000 500 0 0.7°C.7°C): (a) Evaporative heat transfer coefficient.0077 kg/s.1 0.4 0.6 ≤ Tamb ≤ 38. 27.45 0.35 0.15 0. Steady state performance characteristics of the stationary LHP versus & transported heat (Qcc = 0 W. (b) Thermal resistance. ar = 0 g. 36.

Bayonet In (b) T (C) (c) 0.Evap Out TC15 .Evap Out TC09 .Cond Out TC14 .12 R (K/W) 1400 1200 1000 800 600 400 200 0 -10 0 10 20 30 Qcc (W) h h .Uninsulated R R . (b) Condenser temperatures.Uninsulated 0. 36.140 130 120 T (C) TC04 .Evap 4 TC08 .14.1 0.Evap 2 TC06 .04 0.06 0.Cond 1 TC11 . Steady state performance characteristics of the stationary LHP versus & compensation chamber heat input (Qin = 500 W.Evap/CC (a) 110 100 90 80 110 100 90 80 70 60 2000 1800 1600 h (W/m2-K) TC08 .Cond 2 TC12 .14 0. 55 . (c) Evaporative heat transfer coefficient and thermal resistance.02 0 60 40 50 Figure 2.Cond In TC10 .Cond 3 TC13 .1°C): (a) Evaporator temperatures.1 ≤ Tamb ≤ 38.16 0. 63.0077 kg/s.Evap 3 TC07 . ar = 0 g.Evap 1 TC05 . mcp = 0.4 ≤ Tcp ≤ 64.08 0.8°C.

TC12) (TC13) (TC14) (TC15) T (C) (TC09.TC09 .4 ≤ Tcp ≤ 64.15. mcp = 0.Bayonet In 110 100 90 80 70 60 110 100 90 80 70 60 110 100 90 80 70 60 110 100 90 80 70 60 110 100 90 80 70 60 110 100 90 80 70 60 0 (a) (TC09.Cond 2 TC13 . (b) Qcc = 30 W. 36.TC11) (TC12) (TC13) (TC14) (TC15) (c) T (C) (TC09. (f) Qcc = 50 W.1°C): (a) Qcc = 25 W. ar = 0 g. 56 .1 ≤ Tamb ≤ 38.TC11) (TC12) (TC13) (TC14) (TC15) (d) T (C) (TC09.TC10) (TC11) (TC12) (TC13) (TC14) (TC15) (f) T (C) 50 100 150 t (s) 200 250 300 Figure 2. Transient temperature profiles in the condenser and bayonet tube of the & stationary LHP for Qcc = 25 to 50 W (Qin = 500 W. (e) Qcc = 45 W.TC12) (TC13) (TC14) (TC15) (b) T (C) (TC09. (c) Qcc = 35 W.Cond 1 TC12 . 63.Cond Out TC10 .8°C. (d) Qcc = 40 W.TC10) (TC11) (TC12) (TC13) (TC14) (TC15) (e) T (C) (TC09.0077 kg/s.Cond 3 TC14 .Cond In TC11 .

6 0.0 g.8 0. Qcc = 0 W. (b) Transition to and steady state at ar = 10.Cond 2 TC12 .Evap 1 TC05 .2 ≤ Tcp ≤ 59.Evap Out TC09 . mcp = 0. (c) Transient rate of change of temperatures.CP Out TC04 . 57 .Cond In 18000 24000 t (s) Figure 2.1°C): (a) ar = 0.2 0 6000 12000 TC00 .4 1.2 0 -0.6 1.7°C.Evap 1 TC08 .1 g startup phase.Cond Out TC14 .0077 kg/s.Cond In TC10 .Evap 2 TC06 .4 0. 27.Evap/CC 3000 6000 t (s) 9000 12000 15000 (b) (TC04) (TC13) (TC09-TC12) (TC15) (TC14) T (C) 100 80 60 40 20 15000 1.Cond 3 TC13 .CP In TC01 .Evap 4 TC08 .Bayonet In TC15 .Evap 3 TC07 .2 dT/dt (K/min) 18000 21000 t (s) 24000 27000 (c) 1 0.Cond 1 TC11 .Evap Out TC09 .16. Transient temperature traces of the LHP at elevated acceleration (Qin = 600 & W. 55.140 120 T (C) (a) 100 80 60 40 20 0 140 120 (TC04) (TC08) (TC09-TC13) (TC15) (TC14) TC04 .9 ≤ Tamb ≤ 30.

Effect of resultant acceleration vector direction on fluid distribution within the LHP: (a) Resultant acceleration vector orientation versus radial acceleration. (c) Liquid pooling in the condenser bends. (b) Liquid pooling in the evaporator. and condenser under elevated acceleration (to scale. compensation chamber. top view).17.90 80 70 60 50 40 30 20 10 0 0 2 4 ar (g) 6 8 10 az θ ar |a| θ (arc degrees) (a) (b) Liquid Pooling V L (c) Figure 2. 58 .

25.2000 1800 1600 h (W/m2-K) (a) 1400 1200 1000 800 600 400 200 0 0. 37.0 g a r = 8.3 0.0077 kg/s.18.1 0.2 0. (b) Thermal resistance.1 ≤ Tamb ≤ 38.0 g a r = 10.0 g a r = 6. 59 .2 ≤ Tcp ≤ 67.7°C.0 g a r = 4.0 g (c) ΔTsh (K) 50 40 30 20 10 0 0 100 200 300 Qout (W) 400 500 600 Figure 2.4 0. Steady state performance characteristics of the LHP versus transported heat & at stationary and elevated acceleration (Qcc = 0 W. (c) Wall superheat.5 0. mcp = 0.25 0.1 g a r = 2.15 0.35 (b) R (K/W) 0.05 0 90 80 70 60 ar= 0 g a r = 0.45 0.7°C): (a) Evaporative heat transfer coefficient.

Cond 1 TC11 .7°C.2 ≤ Tcp ≤ 59.Cond In TC10 . (b) Te.1 g 8g (Qout) (TC06) TC06 . Qcc = 0 W. mcp = 0.225 205 185 165 T (C) 8g 0.1 g 8g 0. Tamb = 28.Evap 3 TC09 .Cond 2 TC12 .1 g (c) 500 450 400 350 300 250 200 150 100 50 145 125 105 85 65 45 25 13000 (Qout) (TC06) (TC09) (TC10) (TC11) (TC12) (TC13) 15000 17000 t (s) 19000 0 21000 Figure 2.max = 200°C.0077 kg/s.1 g (b) 500 450 400 (Qout) (TC06) 350 300 250 200 150 100 50 T (C) 0 25 6000 7000 8000 9000 10000 11000 12000 13000 t (s) 225 205 185 165 T (C) (TC09) (TC10) (TC11) (TC12) (TC13) 0. (c) Te.1 g (a) 500 450 400 Qout (W) Qout (W) Qout (W) 145 125 105 85 65 45 25 0 225 205 185 165 145 125 105 85 65 45 0.Cond 3 TC13 .max = 175°C.Cond Out Qout 350 300 250 200 150 100 50 0 (TC09) (TC10) (TC11) (TC12) (TC13) 1000 2000 3000 4000 5000 6000 t (s) 0.0°C): (a) Te.max = 150°C. 60 . Transient temperature traces of the LHP at elevated acceleration showing & dry-out behavior (Qin = 400 W. 37.19.

mcp = 0. Quasi-steady state temperature traces of the LHP and cold plate at elevated & acceleration for Qin = 200 W (Qcc = 0 W. Tamb = 26.75 70 65 60 T (C) (TC05) (TC06) (TC07) (TC04) (TC08) (TC09) (TC15) Evaporator Vapor Line Evap / CC (a) Condenser Liquid Line Cold Plate 55 50 45 40 35 0 110 100 90 80 (TC04) (TC05) (TC06) (TC07) (TC08) (TC15) (TC09) Evaporator Vapor Line Evap / CC Condenser Liquid Line Cold Plate (TC10) (TC11) (TC12) (TC13) (TC14) (TC00) (TC01) 50 100 150 t (s) 200 250 300 (b) T (C) 70 60 50 40 30 0 (TC10) (TC11) (TC12) (TC13) (TC14) (TC00) (TC01) 100 200 300 400 t (s) 500 600 700 Figure 2.1 g and t = 13834 s. (b) Transient temperature trace at ar = 4.0077 kg/s.9°C.4°C): (a) Transient temperature trace at ar = 0. 61 .0 g and t = 31240 s. Tcp = 41.20.

62 . 25. 37.2°C).1 ≤ Tamb ≤ 30.2 ≤ Tcp ≤ 59.0077 kg/s.12 10 8 ar (g) Steady Operation Quasi-steady Operation Dry-out 6 4 2 0 0 100 200 300 Qout (W) 400 500 600 Figure 2. Steady state performance map of the LHP relating radial acceleration and & heat transported (Qcc = 0 W. mcp = 0.7°C.21.

Table 2.48 × 28. CP Grade 2 Transport Line Material Titanium.1.16 cm Condenser Footprint 30.32 × 10. 243.58 cm Transport Line Lengths Approx. CP Grade 2 Evaporator Wick Material Titanium.54 cm OD up to 25.4 cm long Evaporator Footprint 20. Requirement Parameter Thermal Minimum Heat Load 500 W Minimum Heat Flux 3 W/cm2 Operating Temperature 200°C Condenser Heat Sink Temperature 5 to 140°C Tilt in One G ± 0 inches. horizontal Conductance 50°C/W Proof of Pressure Test 3102 psi (200°C) Materials Evaporator Envelope Material Titanium.8 cm 63 . AFRL/RZPS design requirements. CP Grade 2 Working Fluid Water LHP Dimensions Evaporator Configuration 2.

Table 2.033 cm OD Length 11.6350 OD × 0.8001 cm Grooves 6 Groove Depth 0.4 cm Condenser Line Diameter 0.1524 cm Groove Width 0.9525 OD × 0.43 cm Chamber Location Coaxial with evaporator Wick Properties Effective Pore Radius 9.286 cm Length 20.3 cm Liquid Line Diameter 0.2. Transport Lines Vapor Line Length Approx. 335. 243.2×10-12 m2 Outside Diameter 2.0889 cm wall Condenser Line Length Approx.0889 cm wall Compensation Chamber Diameter 6.1μm Permeability 1.8 cm Vapor Line Diameter 0.0889 cm wall Liquid Line Length Approx. 279.32 cm Inside Diameter 0.9525 OD × 0.1524 cm 64 . ACT LHP geometric design parameters.

Summary of LHP thermocouple locations Thermocouple Location TC04 Evaporator 1 TC05 Evaporator 2 TC06 Evaporator 3 TC07 Evaporator 4 TC08 Evaporator Outlet TC09 Condenser Inlet TC10 Condenser 1 TC11 Condenser 2 TC12 Condenser 3 TC13 Condenser Outlet TC14 Bayonet Inlet TC15 Compensation Chamber / Evaporator 65 .3.Table 2.

01 + confidence interval + (ar. ΔTin) Tcp Te ΔTsh Cp.cl – ar.oe – ar.01 + confidence interval + (ar.4. Summary of uncertainties.Table 2.0000254 m Calculated ± max(ΔTout.cl)) g .PAO ± max(Δ(TC04). Δ(TC06).10) Qout h R 66 . Δ(TC05).(0.28°C ± 0.9) Equation (2. L ± 0. Δ(TC07)) ± 1. Quantity Uncertainty Measured Temperature ± (fixed value + confidence interval) °C Acceleration + (0.00025V + 0.005 + confidence interval) V Constant Resistors ± 0.8) Equation (2.5% * C p (Tcp ) + (C p (Tcp + ΔTcp ) − C p (Tcp )) Equation (2.02%R Ω Wick D.ie)) g Voltages ± (0.

5 174 ± 7.3 78.361 ± 0.12 43.7°C) showing effect of startup path.6 ± 0.12 38.86 166 0-100-133-166 65.9 ± 4.8 ± 4.0 ± 0.12 37.116 ± 0.8 ± 0.351 ± 0.86 0-100 64.34 59.34 65.355 ± 0.6 ± 0.020 0. Steady state operating characteristics for the stationary LHP (Qcc = 0 W.86 0-200 66.2 ± 0.1 ± 4. 27.3 77.1 ± 0.34 59.8 ± 0.9 ± 0.12 36.12 37.3 ± 0.119 ± 0.86 0-100 64.0 ± 4. Path to Final Final Qin Te (°C) Te/cc (°C) (W) Qin 0-100 66.0 ± 0.4 ± 0.12 39.5.2 ± 0.9 ± 0.361 ± 0.12 46.6 ± 0.8 ± 0.34 61.4 ± 0.86 0-100-300-100 66.12 40.019 0.4 ± 0.8 ± 0.12 45.12 45. mcp = 0.1 ± 0.7 ± 0.4 ± 0.217 ± 0.1 ± 4.158 ± 0.3 ± 0.020 0.5 ± 0.6°C.020 0.5 ± 0.86 0-200 65.0055 1670 ± 390 1660 ± 380 1670 ± 390 1540 ± 280 1540 ± 300 1730 ± 560 1620 ± 500 1770 ± 310 1760 ± 250 1560 ± 230 1460 ± 140 1660 ± 250 67 .0077 kg/s.12 38.34 60.3 78.3 79.35 66.9 ± 4.86 Tcp (°C) 39.8 ≤ Tcp ≤ 71.0057 0.011 0.34 60.020 0.4 ± 0.7 174 ± 7.0076 0.020 0.34 62.6 ± 0.1 ± 0.86 100 0-100 65.8 h (W/m2-K) R (K/W) 0.020 0.6 ≤ Tamb ≤ 38.0058 0.& Table 2. 36.5 143 ± 6.34 59.86 0-100 65.34 62.2 ± 0.9 ± 0.350 ± 0.4 ± 0.4 ± 0.342 ± 0.34 58.7 176 ± 7.86 0-100-500-100 67.0 ± 0.3 114 ± 5.86 133 0-100-133 65.5 ± 0.348 ± 0.3 79.3 ± 0.8 ± 0. ar = 0 g.46 66.4 ± 4.12 Qout (W) 79.86 200 0-200 66.2 77.114 ± 0.4 ± 0.

0045 0.0985 ± 0.40 60.127 ± 0.1 ± 0.0077 kg/s.86 400 0-300-400 94.6 ± 0.4 ± 0.86 300 0-200-300 78.12 50. mcp = 0.3 ± 0.157 ± 0.12 67.34 93.8 ± 0.5 ± 0.1 ± 0. Steady state operating characteristics for the stationary LHP (Qcc = 0 W.7 ± 0.40 64.4 ± 0.86 0-600 148 ± 0.6 ≤ Tamb ≤ 38. Path to Final Final Qin Te (°C) Te/cc (°C) (W) Qin 0-100-300 77. 36.135 ± 0.35 75.0 ± 0.12 62.7 ± 0.36 93.0066 1660 ± 160 1670 ± 170 1580 ± 140 1550 ± 160 1400 ± 85 1470 ± 98 1190 ± 60 1050 ± 45 999 ± 41 1050 ± 47 956 ± 42 803 ± 33 867 ± 37 757 ± 33 68 .0053 0.34 83.34 85.105 ± 0.12 67.9 ± 0.0045 0. continued.0045 0.0 ± 0.12 61.8 ≤ Tcp ≤ 71.86 0-400 97.6 ± 0.0053 0.86 0-200-400-600 145 ± 0.1 ± 0.86 0-100-500 119 ± 0.7°C) showing effect of startup path.12 52.12 52.8 ± 0.0060 0.103 ± 0.39 62.5 ± 0.1 ± 0.0045 0.34 82.38 63.8 ± 0.3 ± 0.7 ± 0.36 93.5. ar = 0 g.0057 0.6°C.3 ± 0.8 ± 0.86 Tcp (°C) 51.0 ± 0.126 ± 0.1 ± 0.6 ± 0.0051 0.12 58.34 73.8 ± 0.86 0-200-400 94.& Table 2.34 83.12 57.149 ± 0.4 ± 0. 27.119 ± 0.104 ± 0.131 ± 0.12 Qout (W) 260 ± 11 259 ± 11 258 ± 11 258 ± 11 345 ± 14 345 ± 14 344 ± 14 430 ± 18 429 ± 18 431 ± 18 432 ± 18 515 ± 21 515 ± 21 513 ± 21 h (W/m2-K) R (K/W) 0.4 ± 0.4 ± 0.86 500 0-400-500 117 ± 0.0062 0.86 0-300 79.6 ± 0.86 0-100-300 75.142 ± 0.86 0-500 122 ± 0.0047 0.12 67.12 63.86 600 0-400-500-600 141 ± 0.0055 0.7 ± 0.86 0-100-300-500 116 ± 0.34 72.12 63.4 ± 0.0988 ± 0.12 56.2 ± 0.0046 0.3 ± 0.101 ± 0.

86 85.12 64.0049 0.1°C).86 97.0048 0.0970 ± 0.7 ± 0.103 ± 0. 36.0042 0.8 ± 0.103 ± 0.86 95.3 ± 0.0 ± 0.12 Qout (W) 432 ± 18 438 ± 18 444 ± 18 449 ± 19 456 ± 19 461 ± 19 466 ± 19 468 ± 19 473 ± 20 475 ± 20 477 ± 20 h (W/m2-K) R (K/W) 0.3 ± 0.8 ± 0.0044 0.35 120 ± 0.6 ± 0.34 127 ± 0.8°C.0056 122 ± 0.118 ± 0.Table 2.86 Tcp (°C) 63.12 64.86 104 ± 0.0991 ± 0.114 ± 0.0972 ± 0.0077 kg/s. ar = 0 g.0055 0.34 110 ± 0.135 ± 0.34 129 ± 0.2 ± 0.12 64.86 87.134 ± 0.34 112 ± 0.1 ± 0.4 ± 0.12 63.86 105 ± 0.1 ± 0.12 64.7 ± 0.86 90.86 86.0044 0.12 63.5 ± 0. mcp = 0.117 ± 0.0041 0.86 85.35 107 ± 0.34 115 ± 0.6 ± 0.12 63.0 ± 0.0050 0.86 85. Qcc (W) 0 5 10 15 20 25 30 35 40 45 50 Te (°C) Te/cc (°C) 85. 63.35 108 ± 0.35 956 ± 42 1130 ± 58 1420 ± 110 1610 ± 130 1610 ± 130 1620 ± 130 1610 ± 130 1580 ± 120 1580 ± 120 1530 ± 110 1510 ± 110 69 .6 ± 0.12 64.34 118 ± 0.6.1 ≤ Tamb ≤ 38.4 ≤ Tcp ≤ 64.35 109 ± 0.12 64.5 ± 0.132 ± 0. Steady state operating characteristics for the stationary LHP showing effect of heat input to the compensation chamber (Qin & = 500 W.4 ± 0.5 ± 0.5 ± 0.0041 0.0057 0.12 64.

0077 kg/s. preconditioned CC Qcc (W) 0 0 20 100 decreased to 20 Te (°C) Tcp (°C) 52.142 ± 0.3 62. 115 107 103 878 ± 37 1050 ± 51 1340 ± 77 103 53. Tcp = 52.106 ± 0.8°C. The effect of compensation chamber temperature control on LHP operation (Qin = 500 W.7.0052 0.4 72. Tamb = 26.9 72. 70 . mcp = 0. temperature controlled to Tcc = 72.& Table 2. no temperature control Insulated.0045 250 min.0060 0.5°C. ar = 0 g.4°C) Compensation Chamber Conditions Uninsulated.6 53.106 ± 0. simultaneous heat input startup Insulated.8°C. temperature controlled to Tcc = 72.0045 Time to Steady State 300 min.4 52.8 467 ± 19 1350 ± 77 0. no temperature control Insulated.122 ± 0.5 Tcc (°C) 59.8 Qout (W) 442 ± 19 447 ± 19 470 ± 20 h (W/m -K) 2 R (K/W) 0. 375 min. 310 min.

2 275 ± 12 285 ± 13 360 ± 15 361 ± 15 361 ± 15 361 ± 15 372 ± 16 376 ± 16 376 ± 17 369 ± 16 h (W/m2-K) R (K/W) 0.21 46.050 4.2 ± 0.012 2.12 49.7 ± 0.8 ± 0. 25.1 ≤ Tamb ≤ 30.091 6.0055 0.63 94.86 61.86 59.13 Te (°C) Te/cc (°C) 59.0 ± 0.2 ≤ Tcp ≤ 59.8 172 ± 7.0044 0.3 ± 0.0 ± 0.012 0.0063 0.12 49.86 60.0 ± 0.1 ± 0.9 ± 0.1 ± 0.9 ± 0.39 86.169 ± 0.43 71.0081 0. Steady state operating characteristics of the rotating LHP (Qcc = 0 W.86 Tcp (°C) 38.2 ± 0.12 49.0 ± 0.1 ± 0.0096 0.0051 0.112 ± 0.12 37.16 42.2 ± 0.8 ± 0.446 ± 0.13 41.0 ± 0.0 ± 0.0 ± 0.39 85.029 0.6 ± 0.7°C.126 ± 0.34 64.42 87.0051 0.7 ± 0.& Table 2. Qin (W) 100 100 100 100 200 200 200 200 300 300 400 400 400 400 400 400 400 400 ar (g) 0.103 ± 0.4 ± 0.0045 0.48 81.8 174 ± 7.5 ± 0.5 76.16 49.6 ± 0.86 61.8 188 ± 9.2 ± 0.1 ± 0.34 66.65 76.6 ± 0.2°C).2 ± 0.050 0.1 ± 0.42 91.2 79.101 ± 0.8 ± 0.5 ± 0.5 ± 0.6 ± 0.0053 64.020 0.1 ± 0.86 59.9 ± 0.41 71.091 0. 37.181 ± 0.3 ± 0.122 ± 0.12 Qout (W) 74.6 ± 0.4 ± 0.6 ± 4.15 49.089 4.1 ± 4.5 ± 0. mcp = 0.1 ± 0.86 61.021 0.86 60.012 2.3 ± 0.012 0.0 ± 0.34 1620 ± 340 1810 ± 510 1770 ± 420 1130 ± 150 1400 ± 110 1380 ± 110 1360 ± 110 1220 ± 86 1390 ± 100 1280 ± 80 1300 ± 82 1300 ± 81 1330 ± 84 1330 ± 89 1280 ± 74 1160 ± 56 1130 ± 59 1030 ± 46 71 .0 ± 0.0079 0.8.116 ± 0.8 ± 0.103 ± 0.050 0.86 64.3 ± 0.0 ± 0.3 ± 0.86 60.0077 kg/s.012 2.43 93.012 2.3 ± 0.86 61.8 ± 0.361 ± 0.38 75.1 ± 0.13 42.102 ± 0.1 ± 0.2 ± 4.1 ± 0.34 71.3 ± 0.7 ± 0.8 ± 0.12 38.9 ± 4.86 57.103 ± 0.355 ± 0.1 ± 0.0044 0.12 45.8 ± 0.0 ± 0.4 ± 0.12 49.86 59.76 86.2 74.1 ± 0.0045 0.86 59.050 4.1 ± 0.44 70.8 174 ± 7.0 ± 0.111 ± 0.7 ± 0.86 61.1 ± 0.0045 0.170 ± 0.39 86.3 ± 0.012 0.362 ± 0.012 0.12 37.1 ± 0.012 0.86 61.012 0.86 60.021 0.6 ± 0.165 ± 0.3 ± 0.0078 0.1 ± 0.1 ± 0.86 59.12 49.12 49.3 ± 0.12 41.3 ± 0.

86 72.3 ± 0.34 137 ± 0.0 ± 0.2 ± 0.35 107 ± 0.43 141 ± 0.12 57.012 0.0077 kg/s.34 960 ± 43 1030 ± 48 1030 ± 50 1000 ± 46 1010 ± 47 887 ± 37 982 ± 43 934 ± 42 865 ± 38 824 ± 35 759 ± 32 707 ± 29 703 ± 29 707 ± 30 821 ± 35 726 ± 31 621 ± 26 651 ± 27 662 ± 28 691 ± 29 72 .1 ± 0.4 ± 0.86 64.1 ± 0.9 ± 0.1 ± 0.2 ± 0.012 0.& Table 2.12 56.12 55.86 76.12 52. 25.1 ± 0.34 147 ± 0.13 Qout (W) 440 ± 18 446 ± 19 447 ± 19 447 ± 19 448 ± 19 452 ± 19 459 ± 19 460 ± 20 463 ± 20 452 ± 20 533 ± 22 533 ± 22 533 ± 22 531 ± 22 539 ± 23 547 ± 24 546 ± 23 545 ± 24 546 ± 24 538 ± 24 h (W/m2-K) R (K/W) 0.0 ± 0.7 ± 0. mcp = 0.34 152 ± 0.6 ± 0.43 140 ± 0.123 ± 0.86 68.131 ± 0.1 ± 0.8 ± 0.0072 0. ar (g) Qin (W) 500 500 500 500 500 500 500 500 500 500 600 600 600 600 600 600 600 600 600 600 0.012 2.5 ± 0.152 ± 0.8 ± 0.125 ± 0.123 ± 0.6 ± 0.1 ± 0.17 10 ± 0.122 ± 0.5 ± 0.13 54.86 76.34 111 ± 0.0 ± 0.12 52.1 ≤ Tamb ≤ 30.1 ± 0.158 ± 0.0 ± 0.0063 0.0058 0.86 76.13 8.12 55.21 Te (°C) Te/cc (°C) 69.86 65.0053 0.50 148 ± 0.42 133 ± 0.12 59.12 58.12 57.1 ± 0.0075 0.3 ± 0.136 ± 0.090 6. continued.12 52.34 145 ± 0.0072 0.86 Tcp (°C) 52.12 55.012 0.0070 110 ± 0.160 ± 0.143 ± 0.86 65.35 114 ± 0.13 54. 37.1 ± 0.2 ≤ Tcp ≤ 59.9 ± 0.0052 0.173 ± 0.7°C.152 ± 0.012 2.012 0.41 109 ± 0.1 ± 0.0 ± 0.2 ± 0.012 0.8 ± 0.0053 0.1 ± 0.86 64.125 ± 0.012 0.131 ± 0.2 ± 0.0 ± 0.9 ± 0.34 140 ± 0.0060 0.86 64.12 55.0067 0.1 ± 0.136 ± 0.0052 0.8 ± 0.5 ± 0.090 6.012 0.7 ± 0.0 ± 0.86 76.0055 0.9 ± 0.8 ± 0.0058 0.0060 0.2 ± 0.1 ± 0.159 ± 0.34 115 ± 0.86 71.42 108 ± 0.12 56.86 71.7 ± 0.34 117 ± 0.9 ± 0.0 ± 0.6 ± 0.86 69.5 ± 0.165 ± 0.8 ± 0.0067 0.0052 0.36 141 ± 0.1 ± 0.0 ± 0.86 71.3 ± 0.0055 0.2°C).17 10 ± 0.6 ± 0.42 110 ± 0.160 ± 0.050 4.13 8.7 ± 0.21 0.12 52.12 56.86 69.123 ± 0.012 0.8.0 ± 0.12 53.86 69.162 ± 0.0067 0.86 76.050 4.12 52.6 ± 0.0066 0. Steady state operating characteristics of the rotating LHP (Qcc = 0 W.0 ± 0.36 107 ± 0.86 72.

9.8 0. ar (g) Tamb (°C) Tcp.3 -4.min (°C) h (W/m -K) R (W/K) ΔTsh (°C) 2 73 .6 56.1 26.169 11.1 51.0 1560 0.9 71.0077 kg/s).& Table 2.2 38.max (°C) Tbayonet inlet.8 54.116 11.2 4.1 51.7 41. mcp = 0.2 0.9 60.1 -4.9 47.2 -4.0 41.9 57.4 Tcp (°C) Te (°C) Tbayonet inlet.min (°C) Te/cc.6 Δ -5.1 -180 0.5 -3.7 42.1 1380 0.053 0.max (°C) Te/cc.1 3.out (°C) 0 31. Comparison of quasi-steady states for Qin = 200 W (Qcc = 0 W.in (°C) Tcp.7 4.4 36.1 46.3 50.1 66.

MIL-HDBK-310: Global Climatic Data for Developing Military Products. T. J. C. Portsmouth. 2000.C. Colorado Springs. "Testing of the Geoscience Laser Altimeter System (GLAS) Prototype Loop Heat Pipe. DOD. Tang. AIAA-94-1965. R. 2003. Ku. New York City: Wiley. J. W. HI: IEEE. S. VA: AIAA. D. Honolulu. 2002.. A. Chuang.. Conroy. Incropera. 681-686." Applied Radiation and Isotopes 61. Introduction to Flight. 74 . NV: AIAA. Hanna. A. 1999. Heat Pipe Science and Technology. Materials Park. 1994. Douglas. "Transient Modeling of Loop Heat Pipes. CO: AIAA. Materials Properties Handbook: Titanium Alloys.. Boston: McGraw-Hill. Reno. Washington D. AIAA 99-0473. 2004: 701-705. J." 6th AIAA/ASME Joint Thermophysics and Heat Transfer Conference.REFERENCES Anderson." 37th AIAA Aerospace Sciences Meeting and Exhibit. T. Brenizer. Welsch. F. J. DeWitt. A.. AIAA 20036082.. D. "Comparison of Hydraulic and Thermal Performance of PAO and Coolanol 25R.... "Study of a Loop Heat Pipe Using Neutron Radiography.. Cimbala. J.. El-Ganayni. J. 1997. D. Ghajar. Faghri. Kaya. Ku. Riley. Beam. "A Status of the United States Air Force's More Electric Aircraft Initiative. 1994. 1995. A.. G." Energy Conversion Engineering Conference." 1st International Energy Conversion Engineering Conference. Boyer.. 1997. OH: ASM International.. Cloyd.. Hoang. E. Fundamentals of Heat and Mass Transfer. Collings.: Taylor and Francis. J.

P. R.. Tables." 29th International Conference on Environmental System. 2000b: 2000-01-2489.Kaya. D. "Mathematical Modeling of Loop Heat Pipes. J. V. G. Heat Transfer Apparatus. M. L... United States Patent 4. Rogers. CA: IEEE.. New York City: Wiley. San Diego. Ku.. "Review: Loop Heat Pipes. T. Reno. W. Hoff. 1985. Dolgirev. "Experimental Investigation of Performance Characteristics of Small Loop Heat Pipes. D." Applied Power Electronics Conference and Exposition. Ponnappan. Part 2: Temperature Stability. Vershinin.209. 75 .. NACA.. 2003. J...S. "More Electric Aircraft. Kaya. China. Maidanik. Part 1: Start-Up. Montgomery. Denver." Applied Thermal Engineering 25. NV: AIAA. J. Kaya. J. Quigley. C.515. P. 906-911. 2003. L. Rogers. Cheung. Runger. "Testing of a Loop Heat Pipe Subjected to Variable Accelerating Forces. Government Printing Office. CO: SAE. Beijing.. B-19-1 . C.. Ku. Washington. AIAA 2003-1038. Hoang. Reno.. Ku. T. J. T. Ottenstein. T. 1992.... J. R. 1999. Ottenstein. 1999.." 37th AIAA Aerospace Sciences Meeting and Exhibit.: U. J. NV: AIAA. Ku. Applied Statistics and Probability for Engineers." SAE." SAE. K.. J." Proceedings of the 8th International Heat Pipe Conference. 1999-01-2007... "Operating Characteristics of Loop Heat Pipes. and Charts for Compressible Flow. Ku. 1953. Yerkes. Kholodov. Beam. Hoff." Proceedings of 41st Aerospace Sciences Meeting and Exhibit. "Analysis and Testing of Heat Pipe in Accelerating Environment. 1993. S. Chang.C. "Testing of a Loop Heat Pipe Subjected to Variable Accelerating Forces. 2000a: 2000-012488. Report 1135: Equations. A9916362. 2005: 635657. Kaya. J. T.B-19-6. Maidanik.

" AIAA Journal of Thermophysics. No. 1992: 921944. 120. Vrable. 1988. Sarno. C. K.. No. K. VA: SAE. P-322. White.. 1998. Heat and Mass Transfer. D." AIAA Journal of Thermophysics and Heat Transfer.. 2001: 384-394. "Quasi-Steady-State Performace of a Heat Pipe Subjected to Transient Acceleration Loadings. "The Effects of Transverse Acceleration-Induced Body Forces on the Capillary Limit of Helically Grooved Heat Pipes. Williamsburg. K. Yerkes. Vol. 1996: 306-309. M. 76 . F. 11. Klasing. S. 2. Vol." Aerospace Power Systems Conference Proceedings.. K." SAE. 4. K. Yerkes... Thomas. 1998: 441-451. Yerkes. Yerkes. Beam. J. "Arterial Heat Pipe Performance in a Transient Heat Flux and Body Force Environment. 15.Thomas." ASME Journal of Heat Transfer. Zaghdoudi. "A Thermal Management Concept for More Electric Aircraft Power System Applications. New York City: Addison-Wesley. S. "Investigation of the Effects of Body Force Environment on Flat Heat Pipes.

Check all bolts to ensure all experimental apparatus are tightened properly. and the E-stop button has been depressed. Press the ‘SYSTEM ENABLED’ button on the front panel so that it is illuminated. 2. The control panel box is located in Control Room 2. 3.1 for picture.2 for picture of control panel box. It is up to the experimentalist to write the program used to control data acquisition and table voltage. c. potentiometer is set to zero (turned completely counterclockwise).1. Ensure the control panel box is in “Man” mode. See Figure A. OPERATING PROCEDURES A. 4. Ensure the main power breaker is in the “OFF” position. Prior to each set of experimental testing: a. Mount the appropriate counter balance weight at the appropriate location to place the centroid in the center of the table and level each spoke to minimize vibrational noise and table runout. designed for a payload mass subjected to a maximum table capacity of 12 g.APPENDIX A. Open the LabVIEW VI file needed to control the data acquisition and table voltage. Sample programs that perform these tasks are available. 77 . Mount test payload with sufficient factor of safety to ensure a reliable mounting The retaining method will be configuration consistent with the generated forces. Standard Operating Procedure 0. d. Press the ‘Run’ button located in the top-left of the LabVIEW toolbar. See Figure A. See Figure A. 1. The breaker is located on the 71B H-bay second floor landing.3 for sample. Ensure all centrifuge maintenance has been completed. This will cause the program to become functional. Software Startup Procedure a. b. 5.

6. 7. Secure outer doors. Make final check on the table for tools or loose objects.5 for picture. Clear the table of foreign objects. f. 78 Controlling the System . e. c. See Figure A. Ensure all free standing equipment and furniture are securely placed along the perimeter of the room. Place “Test in Progress Do Not Enter” sign on the outer door. including removing all tools and placing in the appropriate CTK.4 for location.7 for picture. Turn on recirculating chiller located on the east wall.6 to enable operation. b. While the ‘Table Voltage’ switch is ‘ON’.8 for location. A booster pump for the chill bath coolant is available. Check chill bath coolant flows and flow rates of any intermediate flow loops. Ensure the emergency stop button is activated and the potentiometer is turned completely counter-clockwise on the control panel box prior to proceeding with powering the centrifuge table motor. Warning beacon switches are located outside Test Cell 4 and on the west wall of control c. d. Check camera operation. g. Follow the chill bath plumbing schematic in Figure A. See Figure A. See Figure A. The booster pump control is remotely located in Control Room 2. This provides power to the instrumentation and devices on the table. Turn table power switch on. location.1 for Unlock main power breaker and flip to ‘ON’ position. room 1. Engage table motor control switch on the north wall (cooling motor will be operational).b. e. See Figure A. See Figure A. a. Lock inter-connect door. To control the voltage supplied to the table. a. turn the mode switch from ‘MAN’ to ‘AUTO’ on the control panel box and flip the ‘Table Voltage’ switch to the ‘ON’ position. h. d. Turn on warning beacon and evacuate personnel. Check out instrumentation for proper operation.

Press the ‘SYSTEM ENABLED’ button on the LabVIEW VI so that it is no longer illuminated. 10. The program will stop. 9. 79 . 8. Table voltage can be turned on and off as often as desired while the system is engaged. If for any reason an emergency should occur press the ‘E-STOP’ on the control panel box. TC01. On the control panel box. b. A. TC05. • • • Andrew Fleming Larry Byrd Travis Michalak 58942 53238 64429 experimentalist to develop.01 ≤ dT/dt ≤ 0. Wait for the table to stop rotating then press the ‘E-STOP’ button. Flip the ‘Table Voltage’ switch to the ‘OFF’ position on the LabVIEW VI. c.2.01 K/min. TC08. turn the switch from ‘AUTO’ to ‘MAN’ and press the ‘STOP’ button. Allow LHP to achieve steady state operation by examining dT/dt plot of TC00. Set chiller to Teg = 35°C. TC06. TC04. Allow to come to steady state. & Turn high temperature coolant loop on and set to m = 0. c. Test procedures are experiment dependent and up to the Shutdown a. Should the table “run away” or suddenly accelerate the motor will automatically shutdown. and TC09. Steady state is achieved when all of these thermocouples are bracketed by -0. apply desired heat load to evaporator. Emergency Shutdown a. Slowly reduce the table voltage to zero using the slider bar. Stationary Operation a.the voltage can be adjusted using the slider bar on the left side called ‘Table Voltage’. Test Procedures 1. TC07. Simultaneously.0077 kg/s. Conduct test procedure. b. Contact the appropriate personnel prior to a restart after an emergency shutdown.

c. Decrease radial acceleration to ar = 0.0077 kg/s.01 ≤ dT/dt ≤ 0. apply desired heat load to evaporator. 80 . & Turn high temperature coolant loop on and set to m = 0. Remove heat load from the evaporator and turn off the high temperature coolant loop or adjust to next desired heat load and repeat (c). Steady state is achieved when all of these thermocouples are bracketed by -0. If continuing testing. Again.01 K/min.01 ≤ dT/dt ≤ 0. TC04. Allow to come to steady state. Increase radial acceleration to desired level.01 K/min. then remove heat load from the evaporator and turn off the high temperature coolant. d. TC07. and TC09. TC01. allow LHP to achieve steady state operation by examining dT/dt plot of TC00. TC08. TC01. h. TC08. TC07. allow LHP to operate for 30 min. Allow LHP to achieve steady state operation by examining dT/dt plot of TC00. Increase radial acceleration to ar = 0. Steady state is achieved when all of these thermocouples are bracketed by -0..1 g.1 g. Decrease radial acceleration to ar = 0. Simultaneously. TC04. TC07. Set chiller to Teg = 35°C. allow LHP to achieve steady state operation by examining dT/dt plot of TC00. and TC09. e. TC05. g. Steady state is achieved when all of these thermocouples are bracketed by -0. TC06.01 ≤ dT/dt ≤ 0. Rotational Operation a. Repeat (e)-(h). and TC09.0 g. f.d. TC08. TC04. TC06.01 K/min. If shutting down. TC05. TC06. TC05. b. 2. TC01.

(a) (b) Figure A. 81 .1. (b) Centrifuge table main power breaker. Centrifuge table main power breaker: (a) Electrical panel MCC-6.

82 .Figure A.2. Centrifuge table control panel box.

83 .3.Figure A. Sample LabVIEW control program.

4. 84 . Centrifuge table power switch.Figure A.

Figure A. 85 .5. Neslab recirculating chiller.

86 .6.To Centrifuge Table From Centrifuge Table 6 F/M 0-5LPM F/M 0-25LPM F/M 0-50LPM To Drain Recirc Filter Supply F 1 2 4 5 Recirculating Chiller Return Booster Pump P 3 Figure A. Chill bath plumbing schematic.

Booster pump control panel. 87 .7.Figure A.

8. 88 . Centrifuge table motor control power switch.Figure A.

02% and wick diameter and length measurements. 151 data points were collected representing five minutes of data. and n is the number of data points in the sample. and calculated quantities. compensation chamber.2) 89 .1 g and ±0.PAO out in (B. The fixed error of the accelerometer was ±0. constant. σ is the sample standard deviation.1) The confidence interval was based off a statistical t-distribution with confidence level of 0. the uncertainty was 0. For the specific heat of PAO. and h .005 volts for voltage measurements. the largest uncertainty of the individual measurements was used as the uncertainty of the average value. (Montgomery. 2003) Confidence Interval √ (B. For averaged quantities including Tcp .5% of the total value plus difference between the specific heat using the upper limit of the temperature measurement and the average specific heat. and preheater with an uncertainty of ±0. Constant quantities included precision resistors used for current measurement for heat input via evaporator. Calculated quantities used uncertainty methods that would be the most conservative for the experiment. UNCERTAINTY ANALYSIS The uncertainty analysis for this experiment was broken into the following three different categories: measured. For each steady state condition. The calorimetry of the cold plate. The fixed error for thermocouples and mass flow meter was determined via the calibration methods and can be seen in Appendix C. Qout.APPENDIX B.95 and was given by Montgomery where t is a tabulated value based on the confidence level and number of degrees of freedom.0000254 m. Measured values were averaged and uncertainties were calculated based on the fixed error of each instrument plus the confidence interval for the average. Te .00025*V + 0. with an uncertainty of ±0. was given by out p.

PAO (B.5) out out out in p.The uncertainty of Qout was determined to be Δ out out Δ out out out p.9) Δ e Δ e v Δ v where 1 out e v (B.PAO (B.6) p. was given by out e v (B.10) 90 .PAO / out in (B.PAO out in (B. h .7) The average evaporative heat transfer coefficient.PAO out in (B.4) out p.3) Δ out Δ in where out p.8) The uncertainty of h was given by Δ Δ Δ / out out (B.PAO Δ p.

18) 1 cp out (B.12) e v (B.16) e out out cp (B.13) e e v (B.17) 1 e out (B. was given by e out cp (B. R.14) v e v The thermal resistance.11) e v (B.(B.19) 91 .15) The uncertainty of the thermal resistance was given by / Δ where out Δ out Δ e e Δ cp cp (B.

Second.50 silicon oil. Thus.1. As a result. where the error needed to be reduced as much as possible. The Hart Scientific 6330 Calibration Bath was not capable of maintaining a steady temperature below 40°C for 92 . LabVIEW software needed to be written to interface with the calibration bath. The grounded probe thermocouples were used in the calorimetry of the cold plate.APPENDIX C. and the ambient temperature inside the box. complete calibration curves for each of the thermocouples needed to be developed. there were four grounded probe thermocouples and twelve exposed tip type E thermocouples. Control of a calibration bath and RTD were integrated with the centrifuge table’s data acquisition unit. First. The twelve exposed tip thermocouples were mounted on the loop heat pipe in various locations. In the loop heat pipe experimental setup. The bath was capable of producing steady state temperatures from 40 to 280°C. the calibration of the thermocouples needed a second calibration bath that was capable of achieving the temperature range of 20 to 40°C. The calibration bath used Dow Corning 200. A LabVIEW program was generated to simultaneously control all three devices. it was decided to only calibrate these four over the anticipated temperature range. from 20 to 145°C. and data acquisition system. Finally. in 5°C intervals. There were several steps required before a thorough calibration of the thermocouples could be determined. RTD. the characteristics of the calibration bath and RTD needed to be determined with respect to response times and temperature fluctuations. temperature measurement for the flow meter calibration. CALIBRATION OF THERMOCOUPLES AND FLOW METER C. Thermocouple Calibration The calibrations of the thermocouples used on the centrifuge table required new programming since the data acquisition system was upgraded for the centrifuge table. They needed to be calibrated over the full 20 to 230°C temperature range. The thermocouples were calibrated using a Hart Scientific 6330 Calibration Bath and Hart Scientific 1502A NIST-Traceable platinum resistance temperature detector (RTD).

The bath used for this portion of the calibration was a Brinkmann Lauda RCS 20-D calibration bath using Brayco Micronic 889 (polyalphaolefin or PAO) as the bath fluid. This number of RTD readings was selected for a 95% probability and a confidence interval of 0. a method was devised to determine that the calibration bath had reached a steady state. an insulated copper tube with a closed base was placed in the bath. The front panel and wire diagram for the automatic calibration program can be seen in Figure C. the full calibration process began by placing the thermocouples in the Hart Scientific calibration bath for the temperature range 40 to 230°C in 5°C increments ramping up and down with a standard deviation threshold of 0. Once these sub-VI’s were developed.005°C.3. Due to electrical noise in the centrifuge table test cell and fluctuations in the calibration bath temperature. respectively.95 (Montgomery.50 bath fluid. and the standard deviation of the sampling was calculated. the program would indicate that the system had reached steady state and the thermocouples were then read. When the standard deviation dropped below the specified The bath temperature would then be threshold of 0. Initially. and the temperature set point for the calibration bath was set manually rather than through the data acquisition interface. The first step in the development of the thermocouple calibrations was to write the LabVIEW software to interface with the calibration bath.4 while the front panel and wire diagram for the manual calibration program can be seen in Figure C. This significantly reduced the temperature fluctuation in the bath temperature. the thermocouples and RTD were cleaned and placed in the 93 . The previous 100 RTD temperature values were recorded into an array. incremented or decremented as necessary.1 through Figure C.Dow Corning 200. and calibration bath current temperature read can be found in Figure C. After some examination. RTD. Wire diagrams of the RTD read. First. and data acquisition.005°C. As a result. it was noticed that due to the limited capability of the Brinkmann bath to heat and cool. calibration bath temperature set point. Then. and the copper tube was filled with PAO. 2003). sub-VI’s were developed to interface directly with the calibration bath for reading the current temperature and setting the bath temperature set point as well as for reading the RTD. The same RTD was used as before.5. they needed to be incorporated into a larger framework. it was difficult to maintain a fairly constant temperature in the PAO.

were implemented to evaluate maximum deviation.0055°C over the entire temperature range. For data reduction. setting the bath to 40°C. thermocouple in tabular form. from first to fifth order. After this process. 2. and can be seen in Table C. higher-order polynomial trend lines.6. and the confidence interval associated with the 100 data points in the sample of the thermocouples. The data from each process was combined to produce one composite data set constituting the entire temperature range. A sample plot of TC00 is given in Figure C. The error inherent to the RTD was ±0. Turn on the centrifuge table power. as shown in Figure C. The maximum confidence interval of the RTD from the 100 readings was ±0. Mount all of the thermocouples to the RTD probe with the thermocouple and RTD tips as close to each other as possible. due to the increment and decrement of the calibration process.3 along with the total error of each thermocouple.009°C. 4. Calibration Procedure 1.1. Plots of RTD versus each thermocouple were generated. Turn on the RTD and calibration bath. Also. 3. C. 94 Table C. the maximum deviation of the calculated temperature from the measured temperature. A fifth-order polynomial was selected since it reduced the maximum deviation by approximately a factor of 4 over a first-order trend line. and polynomial trend lines were fit for each thermocouple.2 shows the trend line equations for each .Brinkmann bath for the temperature range 20 to 35°C.1. the maximum confidence interval of the RTD temperature over the 100 readings used in the calibration. there were two data points for each nominal temperature value. with the probes not touching any of the bath surfaces. The maximum deviations of the calculated temperature and the measured temperature are thermocouple specific. in an effort to reduce maximum deviation of the actual versus calculated RTD values.7. The uncertainty associated with each thermocouple was determined by accounting for four sources of error: maximum measured uncertainty inherent to the RTD.1 for TC00. Results from this analysis can be seen in Table C. Place the thermocouple and RTD bundle vertically into the Hart Scientific 6330 Calibration Bath. all 100 data points constituting one nominal temperature value were averaged and the confidence interval was calculated for the RTD.

2. 16. and an Omega FTB-9506 turbine flow meter with FLSC-61 signal conditioner. remove and clean the thermocouples and RTD. turn on LabVIEW and open the thermocouple calibration program. 8. Due to the chemistry of PAO. 15. A LabVIEW 95 . 13. 10. Increment the bath temperature by 5°C up to 35°C and back down to 20°C. Set the standard deviation threshold to 0.005°C. Ensure the “Keep Running” button is depressed. A calibration setup was developed using a Lauda RCS-20D calibration bath filled with PAO. Flow Meter Calibration The calibration of the turbine flow meter for the high-temperature fluid loop proved to be a difficult challenge. After this calibration cycle has been completed. 11. In the control room. 7. 9. As a result. C. Open the manual thermocouple calibration VI. Turn the bath on and set to 20°C. and place in the copper tube that is located in the Brinkmann Lauda RCS 20-D calibration bath. 6. then run the program.005. Tuthill pump from the high-temperature fluid loop. Set the lower temperature set point to 40°C and the upper temperature set point to 230°C. Current temperatures can be examined while the calibration is in progress on the main screen. 12. calibrated type-E thermocouple that was used for measuring temperature entering the flow meter in the high temperature fluid loop. with a temperature increment of 5°C. there is a significant difference in density and viscosity with respect to temperature. Wait for the RTD standard deviation to reach below 0. Take the two data files and combine to make one composite data file. Run the VI. SS-56S6 Whitey inline filter with a 140 micron filter.5. the press the ‘Proceed to TC Read’ button. a calibration surface that was dependent on flow meter output voltage and temperature was required to determine the actual mass flow rate. 14. recording temperature values at each location an overall two times. rebundle. An uncertainty analysis was also performed to fully characterize the flow meter.

The voltage from the flow meter and the temperature from the thermocouple were recorded during this time. All data was collected through the centrifuge table slip rings to the data acquisition to capture all error inherent to the centrifuge table test bed. several issues were encountered.code was developed to aid in the calibration of the flow meter (Figure C.4. Flow straightening sections upstream and downstream were placed according to the manufacturer's instructions. The flow meter required an input voltage between 12-28VDC and output voltage between 0-10V. and mass flow rate (Figure C.002 kg/s. During the course of developing the calibration setup and testing. which allowed the entire flow system to reach a steady temperature. and when the basin was full. with raw data files generated. it was noticed that the flow meter would not output a voltage linearly as expected.025 kg/s in intervals of approximately 0. A 3-D paraboloid regression equation was generated using SigmaPlot to relate temperature.0064 to 0. as many data points as possible were collected across the time span with the limiting factor being the iteration time on the LabVIEW software. A calibrated Mettler PC4400 scale was used to determine the mass collected during a given test run. the flow was again diverted to recirculating the PAO back to the chiller. The pump was controlled via a 0-10V input signal to a TECO FM50 motor controller. The minimum number of data points collected for any given run was 437. It was discovered that the motor housing of the pump was not properly grounded. This code served as the stopwatch and data recording for the calibration. The voltages and temperatures were averaged and a confidence interval of 0.95 was calculated based on a statistical t-distribution for each test run.8). These tests were completed over the range of T = 20 to 120°C & in intervals of 25°C and flow rates ranging from m = 0. The test was repeated for a total of five averaged data points for each nominal temperature and flow rate. A three-way valve was installed after the flow meter. Once the temperature was steady. the flow was diverted to a catch basin for a specified amount of time. but rather responded in a quadratic fashion with a local maximum at approximately six volts.10).9. flow meter voltage. The output 96 . A schematic of the setup is shown in Figure C. Temperature and flow meter voltage were recorded. Immediately from the start of working out the bugs in the system. causing electromagnetic interference to disrupt the operation of the flow meter. During each measurement. The equation can be seen in Table C.

With this type of calibration.4) and 97 . Initially an SS-56S6 Whitey inline filter with a 140 micron filter was placed in the flow loop to catch any debris that may have been picked up by the pump. occasionally creating an open circuit loop. another data point would be collected to replace it.0. the output voltage would develop a trend.5 volts at apparently random times.2) which yields the uncertainty / Δ where m/t m/t Δ m/t Δ (C. and deviation of the fit equation from actual data. This was likely due to fluctuations in the flow rate and was observed more prevalently at higher flow rates. error due to the voltage confidence interval and thermocouple error and confidence interval.voltage was extremely inconsistent. changing +/. At apparently random times the flow meter would start outputting random voltage readings. air bubbles were being rejected. All of the fittings were retightened and the flow remained steady. Three types of uncertainty were identified associated with this calibration: error associated with the scale and time. It was then noticed that on the flow loop outlet. It was discovered that screw terminal was not tightened down on the wire connector.11. it is critical to have a firm grasp on the uncertainty associated with the mass flow rate.3) m/t 1 (C. Each of these errors are described below. as seen in Figure C. The total error associated with this uncertainty analysis is given by Δ tot Δ m/t Δ V/T Δ dev (C.1) & The error for Δmm/t was determined by (C. During the course of data collection. When this was observed during data reduction.

9) The error associated with the deviation of the fit equation from the actual data is given by Δ dev a p (C.0% was calculated and imposed over the entire calibration range.8) and V/T 2 (C. 98 . 4. C. Connect monitor. and mouse to extensions in Test Cell 4. Using the LabVIEW program.6) which yields the uncertainty / Δ where V/T V/T Δ V/T Δ (C. Calibration Procedure 1. 6.” 3. Turn on the motor control unit. Allow the flow meter to come to temperature using PAO from the calibration bath that is by-passed back to the bath. 2.7) V/T 2 (C.0). increase the motor control input voltage to the desired setting (1. keyboard.10) Applying this uncertainty analysis to the data set. Ensure three-way valve is in by-pass mode (flow diverting back to tank).m/t (C. Activate LabVIEW code entitled “Flow meter Calibration.2. 5. a maximum error of 4.1.5) & The error for ΔmV/T was determined by performing an error analysis on the fit equation developed using SigmaPlot (C. Turn on Lauda calibration bath and set to 20°C.

When ready. Quickly return the three-way valve to by-pass mode while simultaneously depressing the “Timer” button on the LabVIEW front panel. 99 . Carefully place the filled flask on the scale and record the mass in the program. Tare the scale.0V up to 10. quickly move the three-way valve into measurement mode while simultaneously depressing the “Timer” button on the LabVIEW front panel. 11. increasing the motor control input voltage by increments of 1. Allow the flask to fill for 45 seconds for pump voltage settings one to five.7. 9. then increasing the temperature by 25°C up to 120°C. Repeat steps 6 through 12. Take the empty 2 L flask and place on the scale. Return the collected PAO to the calibration bath.0V for a given temperature. 12. Remove the flask from the scale and place under the three-way valve. 10. 8. and for 30 seconds for pump voltages settings six to ten. User will be prompted to enter the mass collected.

LabVIEW sub-VI wire diagram for calibration bath temperature set. LabVIEW sub-VI wire diagram for RTD read. 100 . Figure C.1.2.Figure C.

101 .3. LabVIEW sub-VI wire diagram for calibration bath temperature read.Figure C.

4. (b) Wire diagram 102 .(a) (b) Figure C. LabVIEW VI for controlling the automatic thermocouple calibration: (a) Front panel.

5.(a) (b) Figure C. 103 . LabVIEW VI for manual thermocouple calibration: (a) Front panel. (b) Wire diagram.

RTD temperature vs.6. time from the thermocouple calibration procedure.250 RTD Temperature (°C) 200 150 100 50 0 0 20 40 Time (hrs) 60 80 Figure C. 104 .

thermocouple plot for TC00. Sample RTD vs. 105 .160 140 120 100 RTD 80 60 40 20 0 0 50 TC00 100 150 Figure C.7.

(a)

(b) Figure C.8. LabVIEW VI for flow meter calibration program: (a) Front panel; (b) Wire diagram.

106

Figure C.9. Schematic of flow meter calibration loop

107

0.03
(kg/s)

0.025 0.02 0.015 0.01 0.005 0 20 45 T (C) 70 95 2.8 120 6 4.4 9.2 7.6 V (V)

Figure C.10. Temperature and flow meter voltage versus mass flow rate calibration curve for the high-temperature fluid loop flow meter.

108

Sample data collected during one time run for the flow meter calibration.(a) (b) Figure C. (b) “Trend” bad data set. (a) “Shotgun Blast” good data set. 109 .11.

16 Fourth 0. Maximum deviation of calculated RTD and experimental RTD corresponding to each order of polynomial for thermocouple TC00.15 Fifth 0.Table C. Polynomial Order Maximum Deviation (°C) First 0.1.48 Second 0.11 110 .45 Third 0.

12552062722017E-05 -4.Table C.19147483368683E-05 -4.57060445428729E-01 -6.18132574798607E-03 -5.70780118608180E-01 7.47827033736023E-10 -2.2.43858633628823E-07 1.72411097504291E-10 -2.81222079068205E-01 7.37758528400691E+00 3.98601477917870E-03 4.96300850243819E-01 7.09162481723344E+00 1.88677461307058E-05 -4.88538977221388E+00 3.21678343228003E-05 -4.84666399157782E+00 3.38190520875348E-03 -4.82732284362794E-07 1.08082585107749E-05 -3.25067859755088E-05 -4.71478424670112E-07 1.29518294797893E-09 -2.46693742753094E-05 -4.64688850212116E-03 4.83472149016743E-01 7. Coefficients for the trend line of each thermocouple.41571348824236E-02 3.80135796631358E-01 7.11271583430979E+00 7.05652835054604E+00 1.48342058180827E-10 -2.22250860114073E-07 -5.75690833725901E-01 7.74438732820078E-07 1.82179382148322E-01 7.72786905067963E-07 1.59329657844732E-03 4.87573183171731E-05 -4.24414330959761E-05 a0 4.54933199706354E-10 -2.56271525700294E-07 1.53272155262545E-10 -2.60500167972660E-10 -2.69915288517716E+00 3.17841766477268E-09 1.48346921511447E+00 a4 -4.34382548031174E-10 -2.45690722980988E-03 -4.58026057043238E-10 -2.71731670248675E-02 1.49099219277048E+00 3.24299704012793E-05 -4.59630318025552E-03 4.43144596520976E+00 3.23532227941934E-09 1.52702317872843E-03 4.19508133665537E-05 -4.02404183034797E-03 4.90836909105568E-01 7.72087402595724E-07 a5 1.74284360401172E-01 a2 -3.67608851857495E-01 3.24547606721552E-03 4.65772989719964E-01 7.58974962539873E+00 3.75843052066857E-01 7.60233647710038E+00 3.48473325833062E+00 3.29581121061401E-10 -2.60604857830875E-05 -4. 2 3 4 5 Tactual = a0 + a1TTC + a 2TTC + a3TTC + a 4TTC + a5TTC 111 Thermocouple TC00 (CP In) TC01 (CP Out) TC02 (Box Ambient) TC03 (Flow Meter In) TC04 (Evap 1) TC05 (Evap 2) TC06 (Evap 3) TC07 (Evap 4) TC08 (Evap Out) TC09 (Cond In) TC10 (Cond 1) TC11 (Cond 2) TC12 (Cond 3) TC13 (Cond Out) TC14 (Bayonet In) TC15 (CC/Evap) a1 1.20519229310971E-05 -3.52121560485026E+00 3.85109600118292E-03 4.57928286232713E-03 4.69069414365516E-07 1.08395588495732E+00 1.55840771155446E-10 -2.70607369149700E-07 1.01719370757371E-09 1.79389295912649E-01 7.63174346039307E-03 4.58781158427047E-03 4.72539813453339E-07 1.54508869559325E-10 .13979027632699E+00 3.88344399395756E-07 -5.58255416015721E-07 1.07111104393385E-05 8.02302203912097E-05 7.88981194007878E-07 1.28556652723263E-03 4.82525305173107E-10 -2.12753870618840E-07 -4.66498637738964E-03 a3 6.67487608290567E-07 1.45596958297964E-05 7.

Table C.111181 TC02 (Box Ambient) 0.295217 0.098736 0.106414 0.278507 0.269226 TC13 (Cond Out) 0.307238 TC05 (Evap 2) 0.304493 TC14 (Bayonet In) 0.280751 0.293427 TC06 (Evap 3) 0.111083 TC04 (Evap 1) 0.285186 0.291867 0.292962 TC07 (Evap 4) 0.334702 TC08 (Evap Out) 0.328080 TC09 (Cond In) 0. Maximum deviation and total error for each thermocouple. Maximum Deviation ±Total Error Thermocouple TC00 (CP In) 0.278103 0.290722 TC12 (Cond 3) 0.111085 TC03 (Flow Meter In) 0.122388 TC01 (CP Out) 0.291299 TC11 (Cond 2) 0.098267 0.256994 0.285927 0.097236 0.298477 TC15 (CC/Evap) 0.3.288721 TC10 (Cond 1) 0.322196 0.274506 0.315632 0.297719 112 .280739 0.

35000226E-03 c 1.4.49811559E-05 113 .91650117E-07 d 1.69012732E-05 b 2.07704738E-03 a -4. 3-D paraboloid regression equation for high-temperature fluid loop flow meter. & m = y 0 + aT + bV + cT 2 + dV 2 y0 2.Table C.

114 . cm i 1 2 ct cm (D. The condenser endpoints were first set to coincide with the centrifuge table radius.1. Rct.1) The length from the center of the centrifuge table to the midpoint of the condenser is cm ct cos (D. However. The half-angle is given by sin c 2 ct (D. Rcm. the straight condenser section. as much as possible over its length. was found as shown in Figure D. LOOP HEAT PIPE MOUNTING The loop heat pipe was mounted onto the centrifuge table such that the centerline of the tubing coincided with the outer table radius as much as possible. The radius of the condenser midpoint. As shown in Figure D. with length Lc.APPENDIX D. A small deviation existed since the condenser section and the evaporator/compensation chamber were both straight.3) The evaporator section and the compensation chamber were also straight and the method to locate these components in relation to the outer table radius is similar to that described above for the condenser section. This induced a non-uniform radial acceleration field over the lengths of these sections that needed to be quantified. was geometrically aligned on the table so that the centerline was as close to the table radius.2) One-half of the change in radius from this point to the centrifuge table radius is used to determine the radius of the midway point of the condenser section. the evaporator section is directly attached to the compensation chamber and the length of the evaporator is different than that of the compensation chamber. Therefore.1 as well. further care was taken in determining the maximum deviation of the centerline radius of these two components from the radius of the centrifuge table.

6(d)).6 cm for a percent radial difference of 3. 115 .7%. evaporator.3 cm at location 15 at the outside edge of the tubing at each location.2 cm at locations 5 and 6 and maximum radius of 123.Stands were designed using G-7 phenolic to mount the loop heat pipe vertically with support at the compensation chamber. The loop heat pipe had a minimum radius of 119. and two on the transport lines (Figure D. condenser. Complete survey data can be seen in Table D.1. A survey was taken at 22 locations on the loop heat pipe to determine how far various portions of the loop heat pipe were from the centerline radius and at what distance that location was from the origin at point 1 (Figure 2. The entire loop heat pipe fitted within 4.2). The tops of these stands were anchored to the table to reduce deflection when the table was rotating.

1 Lc 2 R Δr = R . 116 .1.L1 R ct R cm L cm Figure D. Mounting of LHP to minimize acceleration gradient.

117 .2.18 11 3 1 2 4 6 5 7 10 9 12 8 13 16 15 14 17 19 20 21 22 Figure D. LHP survey locations.

1 10 59.1.9 120.7 122.5 7 47.9 121.9 122. Survey Location s (cm) r (cm) (tube centerline) 1 0.1 14 76.1 118 .7 121.1 18 93.9 3 26.5 5 35.1 123.Table D.7 20 101.5 121.0 2 15.4 123.7 120.1 21 119.8 11 59.1 15 92.1 9 59.4 16 93.9 120.7 19 97.2 4 35.2 122.8 122.5 6 35.0 123.6 122.5 121.3 121.7 121.1 121.5 8 59.6 120.0 122. Loop heat pipe mounting survey data.1 12 64.3 121.3 122.0 13 76.1 17 93.1 22 136.

and k a maximum deviation of less than ±0.174 10 0. 119 .56 0.1) (E. they reported a maximum deviation of +3.2) (E.36 10 4.3) 5.5% from the reported data for the property equations. including density. ρ is in kg/m3. Brayco Micronic 889 technical data from Castrol was compiled and the data was curve-fitted using a least squares approach and varying order polynomials. Cp.280 10 (E.154 p 0.88 10 3.77 10 1.022 For all of these equations. For ν.APPENDIX E. T is in Kelvin. specific heat at constant pressure. The authors report that for ρ. and k is in W/mK.4%. The equations are only valid on the temperature range -54 ≤ T ≤ 135ۤ°C. BRAYCO MICRONIC 889 TECHNICAL DATA Technical data for various properties against temperature for Brayco Micronic 889. The equations for these properties are given as 1. and kinematic viscosity were provided by Ghajar et al (1994). thermal conductivity.0157 0. Cp is in kJ/kg-K.

1. (a) ρ vs. temperature. T. 120 . T.14 0.138 (a) ρ (kg/m3) (b) k (W/m-k) 0.13 0.136 0. (c) Cp vs.128 3000 2500 (c) Cp (J/kg-k) 2000 1500 1000 500 0 -100 -50 0 T (°C) 50 100 150 (c) Figure E.1000 900 800 700 600 500 400 300 200 100 0 0.142 0. T. (b) k vs.134 0.132 0. Brayco Micronic 889 properties vs.

while the terminal strips at the top are on the stationary support. These are connected via a 40-ring slip ring. it was determined that the data acquisition system needed to be upgraded. At the start of this project. The terminal strip at the bottom of this photo is located on the rotating table. documentation for the new data acquisition hardware was reviewed so that the upgrades could be started. as well as 16 channels (four wires per channel) for voltage and current control. 121 . Each circuit from the data acquisition interface panel to the two new data acquisition modules was checked for continuity by hand. it was decided to completely evaluate the existing data acquisition wiring and document this information. Since all of this equipment was going to be upgraded. A new wiring panel was developed for the wiring coming from the stationary terminal strips above the centrifuge table back to the control room. After verifying the wiring configuration was in proper operating condition. Initially.3. Each circuit from the centrifuge table to the slip ring wiring panel in the control room was checked for continuity by hand. This information was completely documented for future reference. a new wiring panel was developed for the new data acquisition interface as shown in Figure F. Wiring on the centrifuge table terminal strip now matches the wiring coming into the centrifuge table control room. All of the wiring on the centrifuge table was rewired and documented so that it would be easier to trace wiring back to the centrifuge table control room. The original computer was a Pentium 386 running ViewDAC for data collection and reduction.1. Jumper cables were created to transfer signals coming in from the slip rings to the data acquisition wiring panel. including the data acquisition unit and computer system. The first step to updating the data acquisition system was to record the original data acquisition wiring. The wiring scheme on the centrifuge table can be seen in Figure F.APPENDIX F. CENTRIFUGE TABLE UPGRADES The previous centrifuge table data acquisition system dated back to the early 1990’s. This panel was designed to accommodate 64 channels (three wires per channel) for data acquisition. as shown in Figure F.2.

a visual computer language. and resistances. 3-wire multiplexer (E1476A). then ASCII data is returned to the user for processing. Commands were sent to read thermocouple temperatures on one of the channels of the multiplexer. Virtual instruments are subprograms that are written with certain inputs and calculated outputs. If the command were for data acquisition. etc. The command module also exchanges data and commands between the computer and data acquisition system. an 8/16 channel D/A converter (E1418A). allowing the user to request a certain output voltage or current from up to sixteen channels. a 5½-digit multimeter (E1411B). the proper output is processed. The command module serves as the main source of communication between all of the cards in the mainframe. the proper conversions for voltage 122 . First. With the new data acquisition system and computer assembled. documentation for the data acquisition system needed to be reviewed to determine the proper commands necessary to use the computer to communicate with the data acquisition system. text commands are sent from the computer to the data acquisition system. pressure transducers. if the command requests a control signal. as well as any externally applied voltages. which can greatly simplify a complicated code. Second. The software used for writing the data acquisition code was LabVIEW. single text lines were sent from the computer using Agilent’s VISA Assistant software. accelerometers.. and a sixty-four channel. Next. currents.The new data acquisition system from Agilent Technologies has a mainframe (E8408A) with four slots into which the following cards were installed: a command module (E1406A). the task of reading several voltage channels and outputting the data to the screen was accomplished. Essentially. Communication between the data acquisition system and the computer takes place via the general purpose interface bus (GPIB) or IEEE-488 protocol. Virtual instruments (VIs) for communicating with the data acquisition system started fairly crudely. the reading of several voltage channels was placed inside a timed loop such that data would be recorded at regular intervals and written to a file that Microsoft Excel could read. The multimeter reads the voltages that the multiplexer collects from thermocouples. The D/A converter is a control type module. Initially. Then.

After control of the centrifuge table was accomplished using a voltage from the D/A converter to control the angular velocity. including the appropriate flow meter.1) Voltage is related to angular velocity by (F. Centrifugal acceleration is given by r (F.2) where B is an experimental constant to be determined. pressure transducer. The next task to be accomplished with this software was control communication. acceleration. accelerometer conversions.3) Solving equation (F. etc. What was desired was a system where a certain voltage could be applied to increase the angular velocity of the centrifuge table. a virtual instrument was written that could control the output voltage of several channels along with the capability of turning them on and off at any time. This VI was then merged with the data acquisition program with appropriate Boolean commands for control.to temperature. A program was tailored for this experiment. This VI was tested and verified when data was collected for a liquid-vapor separator experiment on the centrifuge table.4) 123 . such as pumps. and data recording. Substituting this relation into the expression for centrifugal acceleration yields r (F. and pressure were written into the code so that actual data was recorded to a file. and to control a variety of other devices. it was decided to control the acceleration directly. A relation was then developed between voltage and acceleration. heaters.3) for voltage yields the relation used for deriving a corresponding voltage for a chosen centrifugal acceleration. Following a similar process to the development of the data acquisition software. r (F.

7) Substituting equation (F. an acceleration control slide bar was added to the data acquisition program by deriving a corresponding voltage output. aθ+ = 0 and a z = −1.6) and solving for the square root of centrifugal acceleration r 1 / (F. ar+ . which have an operating temperature range between -250 to 350°C. Then. a plot of voltage versus the square root of centrifugal acceleration was generated and a linear best fit regression was derived with the voltage intercept forced through the origin.7) into (F. A sample plot can be seen in Figure F. A 124 .4) yields the relation between table voltage and the magnitude of the acceleration. Solving for the radial acceleration gives r θ z (F.8) To calculate B.6) + When the centrifuge table is rotating with a constant velocity. After substituting these values in for equation (F.5) + where a θ+ .4. It was desired to have the capability of reading higher temperatures on the centrifuge table. This relation can be extended to relate voltage and the acceleration magnitude. voltage is linearly related to the square root of centrifugal acceleration. radial. experimental data relating table voltage with centrifugal acceleration was collected. a new slope needs to be found. The corresponding slope is then B g / r .Thus. and if the location of the accelerometer is changed. The current thermocouple amplifier on the centrifuge table is for Type T thermocouples. After the slope for this equation was found. and az are the accelerations normalized by gravity in the azimuthal. It is important to note that the slope is experiment specific. ct 1 / (F. and axial directions on the centrifuge table. The magnitude of acceleration is given by θ r z (F.

new Type E thermocouple amplifier (Omega OM7-47-E-07-2-C) has been installed on the underside of the centrifuge table opposite of the existing thermocouple amplifier so that either one can be used. 125 . depending on the experimental requirements specified. providing operating temperatures between -200 to 900°C.

1.Figure F. 126 . Updated wiring on the centrifuge table.

2. 127 .Figure F. Wiring panel from centrifuge table to the centrifuge table control room.

3. Wiring panel for the new data acquisition system. 128 .Figure F.

129 .5 2 1. Centrifuge table voltage versus a r+ .5 1 0.3.4.5 0 0 1 2 3 4 Figure F.5 Output Table Voltage (V) 3 2.

APPENDIX G. LABVIEW PROGRAMS 130 .

(b) Wire diagram.1. LabVIEW VI for the LHP experiment: (a) Front panel. 131 .(a) (b) Figure G.

(a)

(b) Figure G.2. LabVIEW sub-VI wire diagram for voltage output control: (a) Output on; (b) Output off.

132

Figure G.3. LabVIEW sub-VI wire diagram for data acquisition communication.

133

Figure G.4. LabVIEW sub-VI wire diagram for data analyzing.

134

Figure G.5. 135 . LabVIEW sub-VI wire diagram for data recording.

APPENDIX H. CENTRIFUGE WIRING TABLES 136 .

E1418A 8/16-CH D/A Converter wiring.Table H.1. Module Terminal Number 1 Terminal Letter HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS HS HI LO LS GND EXT TRIGn GND GND CAL HS CAL HI CAL LO CAL LS Wire Bundle Number Wire Color red white blue yellow green brown purple orange red white blue yellow green brown purple orange red white blue yellow green brown purple orange red white blue yellow green brown purple orange red white blue yellow green brown purple orange red white blue yellow green brown purple orange red white blue yellow green brown purple orange red white blue yellow green brown purple orange red white blue yellow green brown purple orange Terminal Strip Number Terminal Number 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 01 02 03 04 05 06 07 08 Function Slip Ring Wire Color Centrifuge Table Voltage (+) Centrifuge Table Voltage (-) Green-Blue Shield Black-Blue Shield 23 2 Heater Power Voltage (+) Heater Power Voltage (-) Red-Red Shield Black-Red Shield 23 3 Pump Voltage (+) Pump Voltage (-) 43 44 Red-Red Shield Black-Red Shield 24 4 5 25 6 24 7 26 8 9 27 10 25 11 28 12 13 29 14 26 15 30 16 31 27 137 .

2. Data acqusition terminal board wiring.Table H. E1476A Module Terminal Number 00 E1476A Terminal Letter H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G H L G Main Main Terminal Terminal Strip Number Number 01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09 01 02 03 04 05 06 07 08 09 Function Slip Ring Wire Color Number 3 19 4 19 5 19 6 19 7 19 8 19 9 19 10 19 11 19 12 19 13 19 14 19 15 19 16 19 17 19 18 19 21 20 22 20 23 20 24 25 26 27 28 29 30 31 32 33 34 35 36 37 red black white black blue black yellow black green black brown black purple black orange black red black white black blue black yellow black green black brown black purple black orange black red black white black blue black green black red orange white brown yellow blue red black gray purple pink tan Board Terminal Number 01 GND 02 GND 03 GND 04 GND 05 GND 06 GND 07 GND 08 GND 09 GND 10 GND 11 GND 12 GND 13 GND 14 GND 15 GND 16 GND 2 1 3 1 4 1 Location TC00 TC Ground TC01 TC Ground TC02 TC Ground TC03 TC Ground TC04 TC Ground TC05 TC Ground TC06 TC Ground TC07 TC Ground TC08 TC Ground TC09 TC Ground TC10 TC Ground TC11 TC Ground TC12 TC Ground TC13 TC Ground TC14 TC Ground TC15 TC Ground Accel x-axis Accel GND Accel y-axis Accel GND Accel z-axis Accel GND Flowmeter Flowmeter GND Evap Heater Resistor Voltage (+) Evap Heater Resistor Voltage (-) Evap Heater Voltage (+) Evap Heater Voltage (-) Preheater Resistor Voltage (+) Preheater Resistor Voltage (-) Preheater Voltage (+) Preheater Voltage (-) CC Heater Resistor Voltage (+) CC Heater Resistor Voltage (-) CC Heater Voltage (+) CC Heater Voltage (-) CP In CP Out 01 01 Box Ambient 02 Flow meter In 03 Evap 1 04 02 Evap 2 05 Evap 3 06 Evap 4 07 03 Evap Out 08 Cond In 09 Cond 1 10 04 Cond 2 11 Cond 3 12 Cond 4 13 05 Bayonet In 14 Evap/CC 15 16 06 17 18 19 07 20 21 22 08 23 24 25 09 26 138 .

8)2)+0.039 Ma∞2) + 0.3483E-1)H2 – (9.5Tw = (284.0 m = 408.9917E-2)H + 1.7664E-4)H4 – (2.97 m/s Freestream velocity: U∞ = Ma∞a∞ = (0.37 K / 300 K)0.36 K Air density at the film temperature: ρ*= ρ ∞/(T∞/T*) = (0.36 K) = 0.6870 kg/m3)(270.15 K) T* = 353.9 m2/s2·K)( 284.0868 ρ∞ = 0.5(408.772 × 10-5 N·s/m2 Reynolds number: ReL = (ρ∞U∞L)/μ∞ = (0.76 = (184.5+0.76 μ∞ = 1.8507E+1) T∞ = 284.37 K)/(353.6870 kg/m3)(284. SAMPLE CALCULATIONS Example Calculation of Average Heat Transfer Coefficient for Flat Plate Flow Given: Altitude: Mach number: Wall temperature: Plate length: Calculations: Freestream temperature: T∞ = (7.05 × 107 [TURBULENT] 139 H = 5 km Ma∞ = 0.5033)H + (4.37 K Freestream density: ρ∞ = (-4.5528 kg/m3 Freestream speed of sound: a∞ = √(γRT∞) = √((1.8994E-2)H3 + (5.4)(286.6 × 10-7 N·s/m2)(284.9336E-6)H3 + (2.37 m/s)(1 m) / (1.APPENDIX I.97 m/s) = 270.0898E-3)H2 – (8.039(0.37 K)(0.6870 kg/m3 Film temperature: T* = T∞(0.15 K .8 Tw = 135 °C L = 1.8)(337. 2002): μ∞ = μR(T∞/TR)0.37 K)) = 337.5 + 0.37 m/s Freestream absolute viscosity (Reference values from Incropera and DeWitt.772 × 10-5 N· s/m2) ReL = 1.

7 W/m2 140 .001502 Local heat transfer coefficient at the end of the plate: hL = StL*ρ*U∞Cp* = (0.06ρ*U∞L/ μ*)) = 0.7 W/m2·K Average heat transfer coefficient over the length of the plate: h = 1.37 K)(1+(0.44 J/kg·K) = 226.001502)(0.7((0.2593E-9)T3 + (1.66 K Absolute viscosity at the film temperature: μ* = μR(T*/TR)0.76 = (184.002705/2)1/2) = 0.15 K .06(0.5528 kg/m3)(270.7 W/m2·K) = 260.6418E-1) Pr* = 0.37 m/s)(1 m) / (2.L* = 0.8)2) = 316.316.6978)1/3 = 0.6978)2/3-1)(0.091 × 10-5 N· s/m2)) = 0.L*/2)1/2 = (0.8870 Adiabatic wall temperature: Taw = T∞[1+r((γ-1)/2)Ma∞2] = (284.37 m/s)(1009.4177E-4)T + (8.5528 kg/m3)(270.Prandtl number at the film temperature: Pr* = (-1.7(Pr*2/3-1)(Cf.7 W/m2·K Average heat flux dissipated over the plate: qw = h (Tw .091 × 10-5 N·s/m2 Local skin friction coefficient at the end of the plate: Cf.76 = 2.6 × 10-7 N· s/m2)(353.455/(ln2(0.8870)(0.0187E+3) Cp* = 1009.L*/2) / (1+12.002705/2) / (1+12.66 K) = 23847.9921E-2)T + (1.15(226.3333E-5)T2 – (6.4444E-7)T3 – (3.455/(ln2(0.Taw) = (260.7778E-6)T2 – (9.15 iL = 1.002705 Local Stanton number at the end of the plate: StL* = (Cf.7 W/m2·K)(408.44 J/(kg·K) Recovery factor for turbulent flow: r = Pr*1/3 = (0.4/2)(0.6978 Specific heat at the film temperature: Cp* = (4.36 K / 300 K)0.