Temperature Oscillation in a Loop Heat Pipe ith G it A i t
J t K M tt G i D k P t l
with Gravity Assist
Jentung Ku, Matt Garrison, Deepak PatelLaura Ottenstein, Frank Robinson
NASA/GSFCCode 545
Spacecraft Thermal Control Workshop, The Aerospace Corporationp p p pEl Segundo, California, March 25-27, 2014
https://ntrs.nasa.gov/search.jsp?R=20140005700 2020-07-20T08:37:43+00:00Z
Outline
• Introduction/Background
• A Theory for Temperature Oscillation with Gravity Assist and• A Theory for Temperature Oscillation with Gravity Assist andHigh Control Heater Power Requirement
ICES t 2 ATLAS LTCS TV T t R lt• ICESat-2 ATLAS LTCS TV Test Results
• Summary and Conclusions
2
Introduction
• During ICESat-2 ATLAS LTCS TV testing, the laser mass simulators could not be controlled at desired set point temperatures.p
• The LHP reservoir control heaters appeared to be under-sized despite flight analysis showing no issuedespite flight analysis showing no issue.
• An investigation of the LHP behaviors found that the root cause of the problem was the temperature oscillation of the reservoirof the problem was the temperature oscillation of the reservoir,which was in turn caused by gravity assist and a combination of other factors.
• Results of the investigation are presented.
3
ICESat-2 and ATLAS
• Ice, Cloud and land Elevation Satellite-2 (ICESat-2) is an earth observing satellite expected to launch in 2016in 2016.
• The Advanced Topographic Laser Altimeter System (ATLAS) will estimateAltimeter System (ATLAS) will estimatesea ice thickness and measure vegetation canopy height.
• Only one of the two redundant lasers in ATLAS will be used at a time. These lasers have stringent thermal control requirementscontrol requirements.
• The thermal control system were designed and fabricated while thedesigned and fabricated while theATLAS lasers were being developed.
4
ATLAS Laser Thermal Control System (LTCS)
Redundant lasers are cooled via a single Laser Thermal Control S t (LTCS) i ti fSystem (LTCS) consisting of a constant conductance heat pipe (CCHP), a loop heat pipe (LHP), and a radiator.
5
ATLAS LTCS TV Test Setup Schematic
RadiatorRadiatorShr
Heat pipe and LHP both operating in reflux
oud
CCHP GSE Heater
6
High Reservoir Control Heater Power
• In thermal balance tests, using 100% of the available heater power could not maintain the commanded reservoir temperature.
• One example:– 136W laser, -101 oC shroud– Expected Laser 1 simulator to
run at +10 oC (and reservoir at +4oC) with 10 6W control power+4oC) with 10.6W control powerbased on ATK analysis at CDR
– Test results: Laser 1 simulator ran at -14 oC (reservoir at -24 oC)ran at 14 C (reservoir at 24 C)with 22W control heater power
• The reservoir displayed persistentp y ptemperature oscillations.
7
Theory for Temperature Oscillation with Gravity Assist and High Control Heater Power Requirement
• A theory has been developed to explain the temperature oscillation and high control heater power requirement based on:– Mass momentum and energy balanceMass, momentum and energy balance– LHP operating principles
Th th i t d i th f ll i d• The theory is presented in the following order:– Thermodynamic constraints in two-phase systems– Pressure drop diagrams in LHP operation– Reservoir energy balance– Physical processes involved during temperature oscillation– Conditions leading to persistent temperature oscillationConditions leading to persistent temperature oscillation– Relevant ATLAS LTCS TV test data that partially verify the
theory
8
Thermodynamic Constraint in Two-Phase Systems
The following relation must be
SaturationCurve
The following relation must besatisfied between any components where liquid and vapor phases coexist in
PAure
PBthermodynamic equilibrium.
TB – TA = (PB – PA)/(dP/dT)A
Vapor
LiquidPre
ssu
Or,
T T = (P P ) (T �v/�)
TB
Vapor
TemperatureTA
TB – TA = (PB – PA) (TB �v/�)
9
Pressure Drop Diagram in LHP Under Gravity Neutral Environment
Vapor Channel
Primary Wick
Secondary Wick
Bayonet
P2 P1
Reservoir EvaporatorPPP6
1 P2 P3 P4P5 P6su
re
Vapor Line
LiquidLine
Condenser
P1P76 6
Pres
s
PP3P5
P4 Location
P7
PP77 (Liquid)(Liquid)
• Evaporator core is considered part of reservoir.• P6 is the reservoir saturation pressure.• All other pressures are governed by P6
PP77
��
��
(Liquid)(Liquid)
All other pressures are governed by P6• All pressure drops are frictional pressure drops.
10
PP11wickwick (Vapor)(Vapor)
Pressure Drop Diagram in LHP With Gravity Assist
P1 PP6
’
Primary Wick
Secondary Wick
Bayonet
Vapor Channel
P2 1 P2 P3
P4Pss
ure
P7’
�Pg
Reservoir Evaporator
2
P P5 P6Pres
P7
Vapor Line
LiquidLine
Condenser
P1P7P6
Location
7Condenser
P3P5P4
• Gravity assist raises the reservoir pressure from P6 to P6’• All other pressures are governed by P6’• Frictional pressure drops remain the same.
When �P > P P (i e P ’ > P ) liq id ill fall from the condenser• When �Pg > P5 – P6 (i.e. P6 > P5), liquid will fall from the condenserto the reservoir.,
11
Pressure Drop Diagram in LHP With Gravity Assist
P6’
Primary Wick
Secondary Wick
Bayonet
Vapor Channel
P1 P2 P3P4re
�Pg
P7’
Reservoir Evaporator
P2
P4 P5 P6
Pres
surReservoir Evaporator
Vapor Line
LiquidLine
P1P7P6
Location
P7Condenser
P3P5P4 Location
• Gravity assist raises the reservoir saturation pressure from P6 to P6’.• All other pressures are governed by P6’
Frictional press re drops remain the same• Frictional pressure drops remain the same.• When P7
’ > P1, liquid will be pushed into evaporator vapor grooves.12
Pressure Drop Diagram in LHP With Gravity Assist and Liquid Reverse Flow
P1 P2 P3
Primary Wick
Secondary Wick
Bayonet
Vapor Channel
P4 P5P6
essu
re
3
P1’
P ’ P5Reservoir Evaporator
P2
Pre
P
�Pg
P ’
P7’
P2 P3’
Reservoir Evaporator
Vapor Line
LiquidLine
P1P7P6
P7P5P4
’Condenser
P3P5P4
Location• Absolute pressures with a reverse liquid flow are shown in red.• The reservation pressure is at P6, which governs all other pressures.• P6 - P5 = frictional pressure drop due to reverse liquid flow.• Reverse liquid flow works against gravity, thus �Pg < 0.
13
Gravity Pressure Head with a Vertical Radiator
Radiator
Gravity
h
y
�Pg = (�l - �v)g �H
Vapor FrontVapor LineH
�H = H – h
LiquidLine
�H�H and �Pg varywith the vapor front position.
ReservoirEvaporator
Condenser
14
Energy Balance in Reservoir• For steady state operation:• For steady state operation:
QCC = Qsub – Qleak
Qsub = mliq Cp (Tcc-Tin) -Qsub QleakCC, TCC
QCC
Evap, TE
mliq = (QE – Qleak)/�
• mliq is not constant during temperature oscillation.
QRad
liq
– mliq and Qsub are increasing when Tcc is decreasing.– A reverse liquid flow occurs when Tcc is increasing, carrying
warm fluid to the condenser.warm fluid to the condenser.– Qrad represents additional heat leak during temperature
oscillation compared to steady state.• During quasi-steady of temperature oscillation*:• During quasi-steady of temperature oscillation :
Total energy loss as reservoir temperature drops from its peak to valley = Total energy provided by the control heater as reservoir temperature rises from its valley to peakfrom its valley to peak
*The control heater is turned on at all times.15
Physical Processes during Temperature OscillationReservoir Temperature Decreasing
Radiator
• Gravity causes liquid to drop from condenser to reservoir.
Radiator
Gravity
• With cold radiator, the liquid carries large subcooling.
• Reservoir temperature decreases rapidly
Vapor FrontVapor Line
h
H
decreases rapidly.• Thermal mass releases sensible
heat, increasing heat load to evaporator. Reservoir
Liquid Line
Evaporator
Condenser
�H
evaporator.• Vapor front inside condenser advances with increasing heat load and
decreasing reservoir temperature.• Liquid mass flow rate increases, causing reservoir temperature to dropLiquid mass flow rate increases, causing reservoir temperature to drop
further.• Vapor front will stop advancing because of energy balance requirement in
condenser and the decreased gravity pressure head.
16
• Control heater is always turned on. Reservoir temperature begins to increase.
Physical Processes during Temperature OscillationReservoir Temperature Increasing
• When reservoir temperature increases, thermal mass stores sensible heat.
Radiator
h
Gravity
• Vapor front inside condenser recedes with decreasing heat load and increasing reservoir temperature.
Vapor FrontVaporLineH
temperature.• Control heater causes reverse
liquid flow along liquid line, filling the space left by vapor
Reservoir
Liquid Line
Evaporator
Condenser
�H
front recession. ReservoirEvaporator
• As vapor front recedes, gravity pressure head increases, slowing down the rate of reverse liquid flow.
• Because a certain length is required to dissipate heat load from evaporator, vapor length reaches it minimum.
• Vapor front stops receding and starts advancing.Liq id drops from condenser to reser oir Reser oir temperat re begins
17
• Liquid drops from condenser to reservoir. Reservoir temperature beginsto decrease, repeating the temperature oscillation.
Sustaining Persistent Temperature OscillationRadiator
• Three driving forces sustain the persistent temperature oscillation.
G it i th
Gravity
– Gravity assist– Liquid reverse flow– Continuous control heater
power
Vapor Front
Liquid
Vapor LineH
�Hpower• Cold radiator and large thermal
mass amplify the effect of these driving forces. Reservoir
LiquidLine
Evaporator
Condenser
g• Without gravity assist, there is no persistent temperature oscillation.• The control heater power must be sufficiently large to cause a reverse
liquid flow. Otherwise, there is no persistent temperature oscillation.• The control heater power is not large enough to maintain the reservoir set
point temperature. Otherwise, intermittent “power-off’ periods will stop the persistent temperature oscillation.C ff f
18
• Causes and effects of temperature oscillation intermingled, leading to a “circular” mechanism.
Root Cause of High Control Heater Power
• During temperature oscillation, the reservoir is subjected to a repeated ingress of cold liquid
Radiator
Gravity
from cold radiator. • As the reservoir temperature is
raised by the control heater, a reverse liquid flow occurs
Vapor FrontVaporLine
h
Hreverse liquid flow occurs,carrying some warm liquid to the cold radiator.
• Before the reservoir set point
Liquid Line
H
�H
ptemperature is reached, the next round of cold liquid is injected into the reservoir.
ReservoirEvaporator
Condenser
• The control heater is tuned on at all times, and its power is consumed largely to warm the reservoir toward its set point temperature which cannot be reached with existing heater power.
• The persistent reservoir temperature oscillation is the root cause of high• The persistent reservoir temperature oscillation is the root cause of highcontrol heater power requirement.
19
Th th t b f ll ifi d b th i ti ATLAS LTCS TV
Verification of the Theory with ATLAS LTCS TV Test Data
• The theory cannot be fully verified by the existing ATLAS LTCS TVtest data.– No temperature sensors on the condenser itself
D t ll ti t f t i t i t ffi i t t if– Data collection rate of once every two minutes is not sufficient to verifyLHP transient behaviors
• Some relevant data are used to provide partial verification of the theorytheory.
• Verification with relevant ATLAS LTCS test data– Liquid drop from the condenser to the reservoir
R li id fl– Reverse liquid flow– No persistent temperature oscillation without sufficient heater power– Effects of some parameters on temperature oscillation
• Part of the theory that cannot be verified by ATLAS LTCS test data.– Vapor front movement– Mass and energy balance in the reservoir– No persistent temperature oscillation if the reservoir set point can be
maintained.20
Some Temperature Sensors on ATLAS LTCS
TCS-10
TCS-17
TCS-11
TCS-15
TCS-16SCA-02 SCB-02
• Only data from TCS 10 and TCS 11 were collected once every four seconds
21
• Only data from TCS-10 and TCS-11 were collected once every four seconds.• All other data were collected once every two minutes.
TC Locations on Radiator50
14
3 2
15
15
7
4
5467
8
54
10
9
1252
530
11
13
22
• TCs 1, 2, 3, 5, 6, 7, 9, 8, 12, 11, 14, 15 follow the condenser footprints.• Data were collected once every two minutes.
Frictional Pressure Drops (CC at -9.8oC and sink at -98 oC)
Summary of Pressure Drop (Pa)Component 49W 74W 109W 136W 175W 196W 249W 300WVapor Line 42 62 152 220 341 415 630 861CondenserCondenser
Two-Phase 27 60 141 267 534 696 1,342 2,137Liquid Phase 1 2 4 7 11 14 23 32
Subcooler 0 1 1 1 1 2 2 3Li id Li 12 17 26 32 40 45 57 69Liquid Line 12 17 26 32 40 45 57 69Capillary Pump
Liquid Core 11 17 25 31 40 44 56 67Wick 637 954 1,397 1,725 2,214 2,476 3,145 3,760Vapor Grooves 42 63 92 114 146 164 208 249Circum. Grooves 0 1 1 1 2 2 2 3
Gravitational Head 0 0 0 0 0 0 0 0Total 772 1 178 1 840 2 398 3 320 3 858 5 465 7 181Total 772 1,178 1,840 2,398 3,320 3,858 5,465 7,181Capillary Limit 41,552 41,552 41,552 41,552 41,552 41,552 41,552 41,552Mass in Condenser (grams)
92.5 89.2 84.7 81.3 76.2 73 67 60
% Pcap Used 2 3 4 6 8 9 13 17
Radiator
A
h
GravityB
Vapor FrontVapor LineH
AB = 63.5 mm (2.5 in)AD = 1212.4 mm (47.7 in)CD = 158.3 mm (6.2 in)
Liquid
H
�H
Maximum �Pg = 7450 PaMinimum �Pg = 1010 Pa
Line
Condenser
C
ReservoirEvaporatorD
• Analytical model predicts the frictional pressure drop along the liquid line
24
y p p p g qis no more than 85Pa for powers up to 300W.• According to the theory, liquid will drop from the condenser to reservoir.
• Test #1: Cold Transition, reservoir temperature was decreasingReverse Flow During Cold Transition Test
, p g• Temperatures of TCS16 and Radiator LL show that liquid drainage and
reverse liquid flow did occur alternately along the liquid line. • Oscillating reservoir temperature decreased toward its quasi-steady
t ttemperature.• Data were collected once every two minutes.
0SCA�02�LHP�CC TM�1LHP�EVAP Radiator�LL
�15
�10
�5°C)
TCS�16�LHP�LL TCS�17�LHP�VLTCS�12�CCHP
136W to thermal mass 1, shroud at -101°C, 22W of
�25
�20
Tempe
rature�(°101 C, 22W of
control heater power with set points of 4°C/5°C
�40
�35
�30
25
�40
• Test #2: Cold Soak, quasi-steady stateReverse Flow During Cold Soak Test
, q y• Temperatures of TCS16 and Radiator LL show that liquid drainage
and reverse liquid flow did occur alternately along the liquid line. • Oscillating reservoir temperature was at a quasi-steady state.• Data were collected once every two minutes.
�15
�10
�20
15ure�(°C
)
136W to thermal mass 1, shroud at -101°C, 22W of
�30
�25
Tempe
ratu101 C, 22W of
control heater power with set points of 4°C/5°C
�40
�35
26
SCA�02�LHP�CC TM�1 LHP�EVAP Radiator�LL TCS�16�LHP�LL TCS�17�LHP�VL TCS�12�CCHP
T t #1 C ld T iti i t t d i
Reservoir Temperature Oscillation During Cold Transition Test
• Test #1: Cold Transition, reservoir temperature was decreasing• Data were collected once every four seconds. • Reservoir temperature was decreasing.• In each cycle reservoir temperature decreased 2 1°C in 24 secondsIn each cycle, reservoir temperature decreased 2.1 C in 24 seconds
and rose 2.1°C in 32 seconds.30
�12
�11
20
25
�16
�15
�14
�13
r�Pow
er�(W
)
ure�(°C)
136W to thermal mass 1, shroud at -101°C, 22W of
5
10
15
21
�20
�19
�18
�17
Control�H
eate
Tempe
rat101 C, 22W of
control heater power with set points of 4°C/5°C
0
5
�23
�22
�21
27
TCS�10 TCS�11 Power
Reservoir Temperature Oscillation During Cold Soak Test
• Test #2: Cold Soak, quasi-steady state• Data were collected once every four seconds. • In each cycle, reservoir temperature decreased 2.1°C in 24 seconds
and rose 2 1°C in 32 secondsand rose 2.1°C in 32 seconds.
25
30
�18
�17
�16
15
20
�22
�21
�20
�19
ater�Pow
er�(W
)
rature�(°C)136W to thermal
mass 1, shroud at -101°C, 22W of
5
10
�26
�25
�24
�23
Control�H
ea
Tempe
r101 C, 22W ofcontrol heater power with set points of 4°C/5°C
0�28
�27
28
TCS�10 TCS�11 Power
• Test #6: Cold Transition
Reservoir Temperatures with and without Reservoir Heater Power
Test #6: Cold Transition• Reservoir temperature oscillated when control heater was turned on
continuously.• Without control heater power, there was no reverse liquid flow and no
temperature oscillation.• Data were collected once every four seconds.
4010
�80
�40
0
�5
0
5
Power�(W
)
re�(°C)196W to thermal
�200
�160
�120
�15
�10
Control�H
eater�P
Tempe
raturmass 1, shroud at
-78°C
�280
�240
�25
�20
29
TCS�10 TCS�11 Power
• Test #6: Cold Transition
Loop Temperatures with and without Reservoir Heater Power
Test #6: Cold Transition• Reservoir temperature oscillated when control heater was turned on
continuously.• Without control heater power, there was no reverse liquid flow and no
temperature oscillation.• Data were collected once every four seconds.
10
�5
0
5C)
�15
�10
Tempe
rature�(°C
196W to thermal mass 1, shroud at -78°C
�30
�25
�20
30
TC1�VL�IN TM�1 LHP�EVAP LHP�VL LHP�LL TCS�12�CCHP SCA�02�LHP�CC TCS�17�LHP�VL
• Test #6: Cold Transition
Radiator Temperatures
• No temperature oscillation on radiator at any time due to conduction and radiation effects.
• Reservoir was at its natural operating temperature without control heater power.
• Data were collected once every two minutes. 0
�15
�10
�5e�(°C)
196W to thermal
�30
�25
�20
Tempe
rature
mass 1, shroud at -78°C
�40
�35
31
TC�2 TC�3 TC�5 TC�7 TC�9 TC�12 TC�14 TC�15 SCA�02�LHP�CC
• Test #4 and #5: 196W to thermal mass 1 shroud at -78°C reservoir
Temperature Oscillation with Various Control Heater Powers
Test #4 and #5: 196W to thermal mass 1, shroud at -78 C, reservoirheaters set points at -2°C/-1°C
• Both heaters used TCS-11 as the control sensor. • Increasing heater power from 22W to 38W raised reservoir temperature
by 4.2 °C. At 38W, one of the heaters was turned on and off. The other was on at all times.
Test #4: 22W control heater power Test #5: 38W control heater power
30
40
9
11
13
15
r�(W)
20
25
30
�3
�2
�1
0
r�(W)
10
20
1
3
5
7
Control�H
eater�P
ower
Tempe
rature�(°C)
10
15
20
�7
�6
�5
�4
Control�H
eater�P
ower
Tempe
rature�(°C)
0�5
�3
�1
0
5
�10
�9
�8
32
TCS�10 TCS�11 PowerTCS�10 TCS�11 Power
• Test #5: Soak
Temperature Oscillation with 38W/19W to ReservoirTest #5: Soak
• 38W/19W of control heater power with set points of -2°C/-1°C• On/off of one control heater (19W) affected the reverse liquid flow.• Data were collected once every two minutes.y
5
10
�5
0
e�(°C
)
196W to thermal mass 1, shroud at -78°C
�20
�15
�10
Tempe
rature78 C
�30
�25
33
SCA�02�LHP�CC TM�1 LHP�EVAP Radiator�LL TCS�16�LHP�LL TCS�17�LHP�VL TCS�12�CCHP
Tracking Vapor Front Movement
• Test #2: Cold Soak• Data were collected once every two minutes. • Vapor front movement could not be tracked without temperature sensors
on the condenser itself The liquid mass flow rate cannot not be derivedon the condenser itself. The liquid mass flow rate cannot not be derived.• Reservoir energy balance cannot be verified.
�20
�35
�30
�25e�(°C)
136W to thermal mass 1, shroud at -101°C, 22W of
�50
�45
�40
Tempe
rature
101 C, 22W ofcontrol heater power with set points of 4°C/5°C
�60
�55
34
TC�2 TC�3 TC�5 TC�7 TC�9 TC�12 TC�14 TC�15
Eff t f th l i h t d
Reservoir Temperature under Various Test Conditions
• Effects of thermal mass power, reservoir heater power, andshroud temperature on reservoir temperature can be inferred from the table.
Test #
LoopStatus
ThermalMassPower (W)
ReservoirHeater Set Points (oC)
ReservoirHeaterPower (W)
ChamberShroudTemperature (oC)
ReservoirTemperature Valley/Peak (oC)( C) ( C)
1 Transient 136 +4/+5 22 -101 Decreasing, oscillation
2 Quasi- 136 +4/+5 22 -101 -25.0/-22.9steady
3 Near quasi-steady
196 -2/-1 22 -101 -16.2/-14.2
4 Near quasi- 196 -2/-1 22 -78 -8.5/-6.5qsteady
5 Near quasi-steady
196 -2/-1 38/19 -78 -4.4/-2.0
6 Transient 196 N/A 0 -78 -20.2 (still
35
(decreasing, no oscillation)
Summary and Conclusions
• The high control heater power in ICESat-2 ATLAS LTCS TV testing was caused by persistent temperature oscillation.
• With persistent temperature oscillation, the reservoir was subjected to a repeated influx of cold liquid from the condensera repeated influx of cold liquid from the condenser.
• When the reservoir temperature was increasing, reverse liquid flow brought warm fluid from reservoir to condenser.
• The control heater was turned on at all times but was unable to• The control heater was turned on at all times, but was unable tomaintain the reservoir set point temperature due to the additional heat leak to the radiator with persistent temperature oscillation.
• Persistent temperature oscillation was sustained by the combination yof gravity assist, reverse liquid flow, and inability of the control heater to maintain the reservoir at the desired set point temperature. Cold radiator temperature and a large thermal mass amplified this effect. C d ff t f i t t t t ill ti i t i l d• Causes and effects of persistent temperature oscillation intermingled.
• The theory of temperature oscillation was only partially verified using data from ATLAS LTCS TV testing due to the lack of condenser temperature data. Additional data from past or future LHP tests aretemperature data. Additional data from past or future LHP tests areneeded to fully verify the theory.
36
