Phase Change CellsT. Shane Topham, Gail Bingham, Harry Latvakoski, Mike Watson

1

Outline• Phase change cell introduction• Summary of ISS phase cell experiments• Flight cell design improvements and ground test results

2

Need for Orbital Temperature Reference• Phase transition cells for absolute temperature reference are

key components of future climate monitoring missions• Mission requirement:

“…an SI-traceable standard for absolute spectrally resolved radiance in the infrared with high accuracy (0.1 K 3σ brightness temperature… Each of the interferometers carry, on-orbit, phase transition cells for absolute temperature,… with SI traceability [1].”

• Because temperature uncertainty will only be one of the contributors to the 0.1 K requirement, absolute temperature uncertainty will need to be lower, on the order of 0.01 K or better

3

Phase Transitions as Thermal ReferencesITS90 Phase Change Materials

4

• Large volume of PCM• Long melt times• Deep re-entrant wells• Fragile containers• Detailed manual heating and cooling

proceduresPractical absolute uncertainty, ~0.2 mK or better [2,3]

T90/K t90/°C Material State Uncertainty mK

234.3146 -38.8344 Hg Triple pt 0.2(0.1)

273.16 0.01 H2O Triple pt 0.05 (0.03)

302.9146 29.7646 Ga Melt pt 0.2 (0.03)

Traditional Triple Point of Water Cell

SDL Phase Change Program History

5

2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014

(2004) Initial internal studies of potential for NPOESS started

VEGA Intl. approached to collaborate on gallium eutectic work

(2006) SDL & IBMP agree to perform ISS testing of PCM cells.

SDL starts IR&D program to get ISS tests with PCM cells and patent technology

Vega Results show Ga eutectics as viable PCMs for calibrations

ESTO office begins funding support to speed development to benefit CLARREO with space qualification results

Launch opportunities delayed by Russian Calibration Institute approvals with ties to VEGA

Hardware Launched on Soyuz, ISS experiments conducted

FY1 FY2 FY3 FY4SDL IR&D program

Hardware returned on Soyuz to IBMP and then to SDL

Mini Orbital Temperature Reference (MOTR) Flight 1 Experiment Design for ISS• PCM temperature controlled by

2 TECs and a temperature-controlled heater enclosure

• Heat exhausted to cabin through forced ventilation

• Automated experiment data stored on CF card, removable only on ground

• Expanding bellow PCM container ~1 mL volume

• Re-entrant sensor well

6

Sensor

PCM

TEC

Flight 1 Experiment Data Analysis• 19 on-orbit melt curves• Temperature calibration from

bath post-flight• Multiple similar data sets

obtained prior to and after flight

• Use average in plotted window as “melt curve temperature”

7

29.76

29.762

29.764

29.766

29.768

29.77

29.772

29.774

29.776

29.778

0.5 1.5 2.5 3.5 4.5 5.5

Tem

pera

ture

(˚C

)

Time from start of melt cycle (hrs)

Flight Melt Curves smoothed to 2.2 minutes[4]

MOTR Data Collection Timeline

29.771

29.772

29.773

29.774

29.775

29.776

29.777

29.778

8/10/2010 2/26/2011 9/14/2011 4/1/2012 10/18/2012 5/6/2013 11/22/2013 6/10/2014 12/27/2014

3 hr

Mel

t Tem

pera

ture

Ave

rage

s (C

)

Data Collection Date

SDL Predelivery(1)

SDL Predelivery (2)

Moscow Delivery

Moscow Pre-launch (1)

Moscow Pre-launch (2)

ISS Flight

Moscow Postflight(1)

Moscow Postflight(2)

SDL Postflight

Temp Trend(1)

Temp Trend(2)

SDL Post ReCal

MoscowPre-launch

ISS Flight

Moscow Post-flight

SDL Post-flight

8

No continual drift trends observed over 4 years of melt data, indicating PCM contamination not a factorAll data sets are within 6 mK. Individual point within a given set vary by < ±1.5 mK

Moscow Delivery

SDL Pre-delivery

Summary of All MOTR Data Sets

29.771

29.772

29.773

29.774

29.775

29.776

29.777

29.778

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Mel

t Tem

pera

ture

(°C

)

Melt Sequence Number

All Data Sets

SDL Predelivery(1)

SDL Predelivery (2)

Moscow Delivery

Moscow Pre-launch (1)

Moscow Pre-launch (2)

ISS Flight

Moscow Postflight(1)

Moscow Postflight(2)

SDL Postflight

Temp Trend(1)

Temp Trend(2)

SDL Post ReCal 1

9

ISS Data: Average Melt Temperature = 29.7731 °C Standard Deviation = 0.0008 ° CALL Data: Average Melt Temperature = 29.7737 ° C Standard Deviation = 0.0017 ° C Within ~0.6 mK

Orbital PCM Reference Value Criteria• Stable, reliable PCM containment• Miniaturized assembly• Includes multiple PCMs to improve calibration knowledge • Capable of reliably and repeatably freezing and melting PCMs• Capable of accurately measuring temperature during melting to

within 10 mK absolute error• Capable of tracking blackbody temperature to within a few mK

10

Transfer of Calibration

PCMSensor

TEC

Transfer:When the TEC is not

powered it acts as a thermal link to the thermal surface.

If adequately insulated it will come to equilibrium with the thermal surface.

The PCM sensor can be compared to thermal surface sensors’ readings.

Thermal Surface(What you really want to measure)

Calibration:During a

recalibration the TEC is powered and the PCM is controlled to a different temperature than the thermal surface to melt the PCM.

Temperature data collected during the melt allows recalibration of the PCM sensor.

Orbital 3 PCM Cell Design Phase-change cell with three cavities to be used for on-orbit temperature calibration

12

Thermistor

Thermistor locator (Teflon)

Cell body (SS 304)

Phase-change material ~0.3 mL

Threaded primary cap (SS 304)

Secondary cap (SS 304)

Redundant welds contain PCM

Ga 29.7646°CGaSn 20.46°CGaIn 15.63°C

Tem

pera

ture

, °C

Time, hours

Bath

Cells

PCM Freezing

PCM Melting

Bi-phase Equilibrium Temperature

Gallium-Tin Calibration Bath Data

Bath Bi-phase Equilibria Temperatures

• Gallium bi-phase equilibria were within <1 mK of the ITS90 Ga fixed point at 29.7646• Most eutectic melt temperatures are reported to >10 mK absolute accuracies

13

PCM AvgerageBi-Phase EquilibriumTemp (°C)

Bi-phase Equilibrium Temperature for 3 Individual Cells Tested (°C)

Ga 29.7647 29.764, 29.765, 29.765

GaSn 20.4610 20.460, 20.461, 20.462

GaIn 15.6323 15.631, 15.631, 15.635

SDL Vacuum Chamber Test Setup

Vacuum Chamber

Copper Block PCM Cell

Surface Heater

Circulated Liquid Cooled Base

Block Isolator

Outer Block Thermistor

Outer Rad Shield

Inner Block Thermister

TEC

Inner Rad Shield

Cell Mount Plate

14

• Unpowered PCM Cells equilibrated to within 2.5 mK of Inner Block Thermistors at 24 °C and at 1 °C.• Agrees with thermal model delta of ~2 mK equilibrium

Thermal Model of a Ga Melt Curve

15

Condition: Blackbody at 273.15 KTemp sensor between orange and red nodes

Thermistor will read somewhere between

these cell nodesPiece of the melting gallium

Gallium Gradients vs. Melt Time

16

Gradients in cell and duration of melt vary with heating power applied.

Mid-melt offsets in measurements agree with thermal model (~45-75 mK)

Ga Melt Point Repeatability• TEC heating with

constant current• Copper block held

at constant 8.5 ±0.01 °C

• Melt durations very consistent

• Plateau mid-point repeatability ~2 mK

17

GaIn Melt Point Repeatability• Melt durations

varied by ~13%• Inconsistency

possibly due to small constant current variations in the TEC

• Mid-points group to ~5 mK

• Clear correlation between mid-point temperatures and melt durations

18

GaSn Melt Point Repeatability• Melt durations varied by

~20%• Inconsistency possibly due

to small constant TEC variations

• Mid-points group to ~5 mK• Clear correlation between

mid-point temperatures and melt durations

19

Conclusions• Orbital testing of Ga phase transitions showed no change in melt

point under 0G conditions and verified long-term repeatability of melt curves

• SDL improvements of design include smaller cells and addition of two new melt temperatures for GaSn and GaIn eutectics

• SDL ground testing of cells with Ga and eutectics demonstrates 2-3 mK repeatability of three melt temperatures

• Ground testing and thermal modeling improve understanding of:– thermal gradients within cells and their effect on interpretation of melt

curves– true bi-phase equilibrium temperatures of melt materials

Questions?

20

References1. Committee on Earth Science and Applications

from Space: A community Assessment and Strategy (2007). Earth Science and Applications from Space: National Imperatives for the Next Decade and Beyond, pp. 93-94, ISBN 0-309-66714-3.

2. Fluke Corporation,Hart Scientific Division, “5901 Series Triple Point of Water Cells” 2005, [Revised 10/2006 2527367 D-EN-N Rev B].

3. Mangum B. W. & Furukawa G. T. (Eds.). Guidelines for Realizing the International Temperature Scale of 1990 (ITS-90), NIST Technical Note 1265, August 1990.

4. Topham T.S., Bingham G.E., Latvakoski H., etal., Observational study: microgravity testing of a phase-change reference on the International Space Station, npj Microgravity 1, Article number 15009 (2015), doi:10.1038/npjmgrav.2015.9

5. GE Sensing, “Thermometric Ultrastable Probe Thermistors,” SP series Datasheet, [copyright 2006].

6. Heraeus, “Platinum Resistance Temperature Detector,” M 222 Datasheet, document name 30910015 Index B, [June 2010].

21

Additional Slides

Sealed Cells vs. Pressure Dependence of Fixed-Point• For contamination issues PCM containers must be sealed• 1 atm pressure changes melt temperature of water by 10 mK [3]• Container must allow PCM expansion without changing fixed-

point temperature

Flexible container:

-No internal voids-PCM can expand container-PCM vacuum filled-complex filling-complex container-moving parts

Rigid container:

-PCM filled at 1 atm-Internal gas voids compress as PCM expands-location of voids in space?

SDL Temperature Sensor Testing• Heraeus PRT and GE

thermistor excellent size and long term stability [5,6]

• GE Thermistors tracked standards PRT ±3mK, with calibration improvement to ~1mK

• Heraeus PRTs tracked ±10-15mK (worse than larger wire PRTs)

• Heraeus shock resistance 40g at 10-2kHz

Bath Cycling of 4 SP60 Thermistors

(mK)

Drift 0.04% 1000hrs 2.1 x 2.3 mm

Heraeus M222 PRT 1.5 mm

3.2 mmGE SP60 Thermistor

Drift 0.02% /yr

Ga Melts of Various Lengths

1-2 hour melt curves are ~50-60 mK higher than melt point at center of melt due to cell thermal gradientsAgrees with thermal models which predict ~50-80mK. Longer melts within 5 mK of true melt temperature

25

GaSn Melts of Various Lengths

26

1-2 hour melts are only ~30-40 mK above eutectic melt point

GaIn Melts of Various Lengths

27

1 hour melts are 50 mK above eutectic melt point>5 hour melt was within 5 mK of eutectic melt point when it was stopped

Supercooling Behaviors

Ga and eutectics cool below their melt points after reaching temperatures well above as liquids.

28

GaIn Behavior Improvements• Tested in bath at

slower ramp rates

• Ga rich mixture shows slightly less supercooling

29

Acknowledgements• Russian Academy of Sciences IBMP for launch and flight

support• VNIIOFI (Andrey Burdakin) for Gallium eutectic

investigations and preparation training• NASA Earth Science and Technology Office (ESTO) for

funding support• NASA Langley’s CLARREO team for CORSAIR work

Questions?

30