Neutron Imaging at NIST: An in situ method for visualizing and quantifying water dynamics in low temperature PEM fuel cells. David Jacobson Daniel Hussey (NIST) Muhammad Arif (NIST) Jon Owejan (RIT) Satish Kandlikar (RIT) Thomas Trabold (GM – FCA) Fuel Cells Neutron Imaging National Institute of Standards and Technology Technology Administration U.S. Department of Commerce
32
Embed
Neutron Imaging at NIST: An in situ method for visualizing and quantifying water dynamics in low temperature PEM fuel cells. David Jacobson Daniel Hussey.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Neutron Imaging at NIST: An in situ method for visualizing and quantifying water dynamics in low temperature PEM fuel cells.
David JacobsonDaniel Hussey (NIST)
Muhammad Arif (NIST)
Jon Owejan (RIT)Satish Kandlikar (RIT)
Thomas Trabold (GM – FCA)
Fuel CellsNeutron Imaging
National Institute of Standards and Technology Technology AdministrationU.S. Department of Commerce
Support
• DOE – Energy Efficiency and Renewable Energy– Interagency agreement # DE\_AI0101EE50660– Nancy Garland Program Coordinator
• NIST Intramural Advanced Technology Program– Gerald Caesar
• NIST Physics Laboratory (www.physics.nist.gov)• NIST Center for Neutron Research (www.ncnr.nist.gov)
– Patrick Gallagher (director), and many others who provide tremendous technical assitance.
Some Neutron Radiography Facilities
• Paul Scherrer Institute - NEUTRA
• Pennsylvania State University - Breazeale Nuclear Reactor Facility
• Institute Laue Langevin (Grenoble, France)
• FRM-II (Munich, Germany)
• JRR-3M (JAERI) (Japan)
• HANARO (KAERI) (Taejon, Korea)
• Many other smaller reactors
1. New facility 14.6 m2 (157 ft2) floor space2. Accessible 2 meters to 6 meters3. Variable L/d ratio
1. At 2 m L/d = 100 → ∞2. At 6 m L/d = 300 → ∞
4. Maximum Intensity without filters1. At 2 m = 1 x 109 n cm-2 sec-1 (L/d =100), 8 cm diameter beam size2. At 6 m = 1 x 108 n cm-2 sec-1 (L/d =300), 25 cm diameter beam size
5. Maximum Intensity with 15 cm LN cooled Bismuth Filter1. At 2 m = 2 x 108 n cm-2 sec-1 (L/d =100), 8 cm diameter beam size2. At 6 m = 2 x 107 n cm-2 sec-1 (L/d =300), 25 cm diameter beam size
6. Support for fuel cell experiments1. Hydrogen flow rates 18.8 lpm2. 50 cm2 fuel cell controller with 5 lpm flow rates.3. Nitrogen, Air, Coolant and Hydrogen Venting
7. Detection capabilities1. Real-Time Varian Paxscan, 30 fps @ 0.254 mm pitch or 7.5 fps @ 0.127 mm pitch2. Second Varian detector will upgrade to 30 fps @ 0.127 mm pitch3. 2048 x 2048 Cooled (50° C) Andor CCD based box with 30 cm maximum field of view.4. 2 more 1024 x 1024 Cooled (30° C) Apogee CCD based
8. Sample Manipulation1. Motor controlled2. 5 axis tomography capability
9. Open for business January 2006
Neutron Imaging Facility (NIF)
Beam Stop
Cable Ports
Drum shutter and collimator
6 meter flight path
LN Cooled Bismuth Filter
2.13 mCable Ports
Steel pellet and wax filled shield walls
Neutron sensitive screen
Point Source
Fuel cell
Neutrons are an excellent probe for hydrogen in metal since metals can have a much smaller cross section to thermal neutrons than hydrogen does.
Comparison of the relative size of the x-ray and thermal neutron scattering cross section for various elements.
x-ray cross sectionH D C O Al Si Fe
neutron cross section
0I tNeII 0
Sample
t
N – numerical density of sample atoms per cm3
I0 - incident neutrons per second per cm2
- neutron cross section in ~ 10-
24 cm2
t - sample thickness
Why Neutrons
Water Sensitivity
=
Wet cuvet Dry cuvet water only
=-ln
• Steps machined with 50 micron.• CCD camera exposure of 1 s yields a
sensitivity of 0.005 g cm-2 s-1
• After 100 s a factor of 10 improvement gives 0.0005 g cm-2 s-1
• New amorphous silicon detector should have a least a factor of 7 improvement in temporal sensitivity
1 s exposure time
50 micron water thickness
Sensitivity required for fuel cells (assumes maximum water content)
• Flow fields 0.020 g cm-2
• Gas diffusion media 0.012 g cm-2
• Electrode 0.0005 g cm-2
• Membrane 0.0005 g cm-2
Neutron scintillator• Converts neutrons to light 6LiF/ZnS:Cu,Al,Au
6Li + n0 4He + 3H + 4.8 MeV
• Light is emitted in the green part of the spectrum
• Neutron absorption cross section for 6Li is huge (940 barns)
CCD
Scintillator
Neutrons inGreen light out
• Neutron to light conversion efficiency is 20%
Real-Time Detector Technology
• Amorphous silicon • Radiation hard• High frame rate (30 fps)• 127 micron spatial resolution• Picture is of water with He
bubbling through it• No optics – scintillator directly
couples to the sensor to optimize light input efficiency
Neutron beam
scintillator
aSi sensor
Side view
Readout electronics
Scintillator aSi sensor
Front view
Helium through water at 30 fps
New technology
• Currently the spatial resolution is of order 100 microns• Not a fundamental limitation, but is due to light blooming
out in the ZnS, which is 0.1 mm – 0.3 mm thick• Currently have tested detectors with 30 micron resolution
(potentially 15 microns).• Major innovation in detection technology• Resolution has been measured to be 30 microns• Final testing and development expected to be completed
• Rectangular channels– Water flow is laminar tending to constrict and plug the channels– Water plugs form as large slugs and can be difficult to remove.
• Triangular channels– Water stays at the corner interface with the diffusion media leaving the
apex of the channel more clear.– Water tends to come out in smaller droplets instead of large slugs,
which require a high pressure differential to remove
Flow Field Properties
Rectangular X-sect
Triangular X-sect
Gold Coated w/PTFEContact Angle = 93°
Graphite Uncoated Gold Coated Gold w/PTFE
0.01170 ohm/cm2 0.00044 ohm/cm2 0.00052 ohm/cm2
Gold Uncoated Contact Angle = 50°
Contact Resistance Values
94°
1.37 mm 1.45 mm
0.7
6 m
m
1.37 mm 1.45 mm
0.3
8 m
m
Xsect Area = 0.52 mm2
Cathode Channel Cross Section Geometry and Surface Energy Study
• For all cells tested, water accumulation in the channels decreased with load, while accumulation in the diffusion media/MEA increased with load.
• There was a significant difference in channel water retention for Toray and SGL materials due to material surface energy.
Key Observations and Conclusions (cont’)• Lower cell performance at 1.0 A/cm2 using Toray is
associated with only 0.05 g more water accumulation in the channels and non-channel regions.
• Channel surface energy has a consistent effect on water slug shape and size. Higher contact angle increases average water mass retained, but distribution of smaller slugs more evenly in the channel area increases performance.
• Triangular cross-sectional geometry accumulates water in the corners adjacent to diffusion media. The center of the channel does not become obstructed by stagnant slugs.