View
7
Download
0
Category
Preview:
Citation preview
Materials Science & Technology
Application of advanced and non destructive testing in
solid oxide fuel cells
Peter Wyss, Erwin HackLaboratory for Electronics/Metrology/Reliability
Artur
Braun, Lorenz Holzer, Christian Soltmann, Peter Holtappels
Laboratory for High Performance Ceramics
Empa, Swiss Federal Laboratories for
Materials Testing
and ResearchÜberlandstr
129 , CH-8600 Switzerland
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 2
Innenringdichtung
Radialdichtung
Innenringdichtung
Radialdichtung
Outline
Test items
and techniques
Non-destructive
testing
neutron
tomography
x-ray
radiography
thermography
Advanced
destructive
testing
FIB tomography
Outlook
metallic components
ceramic
components
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 3
SOFC features&scaling
Electrocatalytic
activityNano/atomic scale
Ionic
conductivityNano/atomic scale
Electronic conductivity„Nano/atomic/micro scale“
(Open) porosityMicroscale
SEM/OM/EDX
Only 2-D imaging
Solid Oxide Fuel
Cell
-
Principles
-
Fuel: CH4
/H2
O CO2
/H2
O
Air: O2
/N2
anode support
Metallic IC Cr-, Ni-Cr-Steels
N2
Ni/YSZ
e-
e-
anode Ni/YSZ
electrolyte OOx YSZ
cathode LSM, LSCF, LSF
Metallic IC Cr-, Ni-Cr-Steels
3-D structures
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 4
Non-destructive
Testing
(NdT)
component state potential problem ndt
test method
metallic machined
part cracks, bad welds X-ray
RT, CT,
UT for
interconn.
metallic assembled
in stack corrosion, contamination
RT, CT X-rays,Neutrons ?
ceramic green
machined
part
porosity
+ homogenity, shrinkage
cracksRT, µCT local
mode,
X-rays, TT
ceramic assembled
and fired
to cellcracks
+ delaminations
(thermal cycling)
RT, µCT local
mode,
X-rays, TT
ceramiccells
mountedin the
stackfatigue
cracks
(thermal cycling)
RT, CT, Neutrons + contrast
fluid
?
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 5
RT system data
(typical)
1. NdT
methods
and test items
source beam
geometryresolution
in µmfield
of view(FOV) in mm
typ. penetrableZrO2 in
mm
Neutrons almost
parallel 100 300 > 200
X-rayMini / Micro
focus
conical 10 400 20
X-ray
Synchrotron parallel 1 10 2
X-ray direction
Field
of view
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 6
Neutron tomography
~ 20 cm
permits
„insight“
into
submillimeter
porosity
of SOFC stack
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 7
2. RT and local
µCT to Empa
YSZ pellets
The
RT / CT microscopy
(macroscopy) system
X-ray
flat
panel Hamamatsu
7942 CA-02
VISCOM TEP 9225 panorama tube head or
TXD 9160 subµ
tube
head
Sample on low
runout
rotation stage
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 8
The
crack
detection
limit
in radiography
α cw
crack
detection
is
possible
if:
c > 0.01 w * sin
αtube spot
size
* 0.5
X-ray
direction
RT and local
µCT to Empa
YSZ pellets
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 9
Visible
light microscopyRadiographies
of a hidden
coarse
grain
RT and local
µCT to Empa
YSZ pellets
FOV 5 x 5 mm, pixelsize
2.5 µm
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 10
Visible
light microscopy
Radiographies
RT and local
µCT to Empa
YSZ pellets
of a shrinkage
crack FOV 5 x 5 mm, pixelsize
2.5 µm
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 11
Local
X-ray
µCT of crack
in pellet Ø
35 x 1 mm
Slicesparallelto main
surface
Field
of view5 x 5 mm,
voxelsize
5 µm
RT and local
µCT to Empa
YSZ pellets
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 12
Local
X-ray
µCT of hidden
coarse
grain
Slicesparallelto main
surface
Field
of view5 x 5 mm,voxelsize
5 µm
RT and local
µCT to Empa
YSZ pellets
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 13
The
thermography
cam, heart
of the
TT system
TT, RT and local
µCT to HTceramics
cellsCamera Cedip
JADE
Camera type Array
Resolution 240 x 320 pix
Wavelength range 3-5 μm
Frame rate Full: 170 HzROI: 9 kHz
Lateral resolution 15 x 15 μm2
NETD 20 mK
Temp. range -20 –
1300 °C
Lock-In frequency < 5 kHz
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 14
TT, RT and local
µCT to HTceramics
cells Thermography
testing (TT), impulse method
bad bond
spot
flash
hits
surface
diffusion
wave propagates
bad bond
spot
stops the
heat
diffusion
wave
and after
some
time a thermic
contrast
appears: positive at the
impulse
side,
negative at the
rear side
time ≈ depth
2
lateral resolution
≈
2 x depth
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 15
Impulse thermography
images
of a region containing
a spot
of high thermal conductivity
TT, RT and local
µCT to HTceramics
cells
Istantaneous
after
flash 20 msec
later
FOV 180 x 150 pixel
or
28 x 25 mm, Pixelsize
167 µm
FOV for
RT +µCT
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 16
The same region (FOV 5 x 5 mm) imaged with:
Radiographies, pixel
size
2.5 µm Local
tomography, voxel
size
5 µm
TT, RT and local
µCT to HTceramics
cells
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 17
Focussed
Ion Beam
(FIB) technique
DLR BekNi
275/3
Advancedpreparation procedure
Conventionalpreparation
procedure
Hi2
2H
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 18
TEM: Imaging
& elemental
analysis
chemical
map:
Ni
Zr
DLR BekNi
275/3
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 19
DLR BekNi
275/3
Electrolytestabilised
zirconia
Fuel
electrodeNi-Cermet
Air electrodeperovskite
volume:40 x 40 x 40 m3
FIB-Nanotomography: 3-D structure
of a FC
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 20
Nanotomography: Informationsgewinn
3D ImagingParticulate
und micro
structure
6 µm
A
6 µm
B
DistinctionCrystallite-Particulate
aggregates
Use
of Information:ModelingUnderstanding
degradation
e.g. sulphur
poisoning
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 21
access
tolarge scale
facilities
Advanced
characterisation
cm / mm micro
meter nano
meter
Source: J.R. Wilson et al.,Nature Materials 5 541-544, 2006
software
sample
preparation
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 22
NdT
methods
and diagnostics
strategy
Visual testing
(VT), including visible
light microscopy
will be
done
in any
case
Ultrasonic
testing
(UT)
requires
flat
surfaces
and low
damping, 3D possibility
Eddy current
testing
(ET)
for
electrical
conductors
only
Magnetic
testing
(MT)
for
ferritic
materials
only
Thermography
(TT)
best for
close
to surface
items, 3D possibility
Radiographic
testing
(RT, CT, XTM)
No requirements
to surfaces
and damping, 3D possibility
Quality
assurance
cell
production
stacking
Life time / Durability
comparison
of pre
and post test state
by
NdT
possible
failures
affecting
mechanical
properties
visible
by
NdT
Combination
of NdT
advantageous
identification
of points
for
destructive
analysis
3D imaging
real structures validation
of 2D analysis
(e.g. SEM, OM)
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 23
Acknowledgement
Defne
Bayraktar, EMPA
Ulrich Vogt, EMPA
Thomas Graule, EMPA
Günter Schiller, DLR
SINQ, PSI
Josef Sfeir, Hexis
HTceramix
Hi2
2H
Empa, Peter Holtappels, FC Tools, Trondheim, 23.6.2009 24
4. A possible
NdT
procedure
for
cells
Impulse thermography
overview
(FOV = 80 x 80 mm) Pixel size
0.33 mm, meas. time per cell
≈
2 min
Impulse thermography
close
up (FOV = 8 x 8 mm) Pixel size
33 µm , meas. time per cell
≈
10 min (automated)
Radiography
overview
(FOV = 80 x 80 mm) Pixel size
20 µm , meas. time per cell
≈
5 min
Radiography
close
up (FOV = 8 x 8 mm) Pixel size
4 µm (OVHM -
Movie) , meas. time per item
≈
10 min (auto)
Local
tomography
(FOV = 8 x 8 mm) , meas. time per item
≈
60 min voxel
size
4 µm or
calculation
of items
depth
by
evaluating
the
trajectories
from
OVHM -Movie
Recommended