Transcript 21.HundCapsuleFabrication

IFE Target Fabrication Update
Presented by Jared Hund1
N. Alexander1, J. Bousquet1, R. Cook1, D. Frey1, D. Goodin1,
J. Karnes2, R. Luo1, R. Paguio1, R. Petzoldt1, N. Petta2, N. Ravelo1,
K. Saito1, D. Schroen1, J. Streit2, A. Cheng3, W. Holloway3,
N. Robertson3, S. Saiedi3, M. Weber3
1General
Atomics, Inertial Fusion Technology, San Diego, CA
2Schafer Corporation, Livermore, CA
3UC San Diego, San Diego, CA
HAPL Workshop
Naval Research Laboratory
Oct 29-30, 2007
IFT\P2007-095
Since the last HAPL Meeting we have:
• Successfully coated HAPL dimension foam shells and have shown
them to be gas tight
• Demonstrated resorcinol formaldehyde foam shells as an option
for a HAPL target
–
–
–
–
–
Made them as a HAPL sized capsule
Improved the wall uniformity
Successfully overcoated with GDP
Shown them to be gas tight
Approaching surface finish specifications
• We are beginning to characterize the capsule beyond the basics
– The foam density uniformity across the shell wall has been
measured for DVB and RF
• Begun a collaboration with U of Rochester and UCLA to investigate
additional control of shell fabrication (Dielectrophoresis)
The current high-gain HAPL target design is a
4.6 mm diameter foam capsule
Thin (300-1200 Å)
High Z coating
10 m CH Overcoat
Important Specifications
• Foam shell:
– Out of round
– Wall uniformity
Foam + DT
DT
DT Vapor
• Plastic Overcoating:
– Gas tight
– Smooth (50 nm RMS)
Foam layer:
• High Z coating
0.18 mm thickness Divinyl Benzene
(DVB) or Resorcinol-Formaldehyde (RF)
The specifications of the NRL/HAPL target have
dramatically evolved
• Foam capsule
20 mg/cc DVB → 100 mg/cc DVB → 100 mg/cc DVB or R/F
• Coatings
1 μm overcoating → 5 μm → 10 μm today
The specifications of the NRL/HAPL target have
dramatically evolved
• Foam capsule
20 mg/cc DVB → 100 mg/cc DVB → 100 mg/cc DVB or R/F
very fragile foam → more robust foam → smaller pore size R/F foam
diameter increased slightly in the evolution of physics design
Elemental composition – DVB (CH) initially preferred because does not contain
oxygen like RF (CHO)
• Coatings
1μm overcoating → 5 μm → 10 μm today
The specifications of the NRL/HAPL target have
dramatically evolved
• Foam capsule
20 mg/cc DVB → 100 mg/cc DVB → 100 mg/cc DVB or R/F
very fragile foam → more robust foam → smaller pore size R/F foam
diameter has increased slightly in the evolution of the physics design
Elemental composition – DVB (CH) initially preferred because does not contain
oxygen like RF (CHO)
• Coatings
1μm overcoating → 5 μm → 10 μm today
increased thickness in order to survive fabrication and filling
initial design depended on primary strength from the foam layer
Do we meet foam capsule requirements?
Attribute
Value
Tolerance
O, N content
0
?
Diameter
4.6 mm
 0.2
Wall
thickness
180 m
 20
Density
20-120
mg/cc
[25%]
Pore size
<3 m
Out of round
<1 % of
radius
--
Nonconcentricity
< 1-3%
of wall
th.
--
Areal density
< 0.3%
Modes 100
to 500
DVB
Pass
(0 at%)
Pass
RF
Comments
Pass
(12 at%)
Pass
0.025 mm
range
± 0.06 mm
range
Pass
In progress
Pass
Pass
(± 20 um)
(97± 5
mg/cc)
(~1 um)
Pass
(~0.01 um)
Based on SEM for DVB. RF
measured with SEM and N2
adsorption
Pass
Borderline
(1%
average)
Measured by single view on dry
foam shells.
Pass
In Progress
(10% of
shells <3%
NC
(<0.3 %)
(~60% of
shells
<3% NC)
Pass
Contact radiography to determine
radial density variation.
How well are we meeting the overcoat specs…?
Attribute
Value
Coating
composition
CHNO
Coating
Thickness
10 m
Power
spectrum
(surface
finish)
Tolerance
DVB
RF
PVP/G
DP
(CHO)
GDP (CH)
+/2um
+/- 2um
<50 nm --
> 500
nm
Getting
close (50
- 200 nm)
Permeability
(gas tight)
and yield
TBD
--
Fail at
10 um
Fail at
10um,
some
good at
> 20 um
For current techniques require
~20 um thickness, working to
minimize
Strength (for
filling)
TBD
--
> 2 atm
For a >20 um thick coating
+/- (30 –
300) nm
Comments
GDP process - Δwall set by
areal density specification
These specs are evolving as more
simulations are done by the designers
The DVB capsule meets the sphericity specification, but
RF still requires work
• The yield of RF shells that meet the 1% of radius Outof-Round (OOR) specification is 70%
DVB Shell Data
OOR (microns)
100% Yield
60
OOR (microns)
RF Shell Data
40
20
70% Yield
60.0
40.0
20.0
0.0
0
0
5
10
Shell #
15
20
0
5
10
15
20
Shell #
max radius
OOR = (max radius – min radius)
A possible fix for this is to increase the interfacial
tension of the RF system before curing
min radius
Currently DVB shells have a better yield of shells that
meet the wall uniformity specification
• Uniformity defined in terms Nonconcentricity (NC)
NC 
offset
avg. shell thickness
Offset = distance between centers
of inner and outer wall
Percentile Plot of Foam Wall Uniformity
100%
75%
Percentile
Current HAPL DVB:
60-70% meet <3% NC
1-3 %
50%
Current HAPL RF
10-15% meet <3% NC
25%
Early HAPL DVB
0%
0
1
2
NC Spec
3
4
5
6
7
8
9
10
% NC
DVB is better for the NC specification (at the moment)
Density Uniformity of shells
• Procedure
– The shell is placed in contact with
glass supported film and exposed to
x-rays
– Film is developed and digitized
– Image analyzed by “unwrapping”
shell and comparing attenuation to
known standard
• Capabilities
Radiograph of
sector of Coated
Foam shell
GDP
RF
Interface
positions
on a poor
NC shell
Radial distance
(um)
– +/- 10% Absolute error, +/- < 1%
relative error within shell
Interior of
shell
1200
1150
Grayscale
Grayscalebetween
betweeninterfaces
interfaces
provides
providesdensity
densityinformation
information
1100
1050
1000
0
90
180
Angle (degrees)
270
360
Many of the DVB shells tested have a density variation
along the thickness of the foam wall
0.12
Density (g/cc)
0.11
0.1
0.09
0.08
0.07
0.06
1500
1600
1700
1800
1900
Radial Distance (microns)
2000
2100
RF shells also have a density variation along the
thickness of the foam wall
0.12
Density (g/cc)
0.11
0.1
0.09
0.08
0.07
0.06
1500
1600
1700
1800
1900
2000
Radial Distance (microns)
Is this a problem for the target physics?
2100
The next steps in characterization of areal density
• 2D Areal density:
– Contact radiography
• Measure density vs. angle
• Wallmapping is being used to get the
overcoating thickness around the shell
– Interferometric technique developed for ICF
• 3D Areal density:
– 3D X-ray Tomography
– Precision radiography
• Directly measure x-ray attenuation of shell
Overcoating
• Currently evaluating GDP coatings
• Could be complemented by RF skin
technique
• Other alternative described in previous HAPL
workshops
– Interfacial polymerization
– Interfacial polymerization/GDP
Measured buckle strength of GDP coated foam shells
Buckle strength of GDP coated RF shells
Buckle strength (psi)
40
35
30
25
y = 0.024x2 + 0.9766x
20
15
10
5
0
0
5
10
15
20
25
30
GDP Coating Thickness (microns)
Tritium inventory implication: ~1 kg for 100° C fill of a capsule with
25 um GDP coating (dominated by layering time)
High temperature fill needs to be tested
A number of GDP/RF shells have been shown to be leak
tight
• Shells filled with D2 and the leak rate measured with mass
spectrometer
• Leak rate measured at room and liquid nitrogen temperatures
–
To distinguish between pinhole and permeation flow
• Test developed at GA for LLE cryo work and validated based on
success of cryo foam targets at LLE
Room
temp
ln (MS ion s ig nal)
-12
Room
temp
LN2
temp
-14
-16
Specs:
-18
4 mm diameter
25 um of GDP on
RF
baseline
-20
3000
2500
2000
1500
1000
500
0
-22
Time (s ec onds )
The shells are tested to be “gas tight” and
can survive cryo cooling and warming cycle
Gas testing results: 4-4.6 mm diameter GDP/RF (HAPL
size foam shell)
Leak Mechanisms:
Good – permeation leak only
LN2 temp
ln (MS ion s ig nal)
-12
1200
Pinhole free
-16
-18
-20
3000
2500
2000
1500
1000
500
0
-22
Time (s ec onds )
800
Bad – pinhole leak
600
Pinholes
-12
ln (MS ion s ig nal)
400
200
-14
-16
LN2 temp
-18
-20
1800
1600
1400
1200
1000
800
600
200
400
-22
0
Time (s ec onds )
0
Ugly – viscous flow leak
Leak too fast to
test at cryo temp
-12
-14
-16
-18
-20
Time (s ec onds )
450
400
350
300
250
150
-22
100
GDP coating thickness (um)
30
50
20
0
Predicted D2
permeation rate
10
ln (MS ion s ig nal)
0
200
D 2 Half life (sec)
1000
-14
Coated R/F foam shells are smoother than over-coated
DVB shells
Surface Roughness of HAPL Coated Foam Shells
(> 4 mm diameter)
Shell/Coating
Type:
DVB/PVP
DVB/PVP/
GDP
RF/GDP
Optical Profiler (WYKO) measurements acquired
at 20x, with a 300 x 200 um area
Spheremapping of GDP coated, 3 mm diameter foam
shell with 10 um GDP wall
• Spheremap – AFM technique for measuring roughness of
spherical capsules
AFM Traces Around Shell
• OOR of shells must be better for
routine spheremapping
Power Spectrum of Surface Roughness
1E+6
Power (nm^2)
1E+5
1E+4
1E+3
1E+2
1E+1
1E+0
1E-1
1
10
100
1000
Mode Number
Meets roughness specification:
42 nm RMS for modes 50-1000
Changing the background pressure during the GDP
coating run affects the final surface
Background pressure:
500 mtorr
Standard condition
250 mtorr
75 mtorr
Dome growth
This may be a
technique for
improving surface finish
of a GDP overcoating
“Coral” DLA
structure
Smooth surface
We have begun a collaboration with U of Rochester and
UCLA to investigate additional control of shell fabrication
• We can manipulate the shells using electric
fields through dielectrophoresis (DEP)1,2
• The electric field also has a centering effect on
the inner droplet
• Currently work is being done for
DEP
Levitation
OMEGA direct drive – HAPL size
Concept
capsules will be done in parallel
cap electrode
• Possibility for HAPL scale up
cusped E field
ring electrode
1.
H. Pohl, Dielectrophoresis : the behavior of neutral matter in
nonuniform electric fields, 1978
2.
T. Jones, Electromechanics of particles, 2005
capsule
Conclusions
• HAPL specifications have evolved into
1. DVB foam option:
– Good NC (yield: 60% < 3% NC)
– DVB with PVP/GDP – still have pinholes at >10
um
– Poor roughness of coated DVB shells
2. RF option:
– NC is worse, but improving (yield: 10% < 3%
NC)
– RF gas tight shells at 20 um (no pinholes)
– The surface roughness is most promising