Update 2015 on target fabrication requirements for NIF

Update 2015 on target fabrication requirements
for NIF layered implosions, with emphasis on
capsule support and oxygen modulations in GDP
S. W. Haan,1 D.S. Clark,1 S.H. Baxamusa,1 J. Biener,1 L. Berzak Hopkins,1 T. Bunn,1 D.A.
Callahan,1 L. Carlson,3 T.R. Dittrich,1 M.J. Edwards,1 B.A. Hammel,1 A. Hamza,1 D. E. Hinkel,1
D.D. Ho,1 D. Hoover,3 W. Hsing,1 H. Huang,3 O.A. Hurricane,1 O.S. Jones,1 A.L. Kritcher,1 O.L.
Landen,1 J.D. Lindl,1 M.M. Marinak,1 A.J. MacKinnon,1 N.B. Meezan,1 J. Milovich,1 A. Nikroo,1
R. Olson,2 L. Peterson,1 P. Patel,1 H.F. Robey,1 J.D. Salmonson,1 V. Smalyuk,1 B.K. Spears,1
M. Stadermann,1 S.V.Weber,1 J.L. Kline,2 D.C. Wilson,2 A.N. Simakov,2 and A. Yi2
1Lawrence Livermore National Laboratory, Livermore, CA 94550, USA
2Los Alamos National Laboratory
3General Atomics (with my apologies for not including in abstract)
21st Target Fabrication Meeting
Las Vegas, Nevada
June 23, 2015
Summary/Outline
We are making good progress toward ignition, and
target
fabrication
is a vital ingredient
For the
three shotsprogress
we have considered
in detail,
simulations show rough agreement with the data but
not a perfect match
This 3D simulation by Dan Clark of shot N120321 nicely shows the issues
keV
N120321
t = 22.85 ns
(bang time)
g/cm3
4.0
500.0
3.0
375.0
2.0
250.0
1.0
125.0
0.0
0.0
fuel-ablator
interface
fill tube
defect
tent
defect
LawrenceImploded
Livermore National
Laboratory
shape
at bang
time
Green surface = fuel-ablator interface
Temperature on left, density on right
We are incorporating our most upTent understanding
feature is 2-3x
bigger
to-date
of the
role than we
1-2 years
ago
ofrealized
the tent, hohlraum
drive
asymmetries, and other
perturbation
sources into 2-D
and
Global out-of-round
from
3-D
simulationsasymmetry
of NIF implosions
hohlraum
While we do not match all of the
Shorter
wavelength
irregularities
data
within the
error bars, there
is
are agreement
probably for
bigger
than
rough
both low
andshown,
high
foot shots,
and
heavily
because
of “clean”
oxygen
nonuniformity
mixed
shots
in GDP
The precise effects of the tent and
The program
is working
to
hohlraum
asymmetries
are
minimize
these drivers
features
and their
probably
the biggest
for the
remaining
growthdiscrepancies
and impactand
subjects for continuing work
Developing other options is also
important: High Density Carbon,
Be, other hohlraum designs
6
2
We believe that we have a fairly good understanding
of the issues in CH implosions
Dan Clark post-shot modeling for N120321 (characteristic of
2012 NIC “low-foot” shot) with various combinations of
degradation factors
When everything we know
1000
about is included (in 3D) we get
pretty close to the experiment
Neutron
100
yield /
For the three shots we have considered in
15
simulations show rough agreement with t
10
10
not a perfect match
keV
1
N120321
t = 22.85 ns
(bang time)
g/cm3
4.0
500.0
3.0
375.0
2.0
250.0
1.0
125.0
0.0
0.0
fuel-ablator
interface
fill tube
defect
tent
defect
We are inco
to-date unde
of the tent, h
asymmetries
perturbation
3-D simulati
While we do
data within t
rough agree
high foot sho
mixed shots
The precise
hohlraum as
probably the
remaining d
subjects for
Lawrence Livermore National Laboratory
D. S. Clark, M. M. Marinak, C. R. Weber, D. C. Eder, S. W. Haan, B. A. Hammel, D. E. Hinkel, O. S. Jones, J. L.Milovich, P. K. Patel, H. F. Robey, J.
D. Salmonson, S. M. Sepke, and C. A. Thomas, Radiation hydrodynamics modeling of the highest compression inertial confinement fusion ignition
experiment from the National Ignition Campaign, Physics of Plasmas 22, 022703 (2015)
3
The tent has the most impact on simulated
performance
Dan Clark post-shot modeling for N120321 (characteristic of
2012 NIC “low-foot” shot) with various combinations of
degradation factors
1000
Neutron
yield /
1015
100
For the three shots we have considered in
simulations show rough agreement with t
not a perfect match
10
keV
1
N120321
t = 22.85 ns
(bang time)
g/cm3
4.0
500.0
3.0
375.0
2.0
250.0
1.0
125.0
0.0
0.0
fuel-ablator
interface
fill tube
defect
tent
defect
We are inco
to-date unde
of the tent, h
asymmetries
perturbation
3-D simulati
While we do
data within t
rough agree
high foot sho
mixed shots
The precise
hohlraum as
probably the
remaining d
subjects for
Lawrence Livermore National Laboratory
Tent is the most important single issue. 3-4x bigger than
originally expected. In-flight backlit implosion
experiments confirm big perturbation.
S. R. Nagel, S. W. Haan, J. R. Rygg, C. Aracne-Ruddle, M. Barrios, L. R. Benedetti, D. K. Bradley, J. E. Field, B. A. Hammel, N. Izumi, O. S. Jones, S. F. Khan, T. Ma, A. E. Pak, K. Segraves, M.
Stadermann, R. J. Strauser , R. Tommasini, and R. P. J. Town , Effect of the mounting membrane on shape in inertial confinement fusion implosions , Physics of Plasmas 22, 039902 (2015)
4
in base units of centimetres, grams, and kiloelectronvolts. For N130927,
fa 5 0.68–0.82. The energy deposited in the hotspot by a-particles is
Ea 5 faEfusion/5, recalling that one-fifth of the D–T fusion energy is
emitted in the form of a-particles (the remaining a-particle energy is
deposited into the cold fuel). We note that, using the values found in
Table 1, Ea/Ehs < 0.56. These energies fully describe the hotspot, but part
of the implosion energy was used to compress the remaining cold D–T
fuel and so we must examine the fuel to get a full picture of the implosion
energy balance.
Because the D–T hotspot is formed by ablating the inner surface of
the cold D–T fuel as electron conduction transports heat from the
forming hotspot into the fuel, we can calculate the amount of D–T
fuel remaining after the hotspot has formed because we know the
bang-time) assuming that the cold fuel and the hotspot are isobaric
5=3
(Pfuel < Phs), in which case we find that a~Pfuel =PF <Phs =0:0021rfuel 5
2.9–3.3 for N130927—the fuel adiabat in flight is lower than this range of
values. The fuel density is also needed to calculate the X-ray losses through
the fuel.
As the hotspot is compressed to high temperatures, the primary
energy loss mechanism is bremsstrahlung X-ray emission because
the D–T hotspot is optically thin to these X-rays. The bremsstrahlung
energy loss is calculated to be24
pffiffiffiffiffi
Ebrems (kJ)~5:34|10{34 n2hs Te Vhs tx
The tent-seeded perturbation is considerably smaller
in less unstable implosions with different pulse shapes
Figure from Riccardo Tommasini’s paper analyzing the tent
feature
in base units of centimetres, kiloelectronvolts and seconds. For N130927,
as seen in backlit implosions (“2D ConA” experiments) E 5 2.3–4.5 kJ, the low end of which is nearly equivalent to the
15
Self-heating yield/
compression yield
20
1.5
Yield (kJ)
RESEARCH
RESEARCH LETTER
LETTER
RESEARCH LETTER
10
Yield from fuel compression
Yield from self-heating
Energy delivered to D–T fuel
N130710
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
GLC
30°
30°
44.5°
44.5° 30°
b
b
b
Inner
Inner cone
cone
Inner cone
23.5°
23.5°
23.5°
44.5°
50°
50° 44.5°
44.5°
50° 44.5°
50°
50°
50°
23.5°
23.5°
23.5°
30°
30°
Outer
Outer cone
cone
44.5°
30°
44.5°
Outer cone
44.5°
44.5°
50°
Low
foot
44.5° 50°
44.5° 50°
Shot
Shot
N130927
Shot
N130927
N130927
D–T ice
ice layer
layer
D–T
D–T ice layer
2.263 mm
mm
2.263
2.263 mm
CH ablator
ablator graded
graded
CH
2%
Si doped
doped
CH2%
ablator
graded
Si
110608
110615
110620
110826
110904
110908
110914
111103
111112
111215
120126
120131
120205
120213
120219
120311
120316
120321
120405
120412
120417
120422
120626
120716
120720
120802
120808
120920
130331
130501
130530
130710
130802
130812
130927
131119
0
0
0
N131119
N130927
N130812
N130501
a
a
a
5
High foot
Low foot
High foot
1.0
0.5
brems
2% Si doped
Cryogenic
Cryogenic cooling
cooling ring
ring
Cryogenic
cooling
ring
Shot
number
This is one reason (possibly the main
reason) why high-foot shots have done
so well, providing 10x more yield and a
lot of nice positive press
(eV)
Trad
TTrad
(eV)
rad(eV)
Figure 3 | Yield and energetics metrics for shots on the NIF. Total fusion
lines, 1s) as calculated from the model of ref. 25. The plot shows that, even with
yield is plotted versus shot number (that is, time). Shots 110608–130331 are
the uncertainty
in our results, shots 130927 and 131119 both yielded more
c
c
300
300 delivered to the
N130927
low-foot shots. Shots 130501–131119 are high-foot shots. The bars showing
fusion energycthan was
D–T. Inset, ratio of self-heating yield to
Au
N130927
Au hohlraum
hohlraum
300
3 shock
shock
total yield are broken into components of yield coming from a-particle selfcompression
yield
versus
generalizedN130927
Lawson criterion (GLC). All error bars,
3
Au hohlraum
3 shock
heating and yield coming from compression. The black dashes denote the
1s.
200
energy delivered to the D–T (fuel plus hotspot) with error bars (black vertical
200
200
Laser
Laser
entrance
Laser
entrance hole
hole
High foot
foot
High
All rights reserved
©2014 Macmillan Publishers
entranceLimited.
hole
100
100
High foot
100
3 4 6 | N AT U R E | VO L 5 0 6 | 2 0 F E B R U A RY 2 0 1 4
Laser
Laser quads
quads
Laser quads
N110914
N110914
4 shock
shock
N110914
4
4 shock
Low foot
foot
Low
0.0
Low foot
0.0
0.0
5.0
10.0
15.0
5.0
10.0
15.0
0.00.0
0.0
5.0
10.0
(ns) 15.0
tt (ns)
20.0
20.0
20.0
t (ns)
Figure 1 | Indirectly driven, inertially confined fusion target for NIF.
fuel layer and surrounding CH (carbon–hydrogen) plastic ablator. c, X
of
b,
X-ray
the
capsule
for
N130927
with
D–T
plastic
capsule with
bundles
incident
the inside
surface
of the
the hohlraum.
hohlraum.
b, representative
X-ray image
image of
oflaser
the actual
actual
capsule
foron
N130927
with
D–T
of the hohlraum. b, X-ray image of the actual capsule for N130927 with D–T
post-NIC high-foot implosion.
Figure 1in
| Indirectly
driven, inertially
confined
fusion target
for NIF. Nature
fuel layer
and surrounding CH (carbon–hydrogen) plastic ablator. c, X
O. A. Hurricane et al., Fuel gain exceeding unity
an
inertially
confined
fusion
implosion,
506,
a,
NIF
target
showing
aa cut-away
of
gold
drive
temperature CH
versus
time for
for the
the NIC
NIC low-foot
low-foot
implosion
Figure
1 | Indirectly
driven,
inertially
fusion
target
for NIF. and
fuel layer
and surrounding
(carbon–hydrogen)
plastic ablator.
c, X
a, Schematic
Schematic
NIF ignition
ignition
target
showingconfined
cut-away
of the
the
gold hohlraum
hohlraum
and radiation
radiation
drive
temperature
versus
time
implosion
plastic
capsule
with
representative
laser
incident
the
inside
high-foot
implosion.
a,
Schematic
NIF
ignition
target showing
a cut-away
of theon
and post-NIC
radiation
temperature
versus time for the NIC low-foot implosion
343–348 (20 February 2014)
plastic
capsule
with
representative
laser bundles
bundles
incident
ongold
thehohlraum
inside surface
surface
post-NICdrive
high-foot
implosion.
13–16
13–16
the hohlraum
. Although the hotspot shape changes that result
5
through wavelength changes alone, and physical changes to the
Final amplitude with adiabat shaping
Tent expectations as of May 2015: tents can work,
but will require progress on several fronts
Smallest possible ring for
respective thicknesses, from
target fab measurements
110 nm, 8-10°
contact angle
45 nm, 10-15°
15 nm, 15-20°
Max allowed assuming optimized
Adiabat-Shaping pulse shape
Max allowed with old low-foot pulse
15 nm, 3-5° contact angle
Radius of contact ring
45° ~
890µm
Depends on thickness, contact angle,
length of contact ring, and pulse shape
•  It is unlikely that >20nm tent thickness
would be OK
•  15nm or less could be OK but we need
to push on all fronts:
Ø  Reduce rms by ~30% with smaller
contact ring
Ø  Reduce contact angle reduces impact
by ~1/2
Ø  We will need an optimized AdiabatShaping pulse shape that reduces tent
growth by ~2x
Ø  Impact can be evaluated and optimized
with HGR experiments, exploring
pulse shape and tent parameters
6
A variety of other support options are being
explored
Capsule can be supported in several ways, each with pros and cons
We now have one fill tube, with impact that’s noticeable but smaller
than tent. Three or four 10 µm tubes (stalks) might be ~ equiv to
15nm tent. Impact grows with diameter, 30 µm diameter looks scary
Line support has considerable length that’s close to the capsule,
probably has to be sub-micron (hundreds of nm would be good)
Supported fill tube might support
capsule with just existing tube
touching the capsule
Foam
Tent, or foam to
hohlraum wall
Foam support looks good on paper, will
depend on uniformity of foam and other
details of density and composition
My best guess right now (and it is a guess) is that the best tent will be better
than foam support, and better than 30µm stalks. Two or three 10 micron stalks
would probably be better than best tent. Foam needs to be tested.
7
Tent is probably much less important for HDC and Be
Higher density of ablator (and lower opacity for Be) makes capsule
ablation pressure higher, pressure wave from exploding tent is relatively
less important
We have seen a tent feature in a couple of HDC symcaps (esp DT 4-step
symcap N130628) but in general it’s much smaller than in CH, in both
simulation and expt
Simulation work is ongoing, but we expect tent to be 2-3x less important,
even better for Be. Does not mean 100nm tent with any pulse shape is
OK! But finding a solution will be easier and tent will probably be OK.
Also, many near-term experiments with CH can use thicker tents (lowgrowth pulses like Near Vacuum Hohlraums, 2-shock pulses)
In the long run, NIF Ignition will require thin tents and accompanying
optimization, or some better support strategy
8
Hohlraum drive asymmetry has the 2nd largest impact
1000
Neutron
yield /
1015
100
For the three shots we have considered in de
simulations show rough agreement with the d
not a perfect match
10
1
keV
N120321
t = 22.85 ns
(bang time)
g/cm3
4.0
500.0
3.0
375.0
2.0
250.0
1.0
125.0
0.0
0.0
fuel-ablator
interface
fill tube
defect
tent
defect
We are incorpora
to-date understa
of the tent, hohlr
asymmetries, an
perturbation sou
3-D simulations o
While we do not
data within the e
rough agreemen
high foot shots, “
mixed shots
The precise effec
hohlraum asymm
probably the bigg
remaining discre
subjects for cont
Lawrence Livermore National Laboratory
Another important related issue is hohlraum efficiency (~15% deficient
in gas-filled hohlraums, implicit in Dan’s post shot model)
This is driving work on a variety of hohlraum options—rugby shape,
near-vacuum fill, various sizes, U w/out Au coating
9
DT
DT
DU hohlraums
drive in the preheatclean
band
(“Mclean reduce
mass
mass
frac. = 0.99
frac. = 0.93
band”), part of optimizing the HDC design
“nominal” M-band, all surfaces
to unlined DU)
0.7 x M-band,(equiv
all surfaces
g/cm3
HDC
HDC
density
density
DT
DT
2-shock
Quantitative modeling of the
mix un-doped
dependsHDC
on accurate
clean mass
mass
knowledge of M-band
fraction and roughness seeds clean
frac. = 0.75
frac. = 0.97
2-shock un-doped design"
surface roughness power spectra"
Clark—HDC Design Mini-Workshop
08/23/2013
0.7 x "
M-band"
“nominal”
24% M-band"
18
Dopant in ablator is the other knob
controlling this instability growth
For HDC, dopant options are limited
and unlined DU hohlraums are a big
help
10
Irregularities in CH oxygen seed hydro instabilities
at a level that is important to NIC performance
There is no conclusive proof of this yet, but there
are several very persuasive strands of evidence
2±0.005 at%
~2-5 at%
We know there is ~0.5 at% oxygen in our CH,
ramping up to 2-5 at% at the outside, but know
little about transverse uniformity
30µm scale
length
0.5 at%
Oxygen in GDP.
Lateral variations
“immeasurably
small,” too small
to see in a plot
like this. 0.01%?
Snapshot of
oxygen
content in
GDP ablator
On this side I
multiplied
lateral
variations x100
just to show
what I am
talking about
With lateral
modulations
x100 for
illustrative
purpose
S. W. Haan, H. Huang, M. A. Johnson, M. Stadermann, S. Baxamusa, S.
Bhandarkar, D. S. Clark, V. Smalyuk, and H. F. Robey, Instability growth
seeded by oxygen in CH shells on the National Ignition Facility,
Physics of Plasmas 22, 032708 (2015)
There is a variety of data from target fab
suggesting transverse inhomogeneity; if those
modulations are variations in oxygen near the
outer surface, then simulations say they seed
late-time perturbations ~2-3x bigger than from
surface roughness.
3D HGR experiments show structure that would
be very mysterious without “oxygen hypothesis”
Ÿ For high-foot implosions, the growth is low
(reduction similar to surface roughness growth)
Ÿ For upcoming “adiabat shaping” implosions, the
growth is also low
11
Growth from oxygen is ~3x bigger than from surface
ripples or equivalent density modulations
Oxygen in GDP increases its density according to Δρ= 0.029*at%
I can normalize to “equivalent ρR modulation” using this density increase
8000
Growth factor
in ρR
Oxygen modulation with
e-d/30 ramp from outside
6000
Oxygen only (same amount of
O as equivalent ρR)
i.e. define
RG(t)=
ρR(rms)
ρR(avg)
4000
and plot
2000
RG(Peak Vel)
RG(0)
0
0
Equivalent density
modulation
Oxygen-only
launches a velocity
deviation, shock is
stronger where
oxygen is smaller. 2/3
of oxygen impact is
Outer surface ripple
from this, and 1/3
50
100
150
200 from the density
increase
Mode number
12
Oxygen seeds the most growth when the
modulation is in the outer few µm
Simulations in which oxygen modulations, or density modulations, are placed at
various depths in the shell. Plotted is relative growth in ρR(rms), mode 60, at time
of peak velocity. Normalize to same initial ρR.
5
Relative
growth of
mode 60
Depth
Seed with
oxygen and
density
4
Oxygen and density
with e-d/30 ramp
3
2
Surface ripple
Density alone
1
0
-200
-150
-100
-50
0
Depth in shell where oxygen (and/or
density) are modulated (µm)
13
Oxygen seeds the most growth when the
modulation is in the outer few µm
Simulations in which oxygen modulations, or density modulations, are placed at
various depths in the shell. Plotted is relative growth in ρR(rms), mode 60, at time
of peak velocity. Normalize to same initial ρR.
5
Relative
growth of
mode 60
For oxygen near outer
surface, where this is ~3,
the best measure is
Oxygen * depth
(ppm-µm) and we care
about
Seed with
oxygen and
density
4
Oxygen and density
with e-d/30 ramp
3
2
0.03 at% * 30µm =
1 at%-µm
at 100 µm lateral scale
Surface ripple
Density alone
1
(equivalent to 100nm
roughness)
0
-200
-150
-100
-50
0
Depth in shell where oxygen (and/or
density) are modulated (µm)
14
Why should the oxygen have lateral variations?
2±0.005 at%
~2 at%
1 at%
30µm radial
scale length
Oxygen in GDP.
Lateral variations
“immeasurably
small,” too small
to see in a plot
like this. 0.01%?
Snapshot of
oxygen
content in
GDP ablator
On this side I
multiplied
lateral
variations x100
just to show
what I am
talking about
With lateral
modulations x100
for illustrative
purpose
Many peoples’ intuition is that the oxygen should
be uniform, because the processes that cause it
seem uniform. However:
§  The composition is almost immeausurably
uniform, modulations would be <1% of the
oxygen that we know is there
§  The contour levels we are considering are
smooth to 10s of nm. What does “uniform” even
mean if surface roughness and contour levels
are similar? Processes that lead to surface
roughness could lead to O variation too.
§  The polymer is a blend of complex molecules,
with various bonds that O can find “free
radicals” i.e. C with less than 4 bonds)
§  We know exposure to x-rays and UV changes
the affinity for O
The burden of proof needs to be on those who say it’s too uniform to
matter. Why should the oxygen be uniform to <0.002 at%?
15
Exposure to x-rays, UV, or visible is known to “paint”
oxygen into the CH
Experiment done by Kevin Sequoia ,
GA, four years ago
PR X-ray beam,
narrowed to ~20°!
5-10keV x-rays
rotation!
First expose the shell to a band
of the PR x-rays
K. L. Sequoia, H. Huang, R. B. Stephens, K. A. Moreno, K. C. Chen, and A. Nikroo, Fusion Sci. Technol. 59, 35 (2011).
16
Exposure to x-rays, UV, or visible is known to “paint”
oxygen into the CH
Experiment done by Kevin Sequoia ,
GA, four years ago
1 mm
(16 rows)
X-ray beam!
5-10keV x-rays
16 detectors!
rotation!
Then rotate it and use the usual
PR setup to diagnose it
K. L. Sequoia, H. Huang, R. B. Stephens, K. A. Moreno, K. C. Chen, and A. Nikroo, Fusion Sci. Technol. 59, 35 (2011).
17
Exposure to x-rays, UV, or visible is known to “paint”
oxygen into the CH
Experiment done by Kevin Sequoia ,
GA, four years ago
1 mm
(16 rows)
X-ray beam!
5-10keV x-rays
16 detectors!
This increased PR absorption,
equivalent to deposition of 0.2 at%
of O in the exposed band (uniform
in radius).
rotation!
Then rotate it and use the usual
PR setup to diagnose it
Similar process occurs from UV and
optical exposure.
More recent work by Sal Baxamusa has identified three processes by which O
is attached into GDP:
Ø  H2O is absorbed from ambient (the 3at%-30µm outer layer & 0.5 at% bulk)
Ø  High photon energies (x-rays, hard UV) create additional sites into which H2O
bonds
Ø  Lower photon energies (soft UV, light) induce absorption of O2, which
otherwise diffuses around unattached within the porous GDP
K. L. Sequoia, H. Huang, R. B. Stephens, K. A. Moreno, K. C. Chen, and A. Nikroo, Fusion Sci. Technol. 59, 35 (2011).
18
Slide from Sal Baxamusa
Unlike most plastics, even low-energy photons
cause GDP to photo-oxidize
0.3
Δ O (at%) = 9.5*A OH/CH
0.25
0.2
Hg arc lamp
365 nm LED
405 nm LED
445 nm LED
532 nm LED
0.15
0.1
0.05
4Pi optical characterization,
blue LED, penetration ~5µm
0
-0.05
2
10
2-3 weeks of typical room light
Fill tube UV glue cure, new protocol*
3
4
10
10
2
Exposure dose (mJ/cm )
5
*Prior to Aug 2014,
10 UV exposure
was several X higher, caused
slumping in N140819
Generating reliable dose-response curves is an ongoing activity
Routine “non-destructive” capsule characterization is
estimated to cause about 0.1 at% photoabsorbed oxygen
Sal’s estimated typical additional
oxidation due to 4pi blue LED
Penetrates ~5µm (~1/6 of our test case)
High-O regions are where diagnostic goes
back for closer look at interesting areas
Ablation front deformation at peak
velocity
0.1 at%
2.5µm
0.03 at%
-1.2µm
This routine characterization step does NOT look like an important
source of oxygen nonuniformity
20
Target characterization has given us a lot of
information about O in GDP, but not enough yet
GDP has oxygen, mostly in the outer ~20µm, ramping from
bulk ~0.5 at% up to 2-5 at% at surface
Oxygen comes in mostly as H2O; molecular O2 diffuses more freely
in shell. Absorption does happen, often photon-assisted
Oxygen in GDP increases its density by 0.029 g/cc/at%. Corresponds
to just fitting in without changing CH structure.
Amount of O depends on history, light/UV exposure, and on the
density of available bonds in the GDP
Transverse variability of something internal is indicated by “PSDI
through the shell” (Mike Johnson) and consistent with Precision
Radiography data (Haibo Huang)
Something like “2±0.01 at% O in outer 30µm” is consistent with
everything we know, suggested by “PSDI through the shell,” and
causes perturbations about 2x those caused by the surface spec
21
3D Rayleigh-Taylor experiments are a great test of
whether there is something like this going on
Experiments by Vladimir Smalyuk,1 Dan Casey, design Steve Weber.
Goal: image the
modulations in single-pass
transmission for “native
roughness” GDP. Previous
experiments (Kumar Raman
et al.) with preimposed
surface ripples had
confirmed our ability to
calculate growth fairly well
As we went into these last fall, the “oxygen-dominant hypothesis”
predicted that we would see larger perturbations than expected from
surface roughness alone. (For previous experiments with large preimposed single-mode ripples, the big ripples dominate, but as we go to
smoother surfaces, the oxygen starts to dominate.)
1V.A.
Smalyuk, S.V. Weber, D.T. Casey, D.S. Clark, J.E. Field, S.W. Haan, A.V. Hamza, D.E. Hoover, O.L. Landen, A. Nikroo, H.F. Robey and C.R. Weber (2015). Hydrodynamic
instability experiments with three-dimensional modulations at the National Ignition Facility. High Power Laser Science and Engineering, 3, e17 doi:10.1017/hpl.2015.12
22
Steve Weber calculated what it should look like, from
the surface roughness
Experiments had pre-imposed divots, 2nd and 3rd had tents
Divots
3D
simulation
with blurring
of 3rd shot
(Steve
Weber)
Accidental
feature
23
The 3D Rayleigh-Taylor experiments would be utterly
mysterious if we hadn’t thought about oxygen seeding
Experiments had pre-imposed divots, 2nd and 3rd had tents
On 3rd shot, we eliminated much of the
UV irradiation. Happened to have a surface
feature right in the field of view.
The first two shots showed
features that could have
3D
been caused by UV during
simulation
assembly. So far we do
with blurring
NOT have a detailed
of 3rd shot
scenario to explain this.
(Steve
Preimposed divots
First shot: “the ring”
and overall roughness
more than expected
Weber)
Tent
Second shot: “the eye”
and overall roughness
more than expected
Accidental
feature
Experiment.
Feature more
pronounced
than expected
(associated
oxygen?). Other
perturbations
also bigger.
400 µm
24
High-foot and “adiabat shaped” implosions are less
sensitive to the oxygen-seeded perturbation growth
The usual growth factor to peak velocity
of conventional surface perturbations
Seeded with oxygen, amplitude at peak
velocity / equivalent initial ρR modulation
2500
Low-foot
Low-foot
“Adiabat shaping”
with reoptimized
peak
800
600
“Adiabat shaping”
pulse
400
200
High-foot
0
-200
0
50
100
Mode number
150
200
Growth factor to ablation front
Growth factor to ablation front
1000
“Adiabat shaping” with
reoptimized peak
2000
“Adiabat shaping”
pulse
1500
1000
500
High-foot
0
0
50
100
150
200
Mode number
•  Generally, O-seeded growth is ~2-3x bigger than from surface ripples with same ρR
•  High-foot growth is 2-3x less than low-foot, for either seed, except at lowest modes
•  “Adiabat shaping” pulse is also less sensitive to oxygen modulations, albeit not
quite as much so as the high-foot implosions, and depends on details
25
There is a somewhat mature campaign using HDC
capsules
There are several ignition designs that differ in hohlraum details, pulse shape. Capsules
differ only in the thickness of the doped layer. Many current experiments use undoped shells.
1110 µm
Planning assumes surface roughness based on high-def
AFM, which is ~3x rougher than we used to expect for HDC
HDC 3.476 g/cc
76 µm thick
0.275 at% W
20 µm thick
(for 3-step design)
DT 0.255 g/cc
55.6 µm thick
Outer surface is wellcharacterized and wellreported. Inner surface and
bulk homogeneity might still
have issues.1
1W.
Recent capsules have nano-crystalline outer coating and
look similar to the “old HDC” curve!
Requieron, yesterday
26
The Be program is just getting underway, under
LANL leadership
1050 µm
Be+Cu
(Total
thickness
160 µm)
925
920
900
895
890
DT ice (70 µm)
DT
gas
0%
0.4%
1.0%
0.4%
0%
This is the design currently being
tested
6 shots so far (3 keyholes, 1
symcap, 2 ConAs). First DT layer
last week, results too late for slides.
Still sorting out hohlraum symmetry,
details of pulse shaping. All results
so far are very positive, and we are
excited about the prospects with Be
J.L. Kline, D.C. Wilson, A.N. Simakov, and A. Yi
Andrei N. Simakov, Douglas C. Wilson, Sunghwan A. Yi, John L. Kline, Daniel S. Clark, Jose L. Milovich, Jay D. Salmonson and
Steven H. Batha, Optimized beryllium target design for indirectly driven inertial confinement fusion experiments on the National
Ignition Facility, Phys. Plasmas 21, 022701 (2014)
27
Be fabrication issues are still being explored
What we are making is good for current experiments, but there are
issues ahead:
Coating process is inclined to make defects
(voids, few µm x 100nm, and “nodules” with
~2% density defect)
Irregular Cu diffusion1 seems to be under
control via oxide layers,2 but diffusion and/
or layer irregularities could still be
unacceptable
We are still iterating on final surface
roughness requirements
Coating includes Ar and O, we need to
ensure that homogeneity is acceptable
1Huang
et al. Fusion Sci. Technol. 63, 190 (2013)
et al. Fusion Sci. Technol. 63, 208 (2013)
2Youngblood
28
Summary/Outline
We are making good progress toward ignition, and
target
fabrication
is a vital ingredient
For the
three shotsprogress
we have considered
in detail,
simulations show rough agreement with the data but
not a perfect match
This 3D simulation by Dan Clark of shot N120321 nicely shows the issues
keV
N120321
t = 22.85 ns
(bang time)
g/cm3
4.0
500.0
3.0
375.0
2.0
250.0
1.0
125.0
0.0
0.0
fuel-ablator
interface
fill tube
defect
tent
defect
LawrenceImploded
Livermore National
Laboratory
shape
at bang
time
Green surface = fuel-ablator interface
Temperature on left, density on right
We are incorporating our most upTent understanding
feature is 2-3x
bigger
to-date
of the
role than we
1-2 years
ago
ofrealized
the tent, hohlraum
drive
asymmetries, and other
perturbation
sources into 2-D
and
Global out-of-round
from
3-D
simulationsasymmetry
of NIF implosions
hohlraum
While we do not match all of the
Shorter
wavelength
irregularities
data
within the
error bars, there
is
are agreement
probably for
bigger
than
rough
both low
andshown,
high
foot shots,
and
heavily
because
of “clean”
oxygen
nonuniformity
mixed
shots
in GDP
The precise effects of the tent and
The program
is working
to
hohlraum
asymmetries
are
minimize
these drivers
features
and their
probably
the biggest
for the
remaining
growthdiscrepancies
and impactand
subjects for continuing work
Developing other options is also
important: High Density Carbon,
Be, other hohlraum designs
6
29