UCRL-CONF-200359
Update on NIF indirect drive
ignition target fabrication
specifications
S. W. Haan, P. A. Amendt, T. R. Dittrich, S. P. Hatchett,
M. C. Herrmann, O. A. Hurricane, M. M. Marinak, D.
Munro, S. M. Pollaine, G. A. Strobel, L. J. Suter
October 17, 2003
Inertial Fusion Sciences & Applications (IFSA)
Monterey, CA, United States
September 7, 2003 through September 12, 2003
Disclaimer
This document was prepared as an account of work sponsored by an agency of the United States
Government. Neither the United States Government nor the University of California nor any of their
employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for
the accuracy, completeness, or usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any
specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise,
does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United
States Government or the University of California. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States Government or the University of California,
and shall not be used for advertising or product endorsement purposes.
Update on NIF indirect drive ignition target fabrication specifications
S. W. Haan, P. A. Amendt, T. R. Dittrich, S. P. Hatchett, M. C. Herrmann,
O. A. Hurricane, M. M. Marinak, D. Munro, S. M. Pollaine, G. A. Strobel,
1
and L. J. Suter
Lawrence Livermore National Laboratory, Livermore CA
1
also at Phys Department, Univ of Georgia, Athens GA
ABSTRACT
Indirect drive ignition target simulations
are described as they are used to determine target
fabrication specifications. Simulations are being
used to explore options for making the targets
more robust, and to develop more detailed
understanding of the performance of a few point
designs. The current array of targets is described.
A new target is described with radially
dependent Cu dopant in Be. This target has
significantly looser specifications for high-mode
perturbations than previous targets. Current
estimates of size limitations for fill tubes, holes,
and isolated defect are discussed. Recent 3D
simulations are described.
I. INTRODUCTION
This presentation is an update on the
fabrication specifications as they result from the
indirect drive ignition target design effort at
LLNL. In general, recent work is in one of two
areas. First, we are exploring options for making
the target designs more robust. This includes
better capsule optimization, graded Cu dopant in
Be, and hohlraum design modifications. Second,
we continue to pursue more detailed modeling of
select point designs in order to enhance our
confidence in their performance. This includes
3D modeling of the targets.
The basic design concept, and our
understanding of the specifications, remains the
same as it has been since NIF was proposed. The
specifications are described in the previous
proceedings.
1,,2
The target is a cylindrical
hohlraum about 1 cm long by 6 mm diameter, of
Au or a "cocktail" of U:Nb
0.14
:Au:Ta:Dy. The
hohlraum is filled with He or H+He gas
contained by 1 µm CH or polyimide windows
over each laser entrance hole (LEH). Each LEH
is about 3mm diameter. The LEH is coated with
30 µm of CH, which comes 600 microns up on
the inside of the hohlraum. The capsule has a
plastic or beryllium single shell ablator, about
120 µm thick, supported by a 0.1 µm thick
polyimide tent. The ablator encloses a cryogenic
solid DT layer about 80 µm thick. The hohlraum
may contain anti-convection foils, thickness less
than 1 µm. While the basic design remains the
same as it has been for a number of years, many
details are being optimized as explained below.
II. TARGET OPTIONS
First regarding the hohlraum, there are
various options that we should be maintaining.
The hohlraum length could be between 7 mm
(for sub-ignition implosions to verify target
physics with half of NIF) up to 16 mm (large
high-yield designs). Note that the LEH diameter
is likely to stay at about 6-8 mm, independent of
hohlraum size, because the laser spot size will
probably be about the same for all the designs.
The hohlraum material could be gold or the
cocktail as mentioned above, and the fill could
be pure He or a H/He mixture. We are actively
considering the specification on tenting. Anti-
convection foils are possible, provided their
mass is significantly less than the mass of the gas
nominally included.
Designs using 0.5 µm light (2ω) may allow
for considerably more energy delivered to the
capsule, albeit possibly at lower drive
temperature. From the point of view of capsule
design and optimization, and target fabrication,
the color of the light is not important; however, it
does affect where one is likely to be in the
parameter space of target size as the 2ω targets
are considerably larger. Outer radii of Be targets
could be as large as 2.5 mm. Of course higher
absorbed energy is very valuable to capsule
performance; one example that has been
examined in some detail is uniformly doped
Be(Cu) capsules at 250 eV, for which the surface
roughness specification increases from 10 nm
(the so-called "NIF standard" surface roughness)
at 250 kJ absorbed, to 60 nm at 600 kJ absorbed.
The range of possibilities is only weakly
dependent on the color of the laser light; either
2ω or 3ω could be used to drive large low
temperature hohlraums, although 2ω would
make it more likely that we would be fielding
very big targets.
UCRL
-CONF-
200359
Finally, the hohlraum designers are
considering variants on the lining and fill of the
hohlraum. Possibilities being considered include
hohlraums filled with foam of SiO
2
, GeO
2
, or
XeO
2
, all at about 1 mg/cc; or linings on the
hohlraum of about 1 µm of Kr or Xe, while the
hohlraum is filled with 0.1 mg/cc of He. Both of
these variations work in simulations, and may
improve laser-plasma interactions and/or
cryogenic fielding.
We continue to consider the full suite of
capsules described in ref. 1. An additional new
design, Be with graded Cu dopant, is described
in more detail below. We also continue to refine
the capsule design. The hohlraum physics and
laser performance will eventually set constraints
on capsule size, pulse shape, and drive
temperature. The main question we need to
answer now, to guide target fabrication and the
other elements of the program, is how to
optimize the capsule performance given these
constraints.
A new way to provide this optimization has
been developed by one of us (Mark Herrmann).
Herrmann's innovations are twofold: a better
implementation of the constraints than
previously, and an automated process that
thoroughly optimizes over the remaining free
parameters. From the point of view of target
fabrication, this reoptimization has two
consequences: the capsule dimensions are being
adjusted, by perhaps 15-20% thickness increase
on both ablator and fuel; and the specification on
surface roughness may be loosened as the target
design becomes more robust. (However, as
mentioned above, the roughness specifications
are a strong function of power and energy, so
that there is likely to be more impact on them
from laser and hohlraum performance than from
this reoptimization.)
This scan technique has been applied to a
variety of targets, including the polyimide
1.1
mm 300 eV capsule, CH + 0.25% Ge at the
same outer radius and peak drive, and four
ablator materials at 250 eV: CH, CH(Ge),
polyimide, and Be(Cu). We find that at 250eV,
undoped CH is comparable in performance to
uniformly doped Be(Cu). Both are considerably
superior to polyimide and doped CH. Full
specifications for these targets are still being
developed.
III. GRADED DOPED BERYLLIUM
At the time of this meeting, in June 2003,
we had just begun looking at graded doped
beryllium capsules. Further work since then has
been quite exciting. Because it is very relevant to
target fabrication, we are including that work in
this paper even though the material was not
presented at the conference.
Graded dopants have been considered
previously, in fact the original NIF designs used
graded Na and Br dopants in Be, as did the
250eV design describe in Lindl.
3
These early
designs were not optimized in detail, and used
monotonically decreasing dopant concentration.
More recently, Dittrich
4
did a detailed
optimization of a graded doped capsule meant to
be driven at very low energy and temperature
(115 kJ absorbed and 250 eV). Dittrich used a
grading scheme proposed by Hatchett
(unpublished). At the original very small scale,
Dittrich's design was not very robust, but since
then we continued to find that other 250 eV
targets were noticeably less robust than Dittrich's
design at that scale. Hence we reexamined this
design, looking at larger more typical NIF scales,
as well as considering
a similar 300 eV
design.
These designs
are much less
susceptible to
acceleration-phase
Rayleigh-Taylor
growth than any other
Figure 1. Be capsule with graded Cu dopant. The 300eV design is shown. the 250eV designs are scales of the design
by Dittrich (ref. 2). The table shows the allowed ablator rms roughness (1/3 of the ablator perturbation that causes
50% yield degradation.
848
928
1105 µm
934
940
995
1011
0
0.35%
0.7%
0.35%
0
Cu:
300 eV design:
Be next to fuel is
undoped, dopant
rises to max in two
steps and back down
again
Ablator roughness specs on four designs:
T
R
E
abs
max Cu % Ablator rms
250 115kJ 0.35% 7 nm
250 300kJ 0.35% 100 nm
250 700kJ 0.35% 280 nm
300 200kJ 0.7% 220nm
ignition target that we have modeled. Figure 1
shows the 300 eV design, and the Rayleigh-
Taylor results for the cases we have examined.
The quantity tabulated as "ablator rms" is one
third of the ablator surface roughness, with the
NIF standard spectrum in spherical harmonic
modes 12 and above, that causes yield
degradation to 50% of clean. Typically, past
capsules have required 20-40 nm. Dittrich's
original design is included in the table; its
specification is indicated, which for its size it is
the most stable 250eV design known. The other
scales, and the 300 eV design, can tolerate very
rough surfaces.
Even though sputtered beryllium surfaces
have been considered unacceptably rough, these
designs are stable enough to tolerate existing
sputtered surface roughness. The technology of
constructing graded layers is probably a
straightforward extension of sputtered beryllium
fabrication: one changes, as a function of time,
the copper concentration in the growing layer.
Issues remain of filling and characterizing the
layer, as in all beryllium capsules, but the
additional stability of these capsules probably
makes them more forgiving of the possible
solutions to these issues.
The 300eV capsule is also very tolerant of
short wavelength ice perturbations. The thick
ablator and high ablation velocity make for low
feed-out and low growth. The simulations used
the old spectrum as described in ref. 2. A 5 µm
rms ice roughness produces 50% yield; a
specification would be 1.7 µm. For this spectrum
with 1.7 µm rms total, the rms in modes 10-15,
which are likely to be the most important, is 0.46
µm. For comparison, typical recent Omega ice
surfaces can be described as approximately
R
lm
= 10µm l
-1.8
.
Such a spectrum has total rms 5.4 µm,
somewhat higher than the best Omega surfaces
of 1.8 µm rms. With an l
-1.8
spectrum, with a total
rms of 5.4 µm, the rms in modes 10-15 is 0.39
µm. Hence we can tentatively draw the
remarkable conclusion that this capsule could
tolerate "typical" Omega ice roughness,
including the nominal safety factor of three. This
conclusion does not extend to modes below 10,
for which (see below) we expect the
specification to be similar to the typical
requirement as given in ref. 1. Hence achieving
adequate low modes will be the primary
challenge for the ice in this capsule. Any surface
that has adequate low modes will very likely
have more than adequate amplitudes for modes
above 10.
As just implied, we examined the
sensitivity of the 300 eV graded-doped capsule
to low modes in the ice and ablator. Its
sensitivity is very similar to the baseline
polyimide capsule —the growth factors for low
modes, and the amount of bang-time low mode
perturbation that the capsule can tolerate, are
virtually identical to the baseline. The
extraordinary features of this capsule are only in
the growth of perturbations with mode numbers
greater than about 15.
Fielding scenarios for these capsules
include the possibilities of fill tubes and fill
holes. Their remarkable high-mode stability,
together with the high-mode structure that is
characteristic of fill tubes and holes, suggest that
these capsules may be able to tolerate relatively
large fill tubes and holes. While we have not had
the opportunity to do full simulations, we can
use growth factors inferred from less difficult
simulations to estimate how large a tube or hole
might be tolerated. These estimates are based on
linear analysis, and a fill feature is certainly not
linear. The only reason to hope that the analysis
might be approximately correct is if the very
short wavelength characteristics do not matter,
and the evolution is dominated by the average
mass defect as it seeds the fastest-growing
modes, which are much larger than the lateral
size of the feature. For other targets for which we
have done detailed fill hole simulations, linear
analysis estimates give reasonable results: for
baseline polyimide, linear analysis indicates that
a fill hole should be less than 2-4 µm diameter,
and a fill tube diameter, divided by the tube's
ρ
0.333
to give equivalent mass defect, should be
less than 8-12 µm. Using growth factors for the
300eV graded doped capsule, we find that a hole
should be less than 10-15 µm diameter, and a
tube less than 15-20 µm diameter. (There is less
leverage of the growth factors on the tube than
on the hole: the effective mass defect for the tube
goes as the cube of the tube size, since this
approximation assumes that a few tube diameters
contribute to the mass defect, while the mass
defect for the hole goes as the square of the hole
diameter.) Of course these numbers are very
preliminary, but they provide a rough idea of
how important the increased stability might be to
fill tube and hole requirements.
Because these capsules are so robust with
respect to short wavelengths, there may be a