Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Fuel Gain Exceeding Unity in An Inertially Confined Fusion Implosion

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

LETTER

doi:10.1038/nature13008
Fuel gain exceeding unity in an inertially confined
fusion implosion
O. A. Hurricane
1
, D. A. Callahan
1
, D. T. Casey
1
, P. M. Celliers
1
, C. Cerjan
1
, E. L. Dewald
1
, T. R. Dittrich
1
, T. Doppner
1
, D. E. Hinkel
1
,
L. F. Berzak Hopkins
1
, J. L. Kline
2
, S. Le Pape
1
, T. Ma
1
, A. G. MacPhee
1
, J. L. Milovich
1
, A. Pak
1
, H.-S. Park
1
, P. K. Patel
1
,
B. A. Remington
1
, J. D. Salmonson
1
, P. T. Springer
1
& R. Tommasini
1
Ignition is needed to make fusion energy a viable alternative energy
source, but has yet to be achieved
1
. Akey step on the way to ignition
is to have the energy generated through fusion reactions in an
inertially confined fusion plasma exceed the amount of energy
deposited into the deuteriumtritium fusion fuel and hotspot dur-
ing the implosion process, resulting in a fuel gain greater than
unity. Here we report the achievement of fusion fuel gains exceed-
ing unity on the US National Ignition Facility using a high-foot
implosion method
2,3
, which is a manipulation of the laser pulse
shape in a way that reduces instability in the implosion. These
experiments show an order-of-magnitude improvement in yield
performance over past deuteriumtritium implosion experiments.
We also see a significant contribution to the yield from a-particle
self-heating and evidence for the bootstrapping required to accel-
erate the deuteriumtritium fusion burn to eventually run away
and ignite.
At the National Ignition Facility (NIF), 192 lasers deliver up to
1.9 MJ of light into a gold hohlraum, a cylindrically shaped radiation
cavity (Fig. 1), that converts the energy into a nearly Planckian X-ray
bath. A fraction of the X-rays are absorbed by a capsule generating
,100 Mbar of pressure in the ablator (the outer shell of the capsule).
This ablation pressure, delivered as a series of weak shocks, accelerates
the capsule inwards. Against the inside of the ablator is the deuterium
tritium(DT) fuel shell, whichis initially ina cryogenic ice state. When
the implosion achieves peak velocity, the fuel has a kinetic energy that
is a fraction of the X-ray energy absorbed by the capsule. As the fuel
stagnates (stops moving), abruptly arrested by the high pressures
forming at the centre of the implosion, the DT forms a hotspot from
the fuels inner surface andPdVwork(pressure times volume change) is
done on the hotspot. The hotspot initiates the fusion reactions, pro-
ducing neutrons and a-particles as the hotspot ion temperature climbs
to many kiloelectronvolts. At sufficient hotspot areal density, (rr)
hs
.
0.3 g cm
22
, and ion temperature, T
ion
.4 keV (where Boltzmanns
constant has been suppressed), the hotspot will ignite as a-particles
redeposit their energy locally. If the fuel areal density, (rr)
fuel
.1 g
cm
22
the burn will propagate (heat adjacent cold fuel, causing it also
to fuse and burn) and a run-away self-heating process releases energy
many times greater than that absorbed by the capsule.
Mix of the ablator and DT can degrade the ability of an inertially
confinedfusionimplosionto compress the DTfuel and canalso cause
undesirable cooling because high-atomic-number (high-Z) materials
inthe DThotspot will rapidly radiate away energy inthe formof brem-
sstrahlung emission, the power of emission scaling in proportion to Z
2
.
Among many motivations, the high-foot implosion
2,3
was developed in
the wake of the National Ignition Campaign
4,5
(NIC) primarily to
address the possibility that ablation-front-driven instability
6,7
was
responsible for part of the observed degraded yield performance
1
and
the ablatorfuel mix inferred fromX-ray emissions in combinationwith
primary neutron yield
8,9
.
The high-foot implosion is designed to reduce ablation-front-
driven instability growth and thereby inhibit ablator plastic (carbon
hydrogen and silicon dopants) from mixing into and contaminating
the DThotspot. The laser pulse shape is designedto obtaina relatively
high hohlraum radiation temperature (T
rad
<90100 eV) during the
foot of the pulse (Fig. 1) and launches three shocks. In contrast, the
NICimplosionpulse shape drives a lower radiationtemperature (T
rad
<
60 eV) inthe foot (hence low-foot) for longer andlaunches four shocks.
The essential stability benefits of the high-foot scheme can be under-
stood from examining an expression for the linear growth rate of the
ablation-driven RayleighTaylor instability
10
c
ARTI
~a
2
Fr, n

kg
1zkL
r

{b
2
Fr, n kv
a
1
where k is the perturbation wavenumber, g is the ablator acceleration, L
r
is the density gradient scale length of the ablation front, v
a
is the ablation
velocity, and a
2
and b
2
are parameters of order unity whose exact values
depend on a heat conduction scale-length parameter, n, and the Froude
number, Fr~v
2
a
_
gL
r
_ _
. The key stabilizing effects of the high-foot drive
enter throughthe higher ablationvelocity, whichscales as T
9=10
rad
, increas-
ing the b
2
kv
a
ablative stabilization term of equation(1), and through an
increase in L
r
, which reduces the instability term proportional to

kg
_
.
The increase in L
r
is primarily due to a stronger first shock, which
increases the adiabat of the implosion and prevents the ablator from
becoming so highly compressed (risking break-up) during the implo-
sion. Theenhancedstabilitycanbe further understoodby comparing the
respective in-flight aspect ratios (R
in
/DR, where R
in
is the ablator inner
radius and DR is the ablator thickness) of the high- and low-foot implo-
sions: for the high-foot implosion, the in-flight aspect ratio is roughly
half that of the low-foot implosionthe amplitude of instability growth
is directly related to the exponential of

R
in
=DR
_ _
2 (ref. 11). The trade-
off made to obtain the improved stability of the high-foot implosion is
that the DT fuel adiabat, a 5P/P
F
(usually denoted a), where P is
pressure and P
F
is the Fermi pressure, is higher, making the fuel less
compressible for a given amount of absorbed energy. (An alternate
definition of the adiabat using P
cold
, the minimum DT pressure at
1000 g cm
23
, is sometimes used
12
.) Details of the stability benefits, other
theoretical motivations and trade-offs involved in the high-foot implo-
sion, and the initial results from the first set of five DT implosion
experiments are described elsewhere
2,3
.
Deuteriumtritium implosions N130927 and N131119 (NIF shot
number in yearmonthday format YYMMDD) build on the previous
high-foot shot, N130812
3
, by modestly increasing the NIF laser power
and energy (Table 1) and by redistributing energy between different
laser beams, through laser wavelength changes that affect the cross-
beam transfer (the transfer of power from one beam to another via
induced Brillouin scattering), to optimize the illumination pattern in
1
Lawrence Livermore National Laboratory, PO Box 808, Livermore, California 94551, USA.
2
Los Alamos National Laboratory, Los Alamos, New Mexico 87545, USA.
0 0 M O N T H 2 0 1 4 | V O L 0 0 0 | N A T U R E | 1
Macmillan Publishers Limited. All rights reserved 2014
the hohlraum
1316
. Although the hotspot shape changes that result
from these wavelength changes can be predicted to some extent
17
, in
practice the precise wavelengths needed to achieve the desired (that is,
round) shape are found empirically. For N130927, the choice of
l
23.5
2l
30
50.7 A

between the 23.5u and 30.0u inner-cone beams was


chosen for azimuthal symmetry control, with Dl
23.5outer
59.2 A

and
Dl
30outer
58.5 A

(the respective laser wavelength differences between


the 23.5u and 30u inner-cone beams and the outer-cone beams) used for
equatorial symmetry control (see Fig. 1 for beam angles). For N131119,
Dl
23.5outer
59.5 A

and Dl
30outer
58.8A

. These wavelength choices


were critical for keeping the hotspot shape under control as the implo-
sion was pushed to higher velocities, because previous experiments had
already shown the tendency for the hotspot to deform into an oblate
toroidal shape when laser power was increased
3
. There are limits to the
amount of control that can be exerted over the hotspot shape just
through wavelength changes alone, and physical changes to the hohl-
raum may also be required in future experiments to maintain hotspot
(and fuel) shapes that will achieve the desired results.
We used a gold hohlraum of 5.75-mm diameter and 9.425-mm
length, which are typical values in most high-foot cryogenic DT
implosion experiments (Fig. 1). The same hohlraum geometry was
used during the NIC for most of the low-foot shots. As is typical for
the high-foot series, the hohlraum was filled with heliumgas of 1.6 mg
cm
23
density (as compared with 0.96 mg cm
23
for the NIC), the pur-
pose of which is to restrict and delay ingress of gold plasma from the
inside wall of the hohlraum, which can impede laser beam propaga-
tion. The plastic capsule at the centre of the hohlraumfor N130927 and
N131119 respectively had outer-shell radii of 1.1315 and 1.1241 mm
and inner-shell radii of 0.9365 and 0.9303 mm (Fig. 1). Layered on the
inner surface of the capsule shell for N130927 and N131119 were 71.4
Shot
N130927
Au hohlraum
Cryogenic cooling ring
Laser
entrance hole
Inner cone
Laser quads
Outer cone
Low foot
High foot
N110914
4 shock
300
200
100
T
r
a
d

(
e
V
)

t (ns)
0.0 5.0 10.0 15.0 20.0
0.0
N130927
3 shock
a b
c
23.5
30
44.5
50
44.5
50
30
23.5
50
44.5
44.5
23.5
30
44.5
50
44.5
50
30
23.5
50
44.5
44.5
DT ice layer
CH ablator graded
2% Si doped
2.263 mm
Figure 1 | Indirectly driven, inertially confined fusion target for NIF.
a, Schematic NIF ignition target showing a cut-away of the gold hohlraumand
plastic capsule with representative laser bundles incident on the inside surface
of the hohlraum. b, X-ray image of the actual capsule for N130927 with DT
fuel layer and surrounding CH (carbonhydrogen) plastic ablator. c, X-ray
radiationdrive temperature versus time for the NIClow-foot implosionandthe
post-NIC high-foot implosion.
Table 1 | Measured and derived implosion performance metrics
Quantity N131119
425 TW
1:9 MJ
N130927
390 TW
1:8 MJ
N130927
25
N130927
26
N130927 (sim.)
Y
1315
(neutron) (5.260.097) 310
15
(4.460.11) 310
15
7.6 310
15
T
ion
(keV) DT 5.060.2 4.6360.31 4.2
T
ion
(keV) DD 4.360.2 3.7760.2 3.9
DSR (%) 4.060.4 3.8560.41 4.1
t
x
(ps) 152.0633.0 161.0633.0 137
P0
x
, P0
n
(mm) 35.861.0, 3464 35.361.1, 3264 32
P2/P0
x
20.3460.039 20.14360.044
P3/P0
x
0.01560.027 20.00460.023
P4/P0
x
20.00960.039 20.0560.023
Y
total
(neutron) 6.1 310
15
5.1 310
15
8.9 310
15
E
fusion
(kJ) 17.3 14.4 25.1
r
hs
(mm) 36.6 35.5 34.442.3 35.736.0 32.2
(rr)
hs
(gcm
22
) 0.120.15 0.120.18 0.130.19 0.10.14 0.15
E
hs
(kJ) 3.94.4 3.54.2 3.75.5 3.714.56 4.1
E
a
(kJ) 2.22.6 2.02.4 2.02.4 2.02.5 2.8
E
DT,total
(kJ) 8.59.4 10.212.0 10.013.9 10.9211.19 13.4
G
fuel
1.82.0 1.21.4 1.041.44 1.281.31 1.9
Lines 19 for columns 2 and 3 are directly measured quantities; others are derived from the data. Columns 46 show results from two data-driven models and simulation, respectively.
RESEARCH LETTER
2 | N A T U R E | V O L 0 0 0 | 0 0 M O N T H 2 0 1 4
Macmillan Publishers Limited. All rights reserved 2014
and, respectively, 69.3 mm of cryogenic DT ice that was held at 0.8 K
belowits triple point for a shot temperature of 18.6 K, like all high-foot
DT shots. This mini-quench condition generally produces an ice
layer with fewer ice cracks than that produced by a full quench
18
.
Characterization of the capsule surface showed a roughness typical
of implosion capsules for NIF, and characterization of the DT ice
showed a roughness well within requirements. The very high quality of
the DT ice layer on N130927 was probably not a significant factor in
its performance because the third-best shot on the NIF (N130812) had
an ice layer that was somewhat worse (with greater surface roughness)
than average. The quality of the DT layer for N131119 was between
the qualities of the layers for N130812 and N130927.
Table 1 shows the key measurements and performance metrics for
NIF shots N130927 and N131119. Key measured quantities are the
neutron yield, Y
1315
, in the 1315-MeVenergy band around the char-
acteristic 14.1-MeVDTfusionneutronenergy; the burn-averagedion
temperature, T
ion
; the neutron and X-ray burn widths, respectively t
n
and t
x
; the down-scatter ratio (DSR); and the time of peak neutron
brightness, or bang-time, t
b
. On the NIF, Y
1315
is an average of many
diagnostics, including four neutron time-of-flight (NToF) detectors
19
,
numerous radiochemical activation measurements
20
and a magnetic
recoil spectrometer
21
. The temperature T
ion
is directly related to the
temporal spread obtained from the full-width at half-maximum of the
NToF detectors. A temporal c-ray history gives t
n
(for the high-foot
experiments, t
x
and t
n
are consistent to within their errors). The DSR
comes from measuring, via NToF and the magnetic recoil spectro-
meter, the number of neutrons scattered into the energy range 10
12 MeV, and is directly relatedto the areal density of the cold DTfuel,
(rr)
fuel
<20.3f 3DSR (where f depends upon the amount of ablator
mass remaining but is typically 0.95 60.05 (ref. 21 and B. K. Spears,
personal communication)). Other diagnostics such as X-ray imaging
and neutron imaging (Fig. 2) give information on the shape of the
implosion.
In what follows, we will use the aforementioned observables, which
are measured over the duration of the fusion burn, to infer the amount
of energy that was deposited into the DT (both fuel and hotspot), to
make a comparison with the amount of energy generated from fusion.
The details of the analysis will focus on N130927; the results for
N131119, which exceeded the performance of N130927, are quoted
in Table 1. The analysis outlined in this letter uses an essentially one-
dimensional onion-skin picture with a hotspot of uniform density
and temperature surrounded by the fuel (with Gaussian or uniform
radial density profile), although the observed three-dimensional hot-
spot shape information is used to obtain the hotspot volume. Also, an
assumption of approximately equal ion and electron temperatures,
T
ion
<T
e
, is made and can be justified post hoc using an expression
for the electronion collision time after the hotspot density is obtained.
Analytical and simulation results based on less simplified assumptions
are also quoted in Table 1 for comparison with what is detailed below.
By analysing the observed hotspot shape (Fig. 2) in terms of Legendre
modes (equatorial view, lines 69 of Table 1) and Fourier modes (polar
view), where the hotspot perimeter, as defined by the 17%-of-peak-
brightness contour, is given by
R
hs
h ~P0 1z

?
~2
P
P0
_ _
P

cos h
_ _
2
where P

(cos h) is the Legendre function, we obtain the hotspot volume,


V
hs
(Methods), and the effective spherical radius, r
hs
5(3V
hs
/4p)
1/3
.
(We note that there is no absolute reference for the X-ray or neutron
images, and so mode 51 is not included in the shape description.
However, 51 and m51 motions can be obtained from the NToF
detectors.) The total neutron yield, Y
total
, can be calculated from
Y
total
5Y
1315
exp(4DSR), which accounts for the neutrons produced
but then scattered out of the measured 1315-MeV energy band by
the cold and dense DT fuel. Because for DT fusion reactions the
energy per fusion is known (14.1MeV per neutron and 3.5 MeV per
a-particle), E
fusion
, the total fusion energy produced, can be calculated
from Y
total
.
Fromthe measured T
ion
, the DTreactionrate per unit volume, sv,
can be calculated using standard formulae
22
(Methods). For N130927,
sv 54.75 310
218
1.03 310
217
cm
3
s
21
The range of values is dri-
ven by the measurement uncertainty in T
ion
. The reported T
ion
values
are actually averages over several detectors. The observed spread inthe
individual detector T
ion
interpretations indicates some motional
broadening contribution, which suggests that the lower temperature
is more representative of the thermal temperature. Throughout this
Letter, the uncertainty ranges given for values for all quoted quantities
are driven by the uncertainty in T
ion
.
For a 50:50 DT mix the fusion power density is _
DT
~7:04|
10
{13
n
2
sv h i in joules per cubic centimetre per second, where n is the
yet-unknown number density of the fusing region. FromE
fusion
, V
hs
and
N130927 equatorial view
Kapton channel
N130927 polar view
Kapton channel

Reconstructed shape for N130927
x, um
y

(

m
)
100 50 0 50 100
100
50
0
50
100
N130927 equatorial
neutron image
1317-MeV neutrons
612-MeV neutrons


100 50 0 50 100
100
50
0
50
100
Reconstructed shape for N131119
x (m)
z

(

m
)

y

(

m
)

y

(

m
)

z

(

m
)

z

(

m
)

x (m)
20
10
30
20
10
30
40
10
5
15
25
20
x (m)
z (m)
y (m)
0
10
20
30
30
20
10
0
20
20
0
20
20
x (m)
z (m)
y (m)
0
20
20
0
20
20
100
100
100
100
50
50
50
50
0
x (m)
100 100 50 50 0
x (m)
100 100 50 50 0
x (m)
100 100 50 50 0
0
100
100
50
50
0
100
100
50
50
0
100
50
50
0
z

(
z

m
)z (( z

x ( x m)
100 100 50 50 0
0000 100 0000
00 50 0
50 5
0000
N131119 equatorial view
Kapton channel
N131119 polar view
Kapton channel
a b c d
10
20
20
10
0
N131119 equatorial
neutron image
1317-MeV neutrons
612-MeV neutrons

100
10
5
15
25
20
Figure 2 | X-ray and neutron images of the hotspot at bang-time.
a, Equatorial (side-on) and polar (top-down) views of the hotspot shape for
N130927. Kapton is the filter material in the imaging system that allows
transmission of X-rays with energies of more than 6 keV. b, As in a, but for
N131119. In these X-ray images, the contour shown in white is taken at the
17%-peak-brightness level (the colour scales show the brightness in arbitrary
units) and is used to obtain a description of the shape in Legendre modes
(equatorial view) and Fourier modes (polar view). c, Three-dimensional
reconstructions of the hotspots. d, Superposition of direct (1317 MeV) and
down-scattered (612 MeV) neutron images from N130927 and N131119.
(X-ray image analysis courtesy of N. Izumi, S. Khan, T. Ma and A. Pak of the
NIF Shape Working Group; neutron image analysis courtesy of D. Fittinghoff,
G. Grim, N. Guler and F. Merrill of the NIF Neutron Imaging SystemWorking
Group.)
LETTER RESEARCH
0 0 M O N T H 2 0 1 4 | V O L 0 0 0 | N A T U R E | 3
Macmillan Publishers Limited. All rights reserved 2014
t
x
the hotspot number density can be calculated:
n~

E
fusion
7:04|10
{13
sv h iV
hs
t
x

For N130927, n58.1310


24
1.2310
25
cm
23
, a value that also pro-
vides the hotspot mass density (assuming a pure DT hotspot, with
average atomic mass number

A~2:5 for DT), r
hs
53450 g cm
23
;
the hotspot mass, m
hs
5r
hs
V
hs
56.49.4 mg; and the areal density,
(rr)
hs
(Table 1).
A number of quantities describing the implosion energetics now
straightforwardly follow. The hotspot pressure can be obtained from
P
hs
~

Zz1 r
hs
T
ion
=

A(

Z~1 for DT), yielding P


hs
5126152 Gbar.
The hotspot energy is then E
hs
~(3=2)P
hs
V
hs
(Table 1). The fraction,
f
a
, of a-particle energy deposited into the hotspot can be calculated
from a classic formula
23
f
a
~1{
1
4 rr
hs
_
rl
a
_ z
1
160 rr
hs
_
rl
a
_
3
where the a-particle stopping range can be found from
24
rl
a
~
0:025T
5=4
e
1z0:0082T
5=4
e
3
in base units of centimetres, grams, and kiloelectronvolts. For N130927,
f
a
50.680.82. The energy deposited in the hotspot by a-particles is
E
a
5f
a
E
fusion
/5, recalling that one-fifth of the DT 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, E
a
/E
hs
<0.56. These energies fully describe the hotspot, but part
of the implosion energy was used to compress the remaining cold DT
fuel andsowe must examine the fuel toget a full picture of the implosion
energy balance.
Because the DT hotspot is formed by ablating the inner surface of
the cold DT fuel as electron conduction transports heat from the
forming hotspot into the fuel, we can calculate the amount of DT
fuel remaining after the hotspot has formed because we know the
initial amount of DT ice layered onto the inside of the capsule,
m
0
5186 mg (for N130927): m
fuel
5m
0
2m
hs
5176179 mg. The cold
DT fuel mass forms a shell surrounding the hotspot with volume
V
fuel
~(4p=3) r
3
out
{r
3
hs
_ _
, where r
out
is the unknown outer fuel radius.
Because m
fuel
~4p
_
r
fuel
r
2
dr and the measured DSR provides a way
to obtainthe fuel density, r
fuel
, from rr
fuel
~
_
r
fuel
dr, by assuming a
fuel profile we can solve for both the fuel layer thickness, r
out
2r
hs
, and
density r
fuel
(Methods). We find that
r
out
{r
hs
~2s~

m
fuel
2p rr
fuel
{r
2
hs
_
{r
hs
4
with a Gaussian density profile
r(r)~(rr)
fuel
exp{(r{r
0
)
2
=2s
2

2p
p
s
where r
0
is the radius of peak fuel density. For N130927, r
out
2r
hs
5
14.715.3 mm, r
fuel
5385402 g cm
23
and V
fuel
5(3.03.2) 310
5
mm
3
.
The fuel outer radius from these arguments, r
out
550.8 mm (at 50%
r
fuel
), is close to that obtained directly fromthe down-scattered neutron
image (Fig. 2), where P0 555.4 mm (at 17% of maximum intensity). By
this time of peak compression, the DT fuel density has increased by a
factor of more than 1,500. The fuel density is not required for calculating
the fuel energy, but it can be used to estimate the adiabat of the fuel (at
bang-time) assuming that the cold fuel and the hotspot are isobaric
(P
fuel
<P
hs
), in which case we find that a~P
fuel
=P
F
<P
hs
=0:0021r
5=3
fuel
5
2.93.3 for N130927the fuel adiabat in flight is lower than this range of
values. The fuel density is alsoneededtocalculate 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 DT hotspot is optically thin to these X-rays. The bremsstrahlung
energy loss is calculated to be
24
E
brems
(kJ)~5:34|10
{34
n
2
hs

T
e
_
V
hs
t
x
inbase units of centimetres, kiloelectronvolts andseconds. For N130927,
E
brems
52.34.5kJ, the low end of which is nearly equivalent to the
0
5
10
15
20
Shot number
1
1
0
6
0
8
1
1
0
6
1
5
1
1
0
6
2
0
1
1
0
8
2
6
1
1
0
9
0
4
1
1
0
9
0
8
1
1
0
9
1
4
1
1
1
1
0
3
1
1
1
1
1
2
1
1
1
2
1
5
1
2
0
1
2
6
1
2
0
1
3
1
1
2
0
2
0
5
1
2
0
2
1
3
1
2
0
2
1
9
1
2
0
3
1
1
1
2
0
3
1
6
1
2
0
3
2
1
1
2
0
4
0
5
1
2
0
4
1
2
1
2
0
4
1
7
1
2
0
4
2
2
1
2
0
6
2
6
1
2
0
7
1
6
1
2
0
7
2
0
1
2
0
8
0
2
1
2
0
8
0
8
1
2
0
9
2
0
1
3
0
3
3
1
1
3
0
5
0
1
1
3
0
5
3
0
1
3
0
7
1
0
1
3
0
8
0
2
1
3
0
8
1
2
1
3
0
9
2
7
1
3
1
1
1
9
Y
i
e
l
d

(
k
J
)



0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8



S
e
l
f
-
h
e
a
t
i
n
g

y
i
e
l
d
/
c
o
m
p
r
e
s
s
i
o
n

y
i
e
l
d

0
1.0
0.5
0
1.5
GLC
Yield from fuel compression
Yield from self-heating
Energy delivered to DT fuel
Low foot
High foot
Low foot
High foot
N130710
N130812
N130501
N130927
N131119
Figure 3 | Yield and energetics metrics for shots on the NIF. Total fusion
yield is plotted versus shot number (that is, time). Shots 110608130331 are
low-foot shots. Shots 130501131119 are high-foot shots. The bars showing
total yield are broken into components of yield coming from a-particle self-
heating and yield coming from compression. The black dashes denote the
energy delivered to the DT (fuel plus hotspot) with error bars (black vertical
lines, 1s) as calculated fromthe model of ref. 25. The plot shows that, even with
the uncertainty in our results, shots 130927 and 131119 both yielded more
fusion energy than was delivered to the DT. Inset, ratio of self-heating yield to
compression yield versus generalized Lawson criterion (GLC). All error bars,
1s.
RESEARCH LETTER
4 | N A T U R E | V O L 0 0 0 | 0 0 M O N T H 2 0 1 4
Macmillan Publishers Limited. All rights reserved 2014
a-particle energydeposited. Toexamine whether or not these X-rays can
escape the dense cold fuel, we can calculate the optical depth of the cold
DT fuel from t
fuel
5r
fuel
k
DT
(r
out
2r
hs
) using a simple DT opacity
model, k
DT
(g cm
22
) 50.352r
fuel
(hn)
23.3
[1 2exp(2hn/T
e
)]. We find
that for X-ray energies of hn <T
ion
<T
e
, the DT fuel layer is almost
one optical depth, t
fuel
50.320.66, implying that some bremsstrahlung
X-rays deposit energy into the cold fuel whereas some escape. Electron
conduction does not have a significant role in the total DT energy loss
fromthe cold fuel at stagnation, but is important for the hotspot energy
loss.
The cold fuel energy at stagnation now follows from the isobaric
assumption, E
fuel
~(3=2)P
fuel
V
fuel
56.97.8 kJ (where we have over-
estimated the fuel internal energy because typically the outer edge of
the fuel has not fully stagnated even at bang-time). The total energy
delivered to the DT by the implosion is then (Table 1)
E
DT,total
~E
hs
zE
fuel
z
1
2
e
{t
fuel
E
brems
{
1
2
E
a
5
The factors of one-half in the radiation term and the a-particle energy
deposition account for having only half the energy emitted or depos-
ited at peak burn. This total DTenergy was calculated with quantities
measured around bang-time, but it represents the kinetic plus internal
energy in the fuel at peak velocity in the implosion. A crosscheck of
E
DT,total
is provided from calculating the fuel kinetic energy, K, using a
direct measurement of implosion velocity from an earlier high-foot
1DConA shot, N130409 (at 350 TW and 1.3 MJ of laser power and
energy), where the peak ablator centre-of-mass velocity was measured
to be 267 615 kms
21
, which is equivalent to a fuel velocity of
297 615 kms
21
(the fuel being at smaller radius and convergence
makes the velocity larger). Scaling the N130409-derived velocity to
the laser power of N130927 (implosion velocity, v
imp
!P
0:41
laser
) gives a
fuel velocity of v
fuel
5311615 kms
21
, and so K~(1=2)m
0
v
2
fuel
~9:0+
0:9 kJ. The difference betweenKand E
DT,total
is the internal energy of the
fuel at peak velocity plus the additional PdVwork done by the ablator on
the fuel during the deceleration.
The total fuel energy gain, G
fuel
5E
fusion
/E
DT,total
is now known and
is 1.21.4 for N130927. For comparison, in Table 1 we also show
results fromother data-derivedmodels of implosionenergetics
25,26
that
are constructed in the spirit of the above analysis but which differ in
some details. A conduction-limited temperature profile in the hotspot
is added to the above development in one case
25
and the other detailed
model case includes a three-dimensional, self-consistent physics
model matched to the data
26
. To complement these analytic data-
driven models, in Table 1 we also show the results from a full, one-
dimensional, radiationhydrodynamics simulation
27
of N130927, with
a multifrequency X-ray drive that is calibrated to shock-timing and
implosion trajectory data, without any mix model applied. The infer-
ences from data and the computer simulation all indicate that
G
fuel
.1. Moreover, we have demonstrated repeatability and improve-
ment with the follow-on shot N131119. It should be understood, how-
ever, that G
fuel
.1 indicates only that the output fusionenergy exceeds
the energy deposited into the fuel. This is not the same as exceeding
either the energy absorbed by the capsule (defined as the ablator shell
plus DTfuel), which absorbed ,150 kJ for N130927 and N131119, or
the energy delivered by the laser to the target (defined as the hohlraum
plus capsule), which was 1.8 MJ for N130927 and 1.9 MJ for N131119.
Key yieldand energy performance metrics are graphically illustrated
in Fig. 3 for N131119, N130927 and all other DT implosions carried
out on the NIF since the summer of 2011. Using a key metric for
ignition, the generalized Lawson criterion
28
x 5(Pt)/(Pt)
ign
(which
is unity at ignition) we see (Fig. 3, inset) that for N131119 we are at
the threshold of achieving yield doubling due to a-particle energy
deposition.
Because most of the quantities associated with inertially confined
fusionthat we seek toimprove toachieve ignitionscale as some positive
power of stagnationpressure, near-termefforts focus on increasing the
implosion speed and controlling the hotspot shape with the present
fuel adiabat. As the implosion speed is increased, we will necessarily
risk giving back some of the gains the high-foot implosion has made in
terms of instabilitycontrol. Newstrategies for the hohlraumwill alsobe
explored because at present hohlraum physics is limiting our ability to
use the full power capability on the NIF while maintaining an accept-
able hotspot shape (higher laser powers are the most direct way to
increase implosion speed). Future efforts may involve more elaborate
schemes to maintain control over ablator instability while recovering a
lower adiabat for the fuel (for example adiabat shaping
29
) or alsousing
an alternate ablator material.
METHODS SUMMARY
Formulae for the hotspot volume and the DT reaction rate, and a discussion of
fuel density profiles, are given in Methods. Neutron image shape coefficients are
also given there.
Online Content Any additional Methods, ExtendedData display items and Source
Data are available in the online version of the paper; references unique to these
sections appear only in the online paper.
Received 1 November 2013; accepted 7 January 2014.
Published online 12 February 2014.
1. Edwards, M. J. et al. Progress towardsignitionontheNational IgnitionFacility. Phys.
Plasmas 20, 070501 (2013).
2. Dittrich, T. R. et al. Design of a high-foot/high-adiabat ICF capsule for the National
Ignition Facility. Phys. Rev. Lett. 112, LJ14108 (2014).
3. Park, H.-S. et al. High-adiabat, high-foot, inertial confinement fusion implosion
experiments on the National Ignition Facility. Phys. Rev. Lett. 112, LK13998
(2014).
4. Lindl, J. D. & Moses, E. I. Plans for the National Ignition Campaign (NIC) on the
National Ignition Facility (NIF): on the threshold of initiating ignition experiments.
Phys. Plasmas 18, 050901 (2011).
5. Glenzer, S. H. et al. Cryogenic thermonuclear fuel implosions on the National
Ignition Facility. Phys. Plasmas 19, 056318 (2012).
6. Bodner, S. E. Rayleigh-Taylor instability and laser-pellet fusion. Phys. Rev. Lett. 33,
761764 (1974).
7. Gonchorov, V. N. & Hurricane, O. A. Panel 3 Report: Implosion Hydrodynamics.
Report LLNL-TR-562104 (Lawrence Livermore National Laboratory, 2012).
8. Ma, T. et al. Onset of hydrodynamic mixinhigh-velocity, highly compressedinertial
confinement fusion implosions. Phys. Rev. Lett. 111, 085004 (2013).
9. Regan, S. P. et al. Hot-spot mix inignition-scale inertial confinement fusiontargets.
Phys. Rev. Lett. 111, 045001 (2013).
10. Betti, R., Goncharov, V. N., McCrory, R. L. & Verdon, C. P. Growth rates of the
Rayleigh-Taylor instability in inertial confinement fusion. Phys. Plasmas 5,
14461454 (1998).
11. Lindl, J. Development of the indirect-drive approachto inertial confinement fusion
and the target physics basis for ignition and gain. Phys. Plasmas 2, 39334024
(1995).
12. Haan, S. et al. Point design targets, specifications, and requirements for the 2010
ignition campaign on the National Ignition Facility. Phys. Plasmas 18, 051001
(2011).
13. Michel, P. et al. Tuning the implosion symmetry of ICF targets via controlled
crossed-beam energy transfer. Phys. Rev. Lett. 102, 025004 (2009).
14. Michel, P. et al. Symmetry tuning via controlled crossed-beamenergy transfer on
the National Ignition Facility. Phys. Plasmas 17, 056305 (2010).
15. Moody, J. D. et al. Multistepredirectionby cross-beampower transfer of ultrahigh-
power lasers in a plasma. Nature Phys. 8, 344349 (2012).
16. Callahan, D. A. et al. The velocity campaign for ignition on NIF. Phys. Plasmas 19,
056305 (2012).
17. Marinak, M. M. et al. Three-dimensional HYDRA simulations of National Ignition
Facility targets. Phys. Plasmas 8, 2275 (2001).
18. Kozioziemski, B. J. et al. Deuterium-tritiumlayer formationfor the National Ignition
Facility. Fusion Sci. Technol. 59, 1425 (2011).
19. Glebov, V. Yu. et al. Development of nuclear diagnostics for the National Ignition
Facility. Rev. Sci. Instrum. 77, 10E715 (2006).
20. Bleuel, D. L. et al. Neutron activation diagnostics at the National Ignition Facility.
Rev. Sci. Instrum. 83, 10D313 (2012).
21. Gatu Johnson, M. et al. Neutron spectrometry an essential tool for
diagnosing implosions at the National Ignition Facility. Rev. Sci. Instrum. 83,
10D308 (2012).
22. Bosch, H.-S. &Hale, G. M. Improved formulas for fusion cross-section and thermal
reactivities. Nucl. Fusion 32, 611631 (1992).
23. Krokhin, O. N. & Rozanov, V. B. Escape of a-particles froma laser-pulse-initiated
thermonuclear reaction. Sov. J. Quantum Electron. 2, 393394 (1973).
24. Atzeni, S. & Meyer-ter-Vehn, J. The Physics of Inertial Fusion 32, 398 (Oxford Univ.
Press, 2004).
25. Patel, P. et al. Performance of DTlayeredimplosions onthe NIF. Bull. Am. Phys. Soc.
58, abstr. NO4.00001 (2013).
LETTER RESEARCH
0 0 M O N T H 2 0 1 4 | V O L 0 0 0 | N A T U R E | 5
Macmillan Publishers Limited. All rights reserved 2014
26. Cerjan, C., Springer, P. T. & Sepke, S. M. Integrated diagnostic analysis of inertial
confinement fusion capsule performance. Phys. Plasmas 20, 056319 (2013).
27. Zimmerman, G. B. & Kruer, W. L. Numerical Simulation of laser initiated fusion.
Comments Plasma Phys. Control. Fusion 2, 8589 (1975).
28. Betti, R. et al. Thermonuclear ignition in inertial confinement fusion and
comparison with magnetic confinement. Phys. Plasmas 17, 058102 (2010).
29. Goncharov, V. N. et al. Improved performance of direct-drive ICF target designs
with adiabat shaping using an intense picket. Phys. Plasmas 10, 19061918
(2003).
Acknowledgements We thank P. Albright, J. Atherton, L. R. Benedetti, D. Bradley,
J. A. Caggiano, R. Dylla-Spears, M. J. Edwards, W. H. Goldstein, B. Goodwin, S. Haan,
A. Hamza, W. Hsing, P. Kervin, J. Kilkenny, B. Kozioziemski, O. Landen, J. Lindl,
B. MacGowan, A. Mackinnon, N. Meezan, J. F. Meeker, J. Moody, E. Moses, D. Pilkington,
T. Parham, J. Ralph, S. Ross, H. Robey, R. Rygg, B. Spears, R. Town, C. Verdon, A. Wanand
B. Van Wonterghem, and the NIF operations, cryogenics and targets teams. We also
thankV. Goncharov andJ. Knauer for their advice, andR. Betti for bringingour attention
to equation(3). Thanks also go to NIFs external collaborators at GA (targets), LLE
(diagnostics), the MIT Plasma Science and Fusion Center (magnetic recoil
spectrometer diagnostic), CEA and AWE. This work was performed under the auspices
of the US Department of Energy by Lawrence Livermore National Laboratory under
contract no. DE-AC52-07NA27344.
Author Contributions O.A.H. was lead scientist for the high-foot campaign, and
performed two-dimensional stability modelling, and one-dimensional pre- and
post-shot analysis. D.A.C. was leadscientist onhotspot shape andhohlraumstrategies.
D.T.C. was part of the DT shot experiment team. P.M.C. performedVISARdata unfolds.
C.C. performed three-dimensional detailed model calculations. E.L.D. was lead
experimenter for 1DConA(R(t) trajectory) tuningexperiments andcapsulere-emission
(early-time symmetry) tuning experiments. T.R.D. performed initial one-dimensional
capsuledesign, scoping, andone-dimensional pre- andpost-shot simulations. T.D. was
lead experimentalist on a 2DConA ablator shape experiment and was part of the DT
shot team. D.E.H. was the pulse shape design physicist and performed all
two-dimensional integrated hohlraum-capsule simulations. L.F.B.H. was design
physicist for keyhole (shock-timing) tuning experiments. J.L.K. was lead
experimentalist for symcap (hotspot shape) tuning experiments. S.L. was lead
experimentalist for the keyhole (shock-timing) tuning experiments. T.M. was lead
experimentalist for several 2DConA ablator shape experiments, was part of the DT
shot team, and was lead experimentalist on shot N131119. A.G.M. was part of the
1DConA and DT experimental teams. J.L.M. was design physicist for the re-emission
experiment. A.P. was part of the DT shot team. P.K.P. provided a hotspot model
analysis andmetrics plots. H.-S.P. was lead experimentalist on DT implosion shots up
to and including N130927. B.A.R. was overall lead on experiments. J.D.S. constructed
model multifrequency sources normalizedto tuning experiments, andperformedone-
and two-dimensional model scoping. P.T.S. provided a hotspot model analysis. R.T.
provided 1DConA analysis and was shot experimentalist.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial interests.
Readers are welcome to comment on the online version of the paper. Correspondence
and requests for materials should be addressed to O.A.H. (hurricane1@llnl.gov).
RESEARCH LETTER
6 | N A T U R E | V O L 0 0 0 | 0 0 M O N T H 2 0 1 4
Macmillan Publishers Limited. All rights reserved 2014
METHODS
Hotspot volume formulae. Using equation(2) the volume is
V
hs
~2p
_
p
0
_
Rhs h
0
R
2
dRsinh dh
~
4
3
pP0
3
z
4
5
pP0P2
2
z
8
105
pP2
3
z
4
7
pP0P3
2
z
16
105
pP2P3
2
z
4
9
pP0P4
2
z
8
35
pP2
2
P4
z
80
693
pP2P4
2
z
8
77
pP3
2
P4z
24
1,001
pP4
3
z
A simple correction to this volume can be applied (by multiplying the above
expression by the expression below) to include m-modes (azimuthal modes):
1z
3
2
M2
M0
_ _
2
z
3
2
M3
M0
_ _
2
z
3
2
M4
M0
_ _
2
z
3
4
M2
2
M4
M0
3
cos 4 w
2
{w
4

where w
2
and w
4
are phase offsets of m-modes 2 and 4, respectively.
Deuteriumtritium reaction rate formulae. From ref. 22
sv h i~C
1
f
{5=6
j
2
exp {3f
1=3
j
_ _
in cubic centimetres per second, where
j~
C
0
T
1=3
ion
f~1{
C
2
T
ion
zC
4
T
2
ion
zC
6
T
3
ion
1zC
3
T
ion
zC
5
T
2
ion
zC
7
T
3
ion
and C
0
56.6610, C
1
56.4341310
214
, C
2
51.5136310
22
, C
3
57.5189310
22
,
C
4
54.6064310
23
, C
5
51.35310
22
, C
6
521.0675310
24
and C
7
51.366310
25
when T
ion
is expressed in kiloelectronvolts.
Deuteriumtritium fuel density profile. Assuming a different density profile
changes the form of equation(4), but changes the numerical value for the fuel
thickness little. For example, assuming a top-hat distribution for the fuel yields
r
out
{r
hs
~
1
2

3m
fuel
_
p rr
fuel
{3r
2
hs
_
{3r
hs
=2
from which we obtain r
out
2r
in
515.516.2 mm for N130927. The fuel density
does showmore sensitivity, being r
fuel
5(rr)
fuel
/(r
out
2r
in
) 5457478 g cm
23
for
a top-hat distribution and r
fuel
~ rr
fuel
_
2p
p
s 5385402 for a Gaussian. The
Gaussian profile assumption is more consistent with simulated fuel density pro-
files. The lower fuel density associated with the Gaussian profile increases the
inferred fuel adiabat and decreases the fuel optical depth as compared with the
uniformprofile. The lower fuel optical depth makes the X-ray energy contribution
to equation(5) larger; that is, it gives us a more conservative contribution to the
total DT energy.
Neutron image shape analysis. For N130927, the Legendre mode shape coeffi-
cients for the down-scattered neutron image are P0 555 64 mm, P2/
P0 51%65% and P4/P0 522%, and for the direct image P0 532 64 mm, P2/
P0 5235%65% and P4/P0 52%. For the N131119 down-scattered neutron
image, P0 550 64 mm, P2/P0 50%65% and P4/P0 52%, and for the direct
image P0 534 64 mm, P2/P0 5234%65% and P4/P0 51%.
LETTER RESEARCH
Macmillan Publishers Limited. All rights reserved 2014

You might also like