Analysis of a Direct-Drive IgnitionCapsule Design for the National Ignition Facility

42nd Annual Meeting of theAmerican Physical SocietyDivision of Plasma Physics

Qu�bec City, Canada23–27 October 2000

P. W. McKenty et al.University of RochesterLaboratory for Laser Energetics

40

30

20

10

00 1 2

��(�m)

0

200

400

600

800

1000

R(�

m)

Z (�m)

–500 –250 5002500

End of acceleration phase

7.5

3.5

0.5

�(g/cc)

Gai

n

�2 = 0.06 �1–9 + �>102 2

Analysis of a Direct-Drive Ignition Capsule Designed for the NIF

P. W. McKenty

Laboratory for Laser Energetics, U. of Rochester

The current direct-drive ignition capsule design planned by the University of Rochester’s Laboratory for Laser

Energetics to be fielded on the National Ignition Facility (NIF) will be reviewed in this paper. The direct-drive

requirements to establish a propagating thermonuclear burn on the NIF will be discussed in terms of the constraints

on laser-irradiation uniformity and target surface roughness. The ignition design1,2 consists of a cryogenic DT

shell (~350 µm thick and ~3 mm in diameter) contained within a very thin (~2-µm) CH shell. To maintain stability

during the implosion, the target is placed on an isentrope approximately three times that of Fermi-degenerate DT

(α = 3). One-dimensional hydrodynamic studies using LILAC show that the ignition design is robust to

uncertainties in laser power history and fuel composition. The two-dimensional hydrodynamics code ORCHID is

used to examine the target performance under the influence of the main sources of nonuniformity: laser imprint,

power imbalance, and inner- and outer-target-surface roughness. Results from these studies indicate that the

reduction in target gain from all sources of nonuniformity can be described in terms of a single parameter related to

the resultant inner-surface deformation at the end of the acceleration stage of the implosion. This parameter is

constructed from a spectral decomposition of the surface deformation by giving different weights to the long and

short wavelengths of nonuniformity. The physical reason for the difference in weighting is discussed in terms of

the mechanisms for ignition failure. This work was supported by the U.S. Department of Energy Office of Inertial Confinement

Fusion under Cooperative Agreement No. DE-FC03-92SF19460.

1. C. P. Verdon, Bull. Am. Phys. Soc. 38, 2010 (1993).2. S. E. Bodner et al., Phys. Plasmas 5, 1901 (1998).

Collaborators

V. N. Goncharov

R. P. J. Town

S. Skupsky

R. Betti

R. L. McCrory

J. A. Delettrez

D. L. Harding

M. Wittman

R. L. Keck

TC5589

Scaling target gain with ��provides the basis fordeveloping a global nonuniformity budget forthe NIF direct-drive point design

• Results from the nonuniformity budget, accounting for allfour sources of nonuniformities, indicate that direct-drivetargets can achieve gains in excess of 30 using current NIFspecifications and the deployment of SSD with two color cycles.

• Outer surface roughness does not make a significantcontribution to the nonuniformity budget.

• Distortions at stagnation are dominated by low ordermodes, however, high order modes cannot be neglected.

• Gain reduction is caused by target nonuniformities delayingthe onset of ignition thereby wasting margin of thestagnating fuel layer.

Summary

TC5484

Outline

� Point design

� Numerical modeling

� Sources of implosion nonuniformities

– power balance

– ice/vapor surface roughness

– outer surface roughness

– laser imprint

� Failure analysis

� Nonuniformity budget

� Summary

TC5576

The “all-DT” direct-drive target is a thickDT-ice layer enclosed by a thin CH shell

1.5 MJ, � = 3Gain: 45Yield: 2.5 � 1019

�rpeak: 1300 mg/cm2

<Ti>n: 30 keVHot spot CR: 29Peak IFAR: 60

1.69 mm

1.35 mm

CH

DT ice

DT gas

103

Po

wer

(T

W)

Time (ns)

0 2 4 6 8 10

102

101

100

Gain

45

35

25

15

5

R. Town, LLE Review, , 121 (1999).79

Point design

TC2802a

Simulations show that low-order modes canreduce high-gain capsule performance if theperturbation amplitudes are large

Illumination perturbation amplitudePeak-to-valley (%)

Yie

ld (

per

turb

atio

n)

/yie

ld (

un

iform

)

0 1 2 3 4

0.6

0.8

0.4

1.0

� = 16� = 8

� = 4

� = 2

Numerical modeling

�2 � 0.06�1�92 � ��10

2

TC5137

To relate the gain reduction to the mode spectrum,a series of 2-D ORCHID simulations with perturbedinner DT-ice interface has been performed

Mode number

101100 102

101

100

10–1

Am

plit

ud

e (µ

m)

Mode number

10–2101100 102

Initial-modespectrum

End ofacceleration

0 1 20

10

20

30

40G

ain

� (µm)

• �� ��0��

� = 1.5

� = 0

TC5316a

There are four sources of perturbations a direct-drive capsule must tolerate to ignite and burn

Target fabrication issues Laser irradiation issues

Outsidecapsulefinish

Inner DT iceroughness

Drivesymmetry

Laserimprinting

Sources of implosion nonuniformities

� �2�rmslaser imprinting

Max allowedvaluelaser imprinting�

�rmsdrive symmetry

Max allowedvaluedrive symmetry� �

2

�rmsDT ice

Max allowedvalueDT ice� �

2 �rmsoutside capsule finish

Max allowedvalueoutside capsule finish� �

2

�

Heuristically, there are four sources of perturbations adirect-drive capsule must tolerate to ignite and burn

TC4113f

Laser-irradiation-related issuesTarget-fabrication-related issues

� < 1

TC5495

NIF temporal power histories are mapped to targetand spherically decomposed for input into ORCHID

NIF laser power histories provided by Ogden Jones (LLNL)

Perfect beam timing40-ps rms mistiming

0 642 8 10

6

4

2

8

10

0

Time (ns)

Po

wer

imb

alan

ce(%

rms)

30

Laser energy(quad to quad)

Energyon target

Laserpulse

Las

erp

ow

er(a

rbit

rary

un

its)

2% rmson target

1.5

192 NIFBeamletpulses

6

4

2

8

10

00 5 10

Time (ns)

0.00

8% rmsbeam to beam

Mode number

Rm

s (%

)

Power balance

TC5563

Results of ORCHID calculations have validatedthe direct-drive base-line power imbalance specifications

Perfect beam timing30-ps mistimingbetween beams

Initial on targetperturbation (% rms)

4

45

30

15

0

Gai

n

0.0 1.00 2 6 8

NIFIndirect-drive

baseline

2.0 2.50.5 1.5��(�m)

TC5317a

Perturbations located initially on the inside DT-icesurface affects the capsule implosion during boththe acceleration and deceleration phases

Time

Initial capsule

Feedout ofperturbation

to the ablationsurface

Ablation surfacegrowth and feed-

through to theinner surface

during theacceleration

phase

Growth of the hotspot–main fuellayer interface

during thedeceleration

phase

InsideDT-icesurface

Ice/vapor surface roughness

TC5211

ORCHID simulations indicate that target gain dependsstrongly on the development of the low-order modes

Fourier mode number

2 25 50 75

Ice/vapor interface spectra

0.01

0.10

1.00

10.0

Initial spectrum

Final spectrum

Per

turb

atio

n a

mp

litu

de

(µm

)�c = 95% �tot

Gain = 1.75

300

250

200

150

100

50

0–100 0 100

R (�m

)

Z (�m)

� (g/cc)

0

80

160

TC5564

Scaling target gain with ��correctly balances theindividual contributions of the low and high ordermodes during the implosion.

a = a0/l�, � = 0( ), 0.75( ), and 1.50( )

3210��(�m)

45

30

15

0

Gai

n

3210��(�m)

(modes 2–10)

4

TC5565

The smoothness and concentricity of thin-wallpolyimide targets (1.5 to 2.0 �m) have been improved

� High-frequency roughness is a consequence of the coating process;the rms value � > 100 is < 40 nm.

� Low-frequency roughness is caused by weak shells deforming toaccommodate the bending moments that develop during processing.

� Inflating the shell reduces the low-frequecy roughness.

Mode number1

10–1

105

104

103

102

101

100

10 100 1000

nm

2NIF indirect-drive standard

Filled to~90 psi

Empty

Early shell (9/99)

Outer surface roughness

TC5455

Twice the NIF standard specification (~165 nm) forouter surface roughness does not result in anysignificant disruption of the ice/vapor interface

1 1010–4

10–3

10–2

10–1

100

100

Po

wer

spec

tru

m(�

m2 )

Legendre mode number

1000

Z (�m)

0

200

400

600

800

1000

R(�

m)

–500 –250 5002500

Surface spectra

Ablation (0.0 ns)Ablation (8.8 ns)Ice/vapor (8.8 ns)

t = 8.8 ns 9.0

5.3

1.6

�(g/cc)

����0.15 �m

tc* � ��sin k� 2� �� � 1

TC5220a

SSD reduces time-averaged laser nonuniformity

*S. Skupsky, Phys. Plasmas 6, 2157 (1999).

0 20 40 60 80 1000

20

40

60

80

100

< t>

��: bandwidth�: speckle size

rms

no

nu

nif

orm

ity

(%)

rms ~ tc t rms0

Focus lensPhaseplate

Targetplane

Laser imprint

� DPP spectrum� Beam Overlay� PS

TC5587

Application of SSD bandwidth is necessaryfor shell integrity during the entire implosion

Z (�m)

0

200

400

600

800

1000

R(�

m)

–500 –250 5002500

M CP D TD T S u pM CP D TD T S SD TW pM CP D TD T S u pM CP D TD T po nd gn NF m p o pow m b n p d

Z (�m)

–500 –250 5002500

No bandwidth Bandwidth = 1 THz

7.5

3.5

0.5

��(g/cc)7.5

3.5

0.5

��(g/cc)

TC5588

Scaling gain with , taken from ORCHID calculations,indicates that NIF must deploy at least 1-THz bandwidth

45

30

15

0

Gai

n

SSD bandwidth(THz, 1 color cycle)

0.0 1.51.0

2 colorcycles

�

��(�m)

2.00.5 0 321

NIFDirect-drivespecification

ProjectedORCHID results2-D

TC5545

To achieve high gain for NIF capsules ignitionmust occur while the cold fuel still retains a significantfraction (margin) of its peak kinetic energy

Graph taken from Levedahl and Lindl, Nucl. Fusion 37 (2), 170 (1997).

Fu

el b

urn

-up

fra

ctio

n

1

10–1

10–2

10–3 v = 3.0 � 107 cm/sv = 4.5 � 107 cm/s

1 10(Pusher kinetic energy/10 kJ)(�/2)1.7 (vimp/3 ��107 cm/s)–5.5

Plasmablowoff

Ignitedhot spot

Implodingcold fuel (DT ice)

Point designG = 45 fb = 0.13margin = ~ 40%

Gain thresholdG = 1 fb = 0.003margin = ~ 10%

Failure Analysis

G = 1

TC5580

Shell stagnation determines the margin trajectorythat defines the window for high gain

100

75

50

25

0

Mar

gin

(%)

Ignition timing (ps)

0–200

Highgain

Non-burningcapsule

Ignitioncapsule

G = 45

200

TC5544

ORCHID simulations indicate that as ice/vaporinterface perturbations increase, ignition is delayedand gain is reduced

30

20

10

40

50

Ign

itio

nd

elay

(ps)

Mar

gin

(%)

Gai

n

60

40

20

80

00.0 1.51.00.5 2.0 2.5

0

30

20

10

40

50

030200

Margin (%)

2-D ORCHID results

10 40 50��(�m)

TC5577

Scaling gain with allows forming a global nonuniformitybudget for the direct-drive point design

��(�

m)

�Nonuniformity budget

Sum-in-quadrature

Multiplier

0.0 1.51.00.5 2.0

40

20

00 1 2

0

1

2

Applied SSDbandwidth (� 1 THz)

On-target powerimbalance (� 2% rms)

Inner-surfaceroughness (� 1-�m rms)

Outer-surfaceroughness (� 80 nm)

Current NIFspecifications

Two colorcycles

��(�m)

TC5589

Scaling target gain with ��provides the basis fordeveloping a global nonuniformity budget forthe NIF direct-drive point design

• Results from the nonuniformity budget, accounting for allfour sources of nonuniformities, indicate that direct-drivetargets can achieve gains in excess of 30 using current NIFspecifications and the deployment of SSD with two color cycles.

• Outer surface roughness does not make a significantcontribution to the nonuniformity budget.

• Distortions at stagnation are dominated by low ordermodes, however, high order modes cannot be neglected.

• Gain reduction is caused by target nonuniformities delayingthe onset of ignition thereby wasting margin of thestagnating fuel layer.

Summary/Conclusion