presented by: Jeff Latkowski

1
presented by Jeff Latkowski contributors D.T.
Blackfield, T.K. Fowler, W. Meier, D. Hewett, S.
Reyes, and L. J. Perkins (LLNL) D. Welch , T.
Hughes and C. Mostrom (MRC) W. Gekelmann and M.
VanZeeland (UCLA) T. Marshall et al.
(INEEL) High Average Power Laser
Meeting December 5-6, 2002 Work performed under
the auspices of the U. S. Department of Energy by
Lawrence Livermore National Laboratory under
Contract W-7405-Eng-48.
2
LLNL chambers work

• Magnetic deflection
• Safety environment
• Fast ignition
• Systems modeling
• Molecular dynamics simulations for graphite

3
Magnetic deflectionWhy needed?

• Original SOMBRERO used 500 mTorr of Xe gas and
  had relatively soft ion output
• Current understanding of target heating during
  injection limits gas to 25-50 mTorr
• Energetic ions will reach chamber wall and/or
  optics
• Multiple radiation damage issues, but
  exfoliation alone sufficient to cause serious
  problems could result in loss of 2 mm/h
• Enhanced ion stopping in plasma still likely to
  be significant ion fluxes at chamber wall

4
Magnetic deflectionGeneral principles
The unmagnetized ions pull electrons across the
magnetic field by a strong electric field,
producing an E X B current that expels the
background magnetic field.
Initally, a hot, dense plasma with unmagnetized
ions and magnetized electrons.
Radial expansion continues until the excluded
magnetic energy is roughly equal to the expanding
plasmas total kinetic energy. This radius is
called the magnetic confinement or Bubble
Radius.
The plasma continues to diffuse down the field
lines and the magnetic field relaxes to its
original state.
With a background plasma present this is
modified, because the expanding ions are
neutralized by background electrons being pulled
down field lines, possibly reducing the
laser-produced plasmas diamagnetism and
generating currents and waves.
5
Magnetic deflection Analytic calculationsand
review of past literature

• IFE chamber protection studied at LANL
  (1974-1980)
• Energy recovery in D-3He chamber at Kyushu
  University (1991-1993)
• Experimental work at United Aircraft Research
  Laboratory (1970) Pharos II III Laser facility
  at NRL (1983-1990) KI-1 facility in Novosibirsk
  (1987-2002) LLNL (1991)
• T.K. Fowler examined plasma expansion in a
  magnetic field (Starfish, 1966) Provides
  estimates on bubble radius, jet formation and
  chamber clearance time (May-June 2002)

6
Magnetic deflectionCurrent concept
Early time
Late time
7
Magnetic deflectionCurrent concept, (Contd.)

• E is plasma kinetic energy
• B0 is background magnetic field
• Assume high-yield direct-drive target (400 MJ
  yield with 110 MJ in ions)
• Allow mean bubble radius to extend 2/3 of the
  way to the chamber wall (4.3 m), we require
• B0 0.9 T
• B0 0.6 T for flux conserving wall

8
Magnetic deflection Summaryof results from T.K.
Fowler

• Species E (MJ) Rb(m) RMIN(m) RMAX(m)
• All 112 4.4 1.54 3.5
• H 1.85 1.15 0.40 0.88
• D 37.1 3.04 1.07 2.44
• T 42.7 3.20 1.12 2.55
• He 28.7 2.79 0.98 2.23
• C 1.56 1.06 0.37 0.88
• Rb (6E/Bo2)1/3 RMIN 0.35Rb RMAX 0.8Rb

RMAX assumes thin shell compresses field to 1.5Bo
RMIN uses special virial theorem and moment of
inertia to determine average radius accounting
for plasma instability, radiation, and
snow-plowing of background gas
Ref. ONeil and Fowler, Phys. Fluids. 9 (11)
(1966) p. 2219
9
Magnetic deflection Summaryof Fowler results
(Contd.)

• Worse case jetting (at 0.9 T) would contain lt14
  of total ion energy
• Chamber clearing could be an issue
• If orbital bounce motions keep loss cone full,
  plasma recompression must be 5.6? before chamber
  clearing time falls to 0.1 s
• If ion collisions are needed to fill the loss
  cone
• tclear 10 tii tii 1016T3/2 / n n
  1019(Rb/R)3 during recompression
• T lt 150 keV for tclear 0.1 s (requires
  thermalization and some entrainment of Xe plasma,
  which is formed by prompt x-rays)
• Raising field reduces need for recompression and
  thermalization due to smaller mirror ratio
• Still worried about charge exchange needs to be
  more carefully addressed

10
Magnetic deflection Significant effort hasgone
into continual development of LSP

• 3D PIC calculations with 30-cm-radius plasma at
  very high (uniform) density ? conditions severely
  stressed the code
• Switch to 2D calculations resulted in numerical
  instabilities produced by transition of electrons
  from a fluid to kinetic description
• Switch to a 2D description with kinetic ions and
  fluid electrons produced a more stable code, but
  resulted in energy losses (fictional plasma
  cooling)
• LSP has now been configured (by MRC) to allow for
  fluid and kinetic ions as well as electrons

11
Magnetic deflection Brief discussionof
numerical issues with LSP

• LSP has never operated with such dense, hot
  plasmas in meter-scale geometries
• Presence of magnetic field is not the cause of
  numerical problem
• Use of implicit solver can lead to lack of energy
  conservation
• Need to resolve plasma skin depth at
  plasma/vacuum interface results in too small
  cell sizes and too many particles (plasma skin
  depth 0.01 cm)
• Kinetic electrons require too short a time step ?
  cDt lt re ? Dt lt 2?10-4 ns wpeDt lt 1 ? Dt lt
  2.5?10-4 ns

12
Magnetic deflection 2D x vs. y LSP plasma
simulation rchamber 3.5 m Bo 0.8 T, (Contd.)
H only
13
Magnetic deflection 2D x vs. y LSP plasma
simulation rchamber 3.5 m Bo 0.8 T
All species
14
Magnetic deflection 2D x vs. y LSP plasma
simulation rchamber 3.5 m Bo 0.8 T
Cats Eye Nebula (NGC 6543)
15
Magnetic deflection Energy loss with
kineticions 2D x vs. y (rchamber 3.5 m Bo
0.8 T)
Field Energy
Total Energy
Energy loss reduced with fluid treatment for
ions valid?
D Energy
H Energy
Time (ns)
Time (ns)
16
2D r vs. z Lsp plasma simulation
rchamber 3.5 m Bo 0.8 T
All species
Electrons
D
H
17
Magnetic deflection Plasma expansion experiments
at LAPDU
Parameters n 1-4 ? 1012 cm-3 B 0.5-4
kG Plasma Radius 25 cm Plasma Length 17 m
LAPDU Large Plasma Device Upgrade
http//plasma.physics.ucla.edu/bapsf
18
Magnetic deflection Laser plasmaexperiments on
LAPDU

• Spectra Physics Nd-Yag Quanta Ray Pro Laser 1.06
  mm 1.8 J ( 1.5 J) tpulse lt 8 ns 0.5 1.0 Hz
• Target diameter of 0.5mm ? 1011W/cm2
• Targets are 0.5 0.75 ? 1.0 long rods of Al,
  C, or Ba
• NAl 2?1015 Vo 1.5?107cm/s Vo 1?107cm/s
• 0.75 J coupled to target EAl ions 3-4keV
  Ee10-20eV some fast electrons with E lt 100 eV
  Zeff 2
• 1.4 cm (0.4T) lt RbAl lt 26 cm (0.005T)
• 5.3 cm (0.4T) lt RhAl lt 424 cm (0.005T)
• 0.26 (0.4T) gt ebAl gt 0.06 (0.05T)

19
Magnetic deflection IFE-relevantexperiments on
LAPDU

• Key parameter
• eb Rh / Rb
• For IFE ? 7.9 cm / 429 cm .018
• With Al target, LAPDU can achieve 0.26 lt eb lt
  0.06 Key is that value is ltlt1
• Previous work has investigated bubble radius vs.
  magnetic field strength, including effects from
  low-density plasmas
• We would like to explore the effects of higher
  background plasma densities and the rate of
  expansion along the magnetic axis
• Would use experiments to (1) investigate
  IFE-relevant parameter space, and (2) benchmark
  LSP models

Rb (3moEo/2pBo2)1/3 Rh Vo / wci
20
Magnetic deflection LAPDU dataon bubble radius
vs. magnetic field

• There appears to be little difference with a
  background plasma of 1012 cm-3
• At the time of peak diamagnetism, the laser
  plasma is O(1013cm-3)
• For IFE chamber
• Background plasma could be gt1015 cm-3
• At stagnation, density in plasma bubble lt1013 cm-3

The high ratio of the background to bubble plasma
densities (gt102) in the IFE case suggests that we
need to investigate higher background plasma
densities
21
Magnetic deflection Additionalcalculations and
analyses are planned

• 2D calculations, although slow, will be the focus
  of our LSP work
• 3D calculations deferred until new platform
  becomes available (MCR has 2300 CPUs and 4600 GB
  memory)
• Will be moving to non-uniform cases should aid
  chamber clearing
• IFE-relevant plasma expansion data will be
  collected at LAPDU
• Rate of plasma leakage along magnetic field axis
  (chamber clearing)
• Rb vs. B0 for higher background plasma densities
• Non-uniform fields to be studied

22
SE INEELs FSP responded to concernsabout
graphite reaction rate data
23
SE With the new data, an excursion in SOMBRERO
is much less of a concern

• Excellent example of how process is supposed to
  work
• Analysis identifies issue
• Capabilities improved to confirm result
• Experiment conducted using newly available
  material
• New analysis performed to address issue

24
SE Scaled SOMBRERO accident doseresults for
average weather
Without oxidation, SOMBRERO has good chance of
meeting 10 mSv goaleven for conservative weather.
Xe scrubbed to remove Cs, I isotopes
25
Substantial progress is being madewith fast
ignition experiments

• Fuel assembly experiments on Omega and Z,
  modeling/design for NIF concept
• Heating/transport
• Planar experiments have elucidated transport
  physics
• Cone appears to concentrate and facilitate
  energytransport
• Proton heating identified as possible alternative
• Optimistic results from ILE Osaka cone
  targetexperiment 1000? increase in D-D neutron
  yield,heating to 1 keV, 30 efficient
• Improvements in laser technology
• Advanced front ends allowing higher fidelity
  ignitor pulses
• Damage resistant compressor gratings that are
  scalable to larger (1m) size
• Petawatt ignitor additions compatible with
  existing facilities (Omega, Z, NIF)

26
FI poses new challenges for laser final optics
layout and protection

• Compression beam requirements similar to hot spot
  ignition
• (but may not require uniform illumination)
• However, petawatt ignitor beams require
  development of high energy, short-pulse
  compatible gratings and focusing optics
• Need to develop an appropriate solution for FI
  final optics layout number of beams, size,
  stand-off distance

• Critical issues that need to be addressed
• potential for directional target output (for
  cone-focused design)
• cone-focused design would have x-ray and debris
  output rivaling indirect-drive designs
• - optics damage threshold for high intensity
  laser
• - optimum stand off-distance compatible with spot
  size requirement

27
New data on laser damage of multilayerdielectric
optics provides optimism for FI
MLD mirrors are being developed for HEPW (better
LIDT, h, wavefront, and scalable to IFE)

• (A) 2000 LULI MLD grating was 2? better than
  gold gratings for 0.5 ps pulse
• (B) Recent LULI Intermediate dispersion gratings
  with low field enhancement (1.2) from GA are
  encouraging for gt3 J/cm2 operation

28
Cone focusing may be able to ease spot size
requirements

• Sentoku observed focusing of ignitor beam due to
  reflection from oblique cone surface (about
  factor of 4 for 60º cone)
• Could relax the spot size requirement from 30 mm
  to gt300 mm
• This would make FI final optic survival easier by
  increasing the stand-off distance

Need to measure and model the concentration of
electron energy for laser radiation obliquely
incident at relativistic intensities on the inner
surface of the cone
29
Summary future work

• Magnetic deflection effort to be split between
  calculations and experiment (fielded on LAPDU)
• 2D expansions with LSP LSP benchmarking (energy
  losses)
• LSP modeling of LAPDU experiments
• LAPDU experiments Rb vs. B0 and nplasma axial
  leakage rate
• SE work to continue as-needed perform
  additional shielding calculations when
  non-uniform magnet configurations are available
• FI work will focus on development of a integrated
  beam/chamber layout
• MDS work is focused on prediction of thermal
  conductivity vs. dose and tritium retention for
  graphite
• Systems modeling moving from DPSSL to chamber
  scaling work

30
Chambers Phase I Goals

• Develop a viable first wall concept for a fusion
  power plant.
• Produce a viable point design for a fusion
  power plant

Long term material issues are being resolved.
UCSD Wisconsin SNL ORNL LLNL UCSD
Example- Ion exposures on RHEPP
31
LLNL chambers 3-year plan

• Increase options/flexibility by developing
  magnetically protected chamber concept to protect
  against ion radiation damage
• Complete modeling for uniform-field configuration
  (FY03)
• Field IFE-relevant experiments on LAPDU (FY03)
• Propose additional experiments, if needed
• Develop magnetic deflection point design, if
  warranted
• Scope unique aspects and feasibility of fast
  ignition for IFE
• Produce a self-consistent beam/chamber layout for
  an FI-IFE power plant (FY03)
• Continue to work with FI community to provide
  input on FI-IFE power plants

32
LLNL chambers 3-year plan, (Contd.)

• Provide safety environmental assessments
• Assess importance of uncertainties in activation
  cross sections (FY03)
• Support laser-IFE community, as needed
• Investigate fundamental nature of radiation
  damage for materials of interest for laser-IFE
  chambers
• Produce a scaling law for graphite conductivity
  as a function of defect concentration (FY03)
• Study neutron irradiation effects in tungsten or
  other alternate wall materials
• Develop integrated systems modeling capability to
  support overall laser-IFE program future design
  efforts
• Develop scaling relationships for candidate
  laser-IFE chambers (FY03)
• Work with laser-IFE community to improve models
  for drivers and other power plant systems

33
BACK-UP SLIDESADDITIONAL INFORMATION
34
Plasma expansion experiments at LAPDU
Parameters n1- 4 X 1012 cm-3 Te5 eV, Ti1
eV B.5 - 4 kG Plasma Radius 25 cm Plasma
Length 17 m
35
Plasma expansion experiments in LAPDU
The experiment proceeds as follows 1) The probe
is positioned at some point on a predetermined
data acquisition plane using a computer
controlled stepper motor system, 2) Plasma is
pulsed on and allowed to reach a steady state,
3) 0.5 ms before the laser is fired data
acquisition begins and continues for 0.32 ms, 4)
Laser fires, 5) Steps 2-4 are repeated for 5
shots at 1 Hz, 6) Target is rotated and/or
translated, 7) Entire process is repeated at a
new probe location.
Laser incidence
Data planes were taken both parallel and
perpendicular to the background magnetic field.
Perpendicular planes were 25cm X 20cm with
.5cm interpoint spacing
36
2D Observation of the diamagnetic cavity
XY plane of data taken on green line shown above.
1.5cm down Bo from laser impact.
Diamagnetic cavity is clearly visible as well
as enhancement of background field on leading
edge of expanding plasma.
37
LAPD Upgrade Available Diagnostics and
Infrastructure

• Over 450 Access ports
• Computer Controlled Parameters and Data
  Acquisition
• Data Acquisition 20 channels _at_ 2GHz, addl LF
  channels
• Microwave Interferometers/ Reflectometers
• Laser Induced Fluorescence
• State of the Art Data Analysis and Visualization
• Bay 100 X 30, ceiling height 15,overhead
  crane
• Laser Clean Room 20 X 15,(space for 4 lasers)
• Cooling water 4.5 Megawatt
• Electrical Power available 4.5 MW

38
Plasma expansion experiments at LAPD (cont)

• Number of particles in a Debye sphere nl D3
  (variable from 102 to106)
• Magnetization parameter w pe/W e (variable
  from3?10-2 - 5?104)
• Ratio of plasma kinetic energy density to
  magnetic energy density b (variable from 10-7 to
  gt 2)
• Ratio of Alfvén speed to electron thermal
  velocity vA/ve (variable from .1 - 10)
• Number of axial Alfvén wavelengths (.5 - 10)
• ne 5?1012 cm-3 Te 10-20 eV

39
Characteristics of thelaser produced plasma
Laser Beam
Background Magnetic Field 1.5kG
.2ms
.6ms
1.0ms
No B
2 cm
10 ns gated imager time series of lpp expansion
across background magnetic field
1.9cm
1.8ms
2.2ms
1.4ms
Laser Plasma Elpp 50 laser energy 0.75
J Number of particles 2 ? 1015 Vperp-expansion
1.4-2 ? 107 cm/s, Vparllel-expansion 1.0 ?
107 cm/s
40
Results of plasma expansion experiments in LAPDU
1) Bubble scaling has been verified 2)
Ambipolar electric field has been observed 3)
Background plasma has not been observed to effect
laser-plasma diamagnetism 4) Laser-plasma
diamagnetism has been observed in 2-D 5)
Laser-plasma has been observed to generate
current structures in the background plasma, the
magnitude of which are proportional to the
background density. 6) Later time lpp expansion
induces current sheets in the background
plasma 7) The generated current structures
radiate Alfven waves 8) For the conditions
used, approximately 0.5 of the laser-produced
plasmas kinetic energy goes into Alfvén waves.