Spherical Microwave Confinement

1
Spherical Microwave Confinement

• An introduction for TUNL October, 2008
• Bill Robinson

2
History

• February 1995 Scientific American article on
  sonoluminescence and fusion got me started
  looking for exotic energy sources
• 1996-99 investigated various cold fusion ideas,
  usually shock waves through hydride aerosols
  gave up for lots of reasons
• July 2000 started investigating idea of helical
  antennas in a sphereand thought of coming to
  NCSU for physics
• 2003 started interest in Ball Lightning (BL)
• 2004 began grad school in hopes of building a
  reactor

3
More History

• 2004-2006 went through large number of possible
  designs with this geometry (including Inertial
  Electrostatic Confinement IEC) ended up with
  magnetic SMC theory, BL on the side, formal
  papers
• August 2006 started construction in 102-A
  Research II with Dr. Aspnes as advisor
• Spring 2007 obvious that magnets are beyond my
  capacity in cost, manpower, time found flaws in
  theory concentrating on BL and Spherical
  Microwave Confinement with no magnets
• September 2007 first plasma
• October 13 2007 back to SMC as variety of IEC

4
Inertial Electrostatic Confinement
(Ref. 2)
5
IEC single potential well 3
Fig. 2 Single potential well structure. The
minimum normalized potential, Ymin, coincides
with the core potential, Ycore Y(r 0). The
fractional well depth, FWD is defined as FWD
1-Ymin.
6
IEC double potential well 3
Fig. 3 Double spheroidal potential well
structure. The double well depth (DWD) is Ypeak
Ymin. Here, Ypeak coincides with Ycore. A double
well is much more likely in SMC than the single
well.
7
Existing IEC

• Large increase of plasma density in potential
  wells, fosters high rate of reaction there BUT
  net reaction rate 1/pressure
• IEC with grids cannot (yet) go above Q10-5
• Big advantages no B fields, easy high T, simple
  geometry, some fusion does occur at center and in
  mantle (zone between grids)
• High T makes advanced fuels tempting but elusive
  so far
• IEC operates at too low density for power reactor
  (need 1021 m-3 in sizable volume) 5
• IEC is the cheapest way to fusion by a very large
  factor reactors are mostly vacuum, thus low
  mass.
• Existing grid reactor can be a practical,
  portable, simple neutron source (like the STAR
  reactor), but not efficient enough yet for
  sub-critical fission or large-scale
  transmutation. Maximum so far 2x1010
  neutrons/sec by Hirsch in the 60s 6 and Nebel
  in late 90s
• Other attempts for either gridless IEC or to
  protect grids magnetically from collision
  (Bussard) have failed (new Bussard-style
  experiment might do better but far from
  break-even)

8
Critical IEC Scaling Problem 1/n

• As density drops, longer mean free path, more
  acceleration between grids, higher energy,
  increased ltsvgt, fewer ion-neutral collisions,
  tighter focus at center, more head-on collisions.
  9
• Thus fusion reactions scale as 1/n instead of n2.
  IEC reactors operate at very high vacuum ltlt
  fusion reactor range (1021m-3)
• Might not be true of SMC since mfp of runaway
  electrons are long due to velocity acceleration
  from microwaves and grids less focus anyway

9
Critical IEC scaling problem Power 1/a

• a radius of spherical active zone, q total
  charge, fa potential at r a, ne and ni are
  average densities in the active zone, P power
  from fusion
• For grid IEC, q ne ni ni
• fa q/a ni a3/a2 ni a2
• Since fa is within a small range, ni 1/a2
• P ni 2 Volume, so P 1/a
• Probably NOT true for SMC since source of ions,
  electrons, and charge balance is not the same as
  for grids q is not ni
• Proof of this is the use of ion or electron beams
  to alter the charge/density relationship in grid
  IEC to increase P
• Result is IEC devices are very small (a few
  inches) and cannot scale up while SMC probably can

10
Unavoidable Loss Problems in grid IEC

• Collisions with grids Pgridloss/Pfusion gt 3000
  particle paths MUST cross grids to be confined
  5
• Ion upscatter and energetic tail loss time 10-3
  fusion rate
• Ion neutral capture and escape from potential
  well
• Fusion reaction products escape, do not heat
  plasma (direct energy conversion probably wont
  work) 8
• Ion collisions increase angular momentum and
  throw ions out of dense center region (may not be
  so bad, double wells can work)
• No way to keep plasma non-thermal collision
  x-section gtgt fusion x-section by factor of at
  least 105
• Bremsstrahlung same or worse as other reactors,
  makes advanced non-neutronic fuels probably
  impractical (fuel touted as ideal for IEC)
• Both ion and electron loss times ltlt fusion time

11
Some recent IEC experiments
Richard Nebels Los Alamos Triple-gridded POPS
IEC 1010 n/s, 500 k, 25 kW 4
Hitachi IEC, Japan, 7 x 107 n/s
12
Spherical Microwave Confinement

• An RF-powered variation of Electron Accelerated
  Inertial Electrostatic Confinement (EXL IEC)
  without internal grids for conventional fusion
  reactions (D-D, D-T, proton-11B?)
• Critical point is to reverse outward-flowing
  electrons by near-field RF and inward-flowing
  electron waves before they reach the antennas,
  instead of requiring transit through grids
• If this is correct, the existing hardware could
  produce large numbers of neutrons. The concept
  might be developed for power generation in larger
  and more efficient reactors

13
SMC Reactor Design

• 20 conical, helical antennas for 2.45 GHz RF, 1
  wavelength long, 5 turns aluminum sphere is
  groundplane
• 20 magnetrons (1kW each) fire from cap bank (-6kV
  to -4kV), 1/10 sec
• Each hemisphere mounted on independent framework
  on casters
• 2 RF shielded windows 2 diameter
• Polar pipes (1 ¼) for access, gas in/out,
  probes, sparker, fiberoptic
• Might accommodate either hemispherical magnets or
  neutron shields 1 ½ inches off of surface, nearly
  enclosing the sphere

14
A Tour of the Lab 1
15
A Tour of the Lab 2
Video camera
Back of rack-mounted control panel and upper
capacitor bank
From 5 magnetrons to coax
Distributing current to the trons
16
A Tour of the Lab 3 alternatives
Old open coils with fresh ceramic
Plastic filled coils with ss circular grid,
trying shield patterns, effect of ceramic
Spraying ceramic inside hemisphere
Testing alternate grid with teflon disks at base
17
Antennas
Filling old coils with PVC (not a good idea) to
prevent plasma inside helix
Casting coils in epoxy
Conical helix in plaster mold for Mark II
Coating cone with silica composite ceramic
Removed epoxy cone
18
Mark II Antennas
Putting on the shields
Completed Mk II
Copper shield
Mounted antennas
Riveting triangles to base
19
Early Video Stills, Ball Lightning (2007)
Early shot 3 torr, sparker loaded with flour and
graphite 30 fps sparker should be delayed to
have maximum during microwave discharge
1) Sparker explodes aerosol
2) Magnetrons start breakdown
3) One of 3 frames, hot plasma
4) Winding down, helix cores last to cool
20
Evolving video stills
Plastic-filled coils upper shielded,
lower shielded and coated with ceramic
Open coils, thin wire grid
Conical helix in epoxy, ceramic, with bare
shield others no shield, SS grid circles
Upper completed Mark II, middle bare shield,
lower coil and ceramic
21
Antennas as e- accelerators

• Antennas are insulated with ceramic and do not
  short out to plasma
• Will apply -6 kV (or more) bias to base rings, 4
  diameter, 1 from wall.
• Microwaves cause breakdown, rapidly saturates to
  critical density (opaque plasma)
• Electron cascade bunches in waves and flows
  toward center same process turns back electrons
  exiting center
• Uncoordinated antenna phases now may be better
  in phase for inwards-moving spherical waves
• Existing rig 5 x108 e/cycle at 25 keV (0.2
  amp) assuming delivering 5 kW to waves from
  microwaves (efficiency of 0.25)
• Bias on base rings limited to no more than
  electron wave energy virtual cathode potential
  10 kV for D-T reactor, 50 kV for D-D
• Ions also bunch in waves, follow e- inwards
  qi(t) lt-qe(t-d)gt
• Inner charge during microwave increase
  qtotal qi - qe - d
  ltdqe/dtgt (qe inner electrons)
• For each 5 microseconds ion delay, can create 1
  kV potential if low electron loss

22
Current and Future Research (1)

• THEORY see how closely SMC resembles IEC,
  determine mechanism of near field and interaction
  with grid at various pressures
• Determine energy spectrum of slightly
  non-Maxwellian plasma
• Ion heating magnetrons are a few MHz out of
  phase, causes Landau damping 8
• Shock dynamics, if they apply, with antennas in
  phase or random (current setup is random)
  compression, heating
• Confinement mechanism for electrons in SMC (and
  maybe Ball Lightning theory?)
• Investigate EM knots (alternative Maxwell
  Equation solutions)
• Analyze increased depth of potential well via
  exit of fusion product ions and ions expelled via
  POPS
• Effect of antenna synchronization vs. independent
  magnetrons

23
Current and Future Research (2)

• HARDWARE Diagnostic tools are first priority
  computer DAQ, plasma probes, spectrometer, gas
  analysis, and detectors for x-rays, gammas,
  alphas, but concentrating on D-D NEUTRONS for now
• Upgrade of vacuum system for lower pressures and
  higher purity of fuel
• Hard coating on ceramics to avoid dust
• Improvement of microwave circuit (depends on
  funding) single microwave source, phase and
  frequency control, lower power and losses, CW,
  better materials
• Monolithic ceramic/Invar antennas (expensive!),
  better ceramic on inside of sphere compatible
  with expansion
• Given time, might try Ball Lightning experiments

24
Current and Future Research (3)

• GOALS SMC proven if D-D neutrons are produced at
  all, is the critical test, must be done in a
  shielded environment
• Thesis ASAP (spring 2009)
• Diagnostics of potential well and plasma
  temperature, density, kinetics, RF fields,
  reaction volume, energy spectrum
• Design of next D-D or D-T reactor as neutron
  source, en route to--? (Might try p-B11 with
  decaborane)
• Determine how SMC scales in size, plasma density,
  and RF power extrapolate to propose pilot power
  reactor
• FUNDING! And a way to continue doing this after
  graduationin this area if possible post-doc?

25
What I would need at TUNL

• Shielded space (11 x 6 minimum) safe for 2.45
  MeV neutrons, 1010 / sec (with luck!)
• 110 AC power, internet, desk/table space near the
  reactor, room for electronics rack
• 1.5 liter/min tap water for cooling turbo pump
• Can borrow fast neutron detector from NCSU for
  short times but better if can use one from here
  (and would need guidance on how to use it and
  MCA)
• Could be ready to ship the gear in November

26
Appendix APeriodically Oscillating Plasma
Sphere (POPS)
Uses RF modulation of grids and emitters to
oscillate the potential well in resonance with
the orbital frequency of the ions to extend life
of virtual cathode
(a) Temporal evolution of plasma potential at the
center of the virtual cathode with and without rf
modulation. (b) Delay in the virtual cathode
destruction due to rf modulation as a function of
modulation frequency. (Reproduced from Ref. 4.)
This is for just a few hundred volts and 10-6 torr
27
POPS SMC?

• POPS in grid IEC cannot scale to a reactor since
  
• With rvc virtual cathode radius, fo potential
  well depth note change in radius and compression
  ratio
  
• Resonant frequency
• At fusion reactor conditions, 10-30 MHz (D-D)
  milder plasmas down to kHz
• Works by throwing a few ions out of potential
  well. Might use by RF AM modulation of grids or
  microwaves, or rapidly pulsed injected beams of
  electrons or ions
• Grid IEC needs addition of electrons at center to
  reduce ion space charge and allow compression,
  may also in SMC

28
Appendix BMagnetic SMC, a possible future
addition

• Two hemispherical coils, counter-rotating
• Uses cylindrical cusp to make electron cyclotron
  resonance (ECR) on spheroidal B isosurface at 875
  gauss
• Could help make plasma transparent outside
  plasmoid
• Would heat electrons at ECR surface efficiently
  and selectively
• reactor is constructed to accommodate the coils
• Expensive and uses a lot of power if not
  superconducting
• Could funnel reaction products out poles and
  equator for direct energy conversion

Arrows are B field center circle is plasmoid
surface outer circle is magnet coil
29
Magnetic SMC

• Coil windings in amp-turns for test reactor, one
  hemisphere (other hemisphere is negative of this)

Tickmarks are meters contours are B field
magnitudes dark circle is 875 gauss (ECR) outer
circle is magnet next circle in is pressure
wall dotted circle is inner end of antennas
30
References

• 1) A. Siebenforcher, Rev. Sci. Instrum. 67(3),
  March 1996
• 2) Tom Ligon, Infinite Energy Issue 30, 2000
• 3) IEC thesis by Ryan Meyer, U. of
  Missouri-Columbia December 2007
• 4) J. Park, R.A. Nebel, S. Stange, Phys. Plasmas
  12, 056315 (2005)
• 5) A general critique of inertial-electrostatic
  confinement fusion systems, Todd Rider, Phys.
  Plasmas 2 (6), June 1995
• 6) R. L. Hirsch, J. Appl. Physics 38, 4522 (1967)
• 7) M. Rosenbluth, F. Hinton, Plasma Phys.
  Control. Fusion 36 (1994) 1255-1268
• 8) F. Chen, Plasma Physics and Controlled Fusion,
  1984
• 9) Development of a High Fluence Neutron Source
  for Nondestructive Characterization of Nuclear
  Waste, M. Pickrell, LANL Technical Report (1999)
• M. Bourham, class notes
• www.billrobinsonmusic.com/Physics for pictures,
  papers, latest news