MURI Conference Working Group

1
UNM Consortium Overview Basic Research Leading
to Compact, Portable Pulsed Power First
Annual Review of FY2001 AFOSR/DoD MURI
Program Edl Schamiloglu, Gardner-Zemke
Professor, PI Department of Electrical and
Computer Engineering University of New
Mexico Albuquerque, NM 87131 USA edl_at_eece.unm.edu
Our actual proposal title
2
Compact Pulsed Power MURI
Consortium Lead Professor Edl Schamiloglu,
UNM Consortium Partner Professor Karl
Schoenbach, Old Dominion University Consortium
Partner Dr. Robert Vidmar, University of Nevada,
Reno Consortium Affiliates Sandia National
Laboratories (Gerald Yonas), Los Alamos National
Laboratory (Mike Fazio), Air Force Research
Laboratory (DE Directorate), Lockheed Martin
Aeronautics Company (Steve Calico), and
Diversified Technologies, Inc. (Marcel
Gaudreau). Program Began 1 June 2001, 1 M/year
for up to 5 years.
3
OUTLINE

• Overview of Entire Consortium Proposed Research
  Topics
• Details of Interactions Among Consortium Members
  and External Organizations
• Details of Each Members Research During Initial
  Year and Linkages to What we Heard Yesterday
• Plans for Year Two

4
What is Compact Pulsed Power?
Power Output (Watts)
Volume (m3)
5
DEFINITIONS
High Power Microwaves imprecise term, usually
denotes sources of coherent radiation spanning 1
GHz - gt100 GHz at power levels scaling as Pf 2
driven by high-perveance, relativistic electron
beams. Pulsed Power - imprecise term, typically
denotes the combination of a capacitive energy
storage system that is charged over a long
period of time (time scale seconds), and that is
rapidly switched to yield a high power pulse
that is tailored and better matched to a load
using a pulse forming line (final pulse width in
nanosecond to microsecond range). Pulsed Power
is the technology used to generate the
high-perveance beams to drive HPM sources.
6
ALTERNATE MEANINGS
High Power Microwaves a) can mean high average
power sources. A 10 kW average power tube is
considered high power. A 100 kW pulsed
power-driven tube is not very high power. b)
can mean high power ultra-wideband microwave
sources (not coherent, but broadband). Pulsed
Power - a) can mean rapidly switched solid state
components, typically used to drive
ultra-wideband sources. b) can include chemical
explosive-generated voltage pulses (explosive
flux compression generators).
7
History of HPM Linked to History of Pulsed Power
High Power Microwaves born in late 1960s and
early 1970s Prof. John Nation (Cornell
University) and Drs. Petelin, Kovalev
(Institute of Applied Physics, Gorky - Nizhny
Novgorod, USSR, in collaboration with group in
Moscow). Pulsed Power modern pulsed power
attributed to Charlie Martin and colleagues at
AWRE, Aldermaston in England in 1960s. Group
interested in radiography had to use pulsed
power to increase doses. Initial HPM sources
revisited traditional microwave tubes, except
used intense beams to drive them. Professor
Kovalev is Visiting Professor of EECE, June-July,
2002
8
How Do We Place HPM Pulsed Power in an
Academic Context?

• As an academic discipline, our research falls
  within the field of
• High Energy Density (Plasma Physics) Research
  (see National Academy of Sciences study, led by
  Ron Davidson, in progresshttp//www.nas.edu/bpa/p
  rojects/hedpp/rd.pdf)
• Inertial Confinement Fusion
• X-Ray and Gamma-Ray Radiation Sources
• Plasma Accelerators
• High Power Microwaves
• Why is there Interest in High Energy Density
  Research?
• Stockpile Stewardship (DoE radiation sources
  and ICF)
• Threats to the Infrastructure (General
  e-bomb)
• Fusion Energy Research (DoE alternate energy
  source)
• Directed Energy Weapons (DoD chemical lasers,
  HPM)

9
What Goes Around, Comes Around
Old
Pulsed Power
Electrodynamic Structure
Beams
New
Our present focus new materials, sophisticated
diagnostics, better modeling
10
Pulsed Power
Capacitive Energy Storage Systems are Typically
Large and Massive, but Basic Research can Lead to
More Compact Devices
11
What Is Required To Achieve This?
NOTE In the following slides for items with
check mark, BLUE type indicates work in progress,
BLACK type indicates work to begin in future

• Basic Research In
• Compact Pulsed Power Topologies (Folded Blumlein
  Pulsers Attractive Research Ceramics and
  Liquids Support AFRL/SNL Research)

?
12
WHY IS AFRL INTERESTED IN COMPACT BLUMLEIN
TECHNOLOGY?

• Compact folded Blumlein pulse-shaping driver
  being developed as a module to power an HPM
  source on an air frame.
• Four technical problems being addressed by this
  MURI
• Electrical breakdown at edges
• Better switch required
• Prebreakdown in propylene carbonate (w/ODU)
• Research directed towards liquid free (ceramics)
  dielectric
• (Dr. Tom Hussey, AFRL Contact)

13
What Is Required To Achieve This?
Problem 1 Electrical breakdown at
edges (48-inch Blumlein folded into a 15
assembly) Courtesy Ron Pate
14
What Is Required To Achieve This?
Electrostatic calculation of voltage contours at
edge of Blumlein. Breakdown at the edge a
problem.
15
What Is Required To Achieve This?
TOP Closure of resistive grading channel in edge
region due to flaring of adjacent cells.
BOTTOM Edge support modified to maintain open
channel to facilitate resistive grading.
16
What Is Required To Achieve This?
Problem 2 Need better switch (Rail gap switch
assembly) Courtesy Ron Pate
17
Summary of Challenges in Compact Pulser
Development
Problem 3 Prebreakdown problems with propylene
carbonate. Why does bubbling precede any sign of
corona? What can be done to prevent this?
Problem 4 Desire to move to ceramic dielectrics
(Egt400 kV/cm, dielectric constants gt100)
18
What Is Required To Achieve This?

• Basic Research In
• Improved Understanding of Breakdown in Solid and
  Liquid Dielectrics (Time-Dependent)

Photographs of prebreakdown phenomena in tap
water between a tip and plate. The gap is 100 µm,
the applied voltage 2 kV. Exposure time is 10 ns.
The first photograph shows the gap at 27 µs after
pulse application. Micro bubbles are formed at
the tip. 10 µs later (lower photograph) plasma
formation is observed at the location of the air
micro-bubbles, and shockwaves are emitted from
the plasma centers. (Courtesy Karl Schoenbach,
Old Dominion University.)
19
What Is Required To Achieve This?

• Basic Research In
• Use of Electromagnetic Simulation Tools in Design
  of Systems

20
What Is Required To Achieve This?

• Basic Research In
• High Pressure Liquid and Gas Switches

• Critical to
• Transmitter for high power, ultra-wideband
  requires large dV/dt for fixed antenna aperture
  height
• Much data exists
• University of Strathclyde has data on variety of
  dielectrics and fast pulse breakdown. UNM/Sandia
  to analyze. Problem? Relate parameters at point
  of measurement to arc region.

21
What Is Required To Achieve This?

• Basic Research In
• Tapered Transmission Lines with Microchannel
  Cooling
• Thermal Management in Pulsed Power Systems

22
What Is Required To Achieve This?

• Basic Research In
• Compact Pulsed Power Systems using Advanced
  Components (IGBTs, MOSFETS, etc.)

23
What Is Required To Achieve This?

• Basic Research In
• Behavior of Liquid Dielectrics in High Altitude
  Environment
• Developing Connectors for High Current Pulsed
  Power Applications

24
UNM CP3 MURI SYNERGISTIC RELATIONS June
2002
Karl Schoenbach Edl
Schamiloglu
Robert Vidmar Old Dominion University
University of New Mexico University of
Reno - Nevada
Tapered Trans. Lines R. Vidmar (UNR)
Fieldable HPM Systems K. Hackett (AFRL/DEHA) W.
Prather (AFRL/DEHP T. Spencer (AFRL/DEHP) C.
Woods (AFRL/DEHA)
DoD Panel R.J. Barker,AFOSR/NE P. Turchi, AFRL A.
Kehs, ARL B. Ganguly, AFRL J. Sethian, NRL G.
Cooperstein, NRL R. Gullickson, DTRA
COMPACT, PORTABLE, PULSED POWER
SS Switches LANL IGBTs
UM-Col. - PCSS
Dielectric Breakdown
SEE UNM (Manasreh) Calabazos Creek
Blumlein Technology E. Schamiloglu, C.
Christodoulou, J. Gaudet (UNM)
Liquid Breakdown K. Schoenbach R. Joshi (ODU)
Transient Breakdown S. Tyo (UNM)
International U. of Strathcylde (U.K.) IRI -
Saitama U. (Japan) IRI - QinetiQ (U.K.)
DEPSCoR D. Evans (UNM/Chem)
ASR
SNL
UM-Rolla
25
YEAR 1 ACTIVITIES AT OLD DOMINION UNIVERSITY

• Basic research in Liquid Breakdown
• Supports AFRL UNM Compact Blumlein Development
• All-Water Pulsed Power System has parameters
  attractive to AFRL Wright Patterson High
  Altitude Engine Combustion Effort (for prototype
  laboratory proof-of-principle experiments)
• Improved Understanding of this Phenomenology
• Diagnostics
• Thermal Effects

26
Folded Blumlein Transmission Lines (per Sandia
National Laboratories)
Folded Blumlein Pulser Attractive Can achieve
long (50 ns) pulse length in compact volume.
Blumlein configuration delivers charge voltage to
a matched load. Requires improved understanding
of electric field distributions, effects of
anistropies introduced in folds, liquid
breakdown, among other issues.
?
For HPM sources, impedances of interest range
from the order of Z 10 100 ?. Sandia effort
sponsored by AFRL/DEH
27
What are the challenges?
Dielectric Breakdown at the Edges of the Line
Results of HFSS (Agilent/Ansoft) calculation at
UNM
28
What are the challenges?
Percolation of Propylene Carbonate Impregnant
Photograph of prebreakdown phenomenon in tap
water between a tip and plate. The gap is 100 µm,
the applied voltage 2 kV. Exposure time is 10 ns.
The first photograph shows the gap at 27 µs after
pulse application. Microbubbles are formed at the
tip and 10 µs later (lower photograph) plasma
formation is observed at the location of the air
microbubbles, and shockwaves are emitted from the
plasma centers. (Courtesy Karl Schoenbach, Old
Dominion University.) This diagnostic set-up will
be used to test propylene carbonate samples
provided by Sandia and by an independent
laboratory. (For a movie of Schlieren photography
of a water discharge, please see
http//www.eece.unm.edu/cp3/research.htm) Also,
see J. Qian, R.P. Joshi, K.H. Schoenbach, M.
Laroussi, E. Schamiloglu, and C. Christodoulou,
Percolative Model of Electric Breakdown in
Liquid Dielectrics, to appear in IEEE Trans.
Plasma Sci. Special Issue on Pulsed Power Science
and Technology, Oct. 02.
29
A More Desirable, Longer Term Approach Would
Require High Energy-Store Dielectrics
?

Increasing dielectric constant leads to more
compact system with increased energy storage (no
folding necessary)
Challenge To identify such a material that has a
high breakdown strength and can survive many 104
discharges.
30
Candidate CeramicsGlass Composites
Composition Tm Density r K BDS (C) (g/cc) (?
cm) (V/mil) SiO2-B2O3 1200 2.06 1016 4
1900 SiO2-B2O3 1000 2.72 1015 8.5
2060 CaO SiO2-B2O3 1100 2.18 1014 4 gt1900 C
aO, Al2O3 From W. Huebner, S.C. Zhang, B.
Gilmore, M.L. Krogh, B.C. Schultz, R.C. Pate, L.
F. Rinehart, and J. M. Lundstrom, High Breakdown
Strength, Multilayer Ceramics for Compact Pulsed
Power Applications, Proceedings 12th IEEE
International Pulsed Power Conference, pp.
1242-1245 (1999).  
31
Candidate CeramicsTiO2
from W. Huebner, S.C. Zhang, B. Gilmore, M.L.
Krogh, B.C. Schultz, R.C. Pate, L. F. Rinehart,
and J. M. Lundstrom, High Breakdown Strength,
Multilayer Ceramics for Compact Pulsed Power
Applications, Proceedings 12th IEEE
International Pulsed Power Conference, pp.
1242-1245 (1999).  
32
Approach

• use electromagnetic modeling tools to design
  resistively graded structure for use in Blumlein
• iterate with ceramists to develop sample matrices
• use pulsed high voltage tester to assess
  breakdown strength (available at Sandia, Jeff
  Alexander)
• iterate with ceramists

33
Candidate Graded Dielectric
Analyze resistively graded structure proposed by
Missouri-Rolla in Blumlein configuration (B.L.
Gilmore, Ph.D. dissertation, U. Missouri-Rolla,
2002).
34
IGBT Switching

• Marvin Roybal, UNM EECE undergraduate student,
  part of LANL team nominated for 2002 RD 100
  Award (Polyphase Converter-Modulator A Compact
  High-Voltage, High-Power System)
• Work performed as part of undergraduate
  independent study directed by Prof. Schamiloglu
• Marvins contribution was the design and
  construction of a fault detection chassis that
  will turn off the IGBT if an over-current fault,
  shoot-thru fault, and/or unconnected pick-up loop
  is present.

35
IGBT Switching
Marvin will be a full time undergraduate on
campus at UNM beginning Fall 2002 and will be
supported by the MURI program as an undergraduate
RA.
36
SUMMARY YEAR 2 PLANS
As we present our plans for Year 2, we request
that the members of the DoD Review Panel provide
feedback to us in order to focus our activities
and maximize our gain over the next
year. Edl Schamiloglu, Karl Schoenbach,
and Robert Vidmar
37
SUMMARY YEAR 2 PLANS ODU

• Experiment and Modeling (water as base liquid
  pulsed systems with gt1 MV/cm)
• Electrical and optical diagnostics with high
  temporal and spatial resolution
• V(t), I(t), E(t,r), T(t,r), light emission
  (t,r), shadowgraphy, Schlieren, interferometry
• Stochastic, time-dependent, electrical model
  development, field mapping

• Results (water)
• Linear I-V characteristic up to 1 MV/cm
• Prebreakdown (10ns) determined by streamers
• Polarity effects indicate importance of
  liquid-electrode boundary
• Breakdown in microbubbles (di/dt 4 1011 A/s)
• Electrical triggering demonstrated
• Formation and decay of macro vapor bubble
  determines recovery (1 ms)
• Stochastic model predicts streamer development

Streamers in water between wire and plane

• Issues
• dielectric strength of liquid
• Boundary effects liquid/conductor
• Breakdown mechanisms
• Triggering
• Switch recovery

38
SUMMARY YEAR 2 PLANS ODU

• Energy Storage/ Dielectric Strength
• electrode surface effects - coatings
• high dielectric strength liquids (?gt ?water)
• Electrical Breakdown
• Prebreakdown electrode surface, conditions of
  liquid (content of ions and gas content,
  temperature, pressure)
• Triggering (electrical, optical (excimer laser))
• Switch Recovery
• Shock waves, vapor bubble
• effect of flow, pressure

39
SUMMARY YEAR 2 PLANS UN-Reno
BASIC RESEARCH ON TAPERED TRANSMISSION
LINE TRANSFORMERS
AND MICROCHANNEL COOLING OF
COMPONENTS
Microchannel Cooling High pressure fluid in
microchannels efficiently convect heat 3 kW/cm2
Application to conductors, passive components,
and active devices
Exponential Transmission Line
Payoff from UNR Effort High power compact
tapered transmission line transformers. Microchann
el cooling to achieve high-power
high-repetition-rate components
UNR Team Robert Vidmar (Transmission Lines and
Microchannel Cooling)
40
SUMMARY YEAR 2 PLANS UNM

• Study prebreakdown percolation of propylene
  carbonate experimentally and compare with
  modeling results (UNM/ODU).
• Design ceramic dielectrics with tailored
  dielectric constant to optimize electrical
  characteristics for use in Blumlein.
• Transition to microwave suite or other EM
  software.
• Begin collaboration with ceramists to
  manufacture desired material (have established
  relationships with U. Missouri-Rolla and Penn
  State U.).

41
SUMMARY YEAR 2 PLANS UNM

• Use electromagnetic modeling tools to design
  optimal dielectric constant distribution for
  Blumlein transmission line
• Collaborate with ceramists to understand
  limitations on manufacturability of proposed
  composites
• Diffusion of Nb5 in TiO2 critical achieving
  resistivity grading, and needs to be applied to
  BaTiO3
• NOTE The idea of using a graded resistivity is
  not new. This has been used in the pulsed power
  community and is commonly used in the manufacture
  of capacitors. Peter and Zucker recently
  proposed applying a graded resistivity coating to
  high gradient rf structures in accelerators (W.
  Peter and O. Zucker, Resistive-gradient coatings
  and high-voltage rf breakdown, submitted to J.
  Appl. Phys.).

42
SUMMARY YEAR 2 PLANS UNM
Validate the 2-D FETD code with known
cases -Uniform radial transmission line with
pulse excitation -Conical transmission lines
(non-dispersive) Implement time-domain
deconvolution strategies to deal with the
step-like excitation. -Step excitations are
notoriously difficult to work with in discrete
Fourier space because of the assumed periodicity
of discrete waveforms Use the computed H-fields
to determine the E-fields, and study the
inductive properties of the arc. - Determine the
voltage drop, which will give the time-domain
impedance characteristics of the arc. Process the
Strathclyde data set to find the switching
properties of the materials under study.