Transcript Dusty Plasma Solar Sails

The Fission Fragment
Nuclear Rocket
Robert Sheldon and Rod Clark
National Space Science & Technology Center
Grassmere Dynamics, LLC
NSSTC, Huntsville, Alabama
May 13, 2005
Abstract
NASA's Human Exploration Initiative has refocussed on high-efficiency, highthrust rocket propulsion, which has returned attention to the potential of nuclear
rockets to provide a unique, high-efficiency, high-thrust propulsion technology.
There have been many nuclear rocket designs suggested over the past 50 years,
some that were developed here at MSFC, but one that has not received much
attention is the extreme high-efficiency "fission fragment" rocket, first proposed
to our knowledge by George Chapline.
Possessing a specific impulse ISP > 100,000 sec makes fission fragment
propulsion second only to pure light (or anti-matter) for raw efficiency. Previous
designs suffered (as do most nuclear rocket designs today) from concerns about
keeping the nuclear core cool. A recently studied material called "dusty plasma"
(such as Saturn's rings) held the secret to a clever solution to the heating problem,
since it provides a density intermediate between gasses and liquids. That is, basic
research into space physics has provided new materials that can solve old
technological problems resulting in improved space capabilities. Think of it as a
debt repaid. We will discuss the principles of operation, a schematic design with a
weight/size breakdown of the components, and potential mission profiles for this
breakthrough technology, with particular attention to radiation hazards.
The Rocket Equation
– Vexhaust= Isp * g
[d/dt(MV) = 0]
– dV = Vexhaust* log( final mass / initial mass)
Material
Isp
Limitation
Solid fuel
LH2/LOX
Nuclear Thermal
Gas Core Nuclear
MHD
Ion
200-250
350-450
825-925
~2,000
< 5,000
< 10,000
fuel-starved
fuel-starved
efficiency-starved
efficiency-starved
energy-starved
energy-starved
Fission Fragment
~1,500,000 fuel-starved
Matter-Antimatter
Photons
~10,000,000 fuel-starved
30,000,000-both-starved
Mission to the Gravitational
Lens at 550AU
Assume we accelerate half-way, decelerate the other half.
(Not the most intelligent approach, but good for
comparing technologies) so T_trip = 10 years.
2
2
● Acceleration = 550AU / (5yr) = DV / 5yr=.0027 m/s
● So DV = 425,000 m/s
● Isp (m/s/10)
Mrocket / Mpayload
1,500,000
1.029
1,000,000
1.04
500,000
1.09
MORAL of Story:
100,000
1.5  DV ~ V_exhaust
10,000
70.6
450
1.2e41
●
Mf/Mi Comparison Missions
\
Mission
Technology
Grav. Lens
550au in 10y
Oort Cloud
0.5Ly in 30y
AlphaCentauri 4
Ly in 50y
LH2/LOX 450
s ISP
1.2e41
Xe Ion 10,000s
ISP
72
2.69e43
2.9e208
Fission Frag
1,500,000s
1.029
1.95
24
Fusion Frag
2,000,000s
1.021
1.65
11
exp(2222) exp(10666)
Ideal Rockets
Semi-Ideal Carnot Efficiency
●
●
●
So whether we have gas or plasma confinement, hotter is
better for thermal rocket propulsion, but worse for
engineering the confinement. Carnot Effic. = (Tf - Ti)/Tf In
order to achieve better than thermal efficiency, we must
have a non-thermally accelerated rocket.
This can be done with plasma:
–
Hot gas is ionized = a plasma, with much higher temperatures
possible because of magnetic confinement
–
Plasma responds to additional forces, electric & magnetic, so it
can be accelerated (or heated in 1-dimension), with better than
Carnot efficiency.
Non-thermal acceleration = high specific-impulse rocket
How to maximize thrust with
non-ideal rockets
●
Rocket engines convert thermal energy into kinetic energy
by means of a Laval nozzle. Therefore maximizing thermal
velocity => high temperature + hydrogen atoms
–
–
–
●
Chemical: Heat=Propellant (LH2/LOX 350s) =H2O @14kK
Nuclear thermal: Heat+propellant (Nerva 800s) =H2 @2kK
Gas Core Nuclear: High heat+H2 (2000s)
= H @50kK
Plasmas with “magnetic walls” & nonthermal acceleration
–
–
–
MHD Engines use magnetic fields to produce a 1-D magnetofluid nozzle for a gain of about 3X. H~100V ISP~2000
Ion Engines achievie non-thermal velocities from kV electric
field acceleration. Xe at 10kV = 100V/nuc  ISP ~ 10,000
Fission-Fragment achieves non-thermal Velocity from MV
nuclear forces. 2MeV/nucl  ISP ~ 1,500,000
NERVA nuclear thermal
circa 1968
1.5GW Pu239 reactor cooled with GH2 run for >30 minutes,
stopped and restarted without incident at Jackass Flats
Nevada test site. One version made 4.08GW for 12 minutes.
Held the record almost 30 years for the highest power
nuclear reactor on Earth. By comparison, the
largest hydropower dam is 12GW.
–
Mass (dry) = 34 ton
–
Diameter = 10.5 m
–
Thrust = 867 kN in vacuum
–
ISP ~ 820second at 1.2GW
Could place men on Mars by
1980. Cancelled in 1972.
JPL Nuclear-Electric Concept
Shielding, Fuel
Shield shadow terminator
Reactor
Power Lines,
Coolant tubes
Cooling Fins
Ion Thrusters
Instruments
Fission Fragment Concept
●
●
Nuclear-Electric converts nuclear energy to heat,
heat to electricity, then electricity to propulsion. The
overall efficiency isn’t very high. There’s abundant
nuclear power, so low efficiency can be tolerated,
but now we also have much heat to remove, which
in space can only be done with radiators.
If the fission fragments, which contain 90% of the
nuclear energy, can be used directly for propulsion,
not only is the nuclear power extracted more
efficiently, but much less waste heat is generated.
Fuel Fibers
Fuel coated micron-thick fibers, emit >50% of fission
fragments away from fiber. Fragments can be directed out of
the system as propellant. Since 90% of energy
is in fission fragments, then <55% energy is
wasted as heat. Still, fibers get hot.
Carbon fiber
Chapline’s Fission
Fragment Rocket
Magnetic yoke
Moderator &
magnet coils
U235 coated
micron-thick
spoke-fibers
rotating fast
Fission-fragment
exhaust
Enabled Missions
Heat: The hidden killer
●
●
●
●
So the problem with space nuclear propulsion is
NOT raw power, but how to eliminate waste heat.
The more efficiently we can generate thrust, the less
waste heat produced.
Can we have our cake and eat it too? Can we have a
non-thermal nuclear propulsion minimizing waste
heat?
Yes.
Fission fragments can escape < 1 micron U235 dust
without heating the grains much. The dust radiates
heat very effectively, permitting high power levels.
●
Schematic
Chapline’s
rocket with
nuclear
fissioning
dust.
Cool Dust
How do we control,
suspend, manipulate such
a dust grain(s)?
Electrostatically.
Dust Clouds
Since we need a total
amount of U235 to
achieve criticality, how
do we collect enough
dust grains without
heating them?
Organization.
What is a dusty plasma?
Charged dust + plasma = a “plum pudding” Coulomb
crystal, or as Cooper-pairs in BCS theory. Note surface
tension & crystalline interaction.
Auburn University
University of Iowa
Fragment Confinement
More on confinement
●
B=0.6 T over 1-meter bore is an awesome
energy
density = pressure. If we could do
that we’d
be flying a fusion reactor! Instead, we use a multipole
magnet toroid, such that the field strength drops as |R –
R0|-N , with N>2, from the wall.
–
–
●
.
This has a magnetic gradient near the wall, producing a strong
mirror force, “insulating” the wall from fission fragments.
By Liouville’s theorem, n/B=constant, so fission fragment
density peaks at the wall, low in the dusty plasma center. E.g,
one pass through dust.
Because the escaping fragments are positive,  net
negative charge in the dust cloud. An ambipolar electric
field (=some fraction of MeV) develops at edge as well,
confining the fragments.
–
Proper treatment will require full kinetic simulations.
Toroidal Multipole
Magnetic Trap
●
Power & Thrust
One mirror
can be
adjusted for
either better
reflection
(more thrust)
or better
transmission
(electric
power).
Concept
●
●
●
●
Field coils on the
end control thrust &
power
U235 dust
Moderator is
lightweight LiH
Multipole
permanent magnets
on sides contain
fragments
Dust suspension FAQs
●
●
●
●
Can the dust be
suspended while the
rocket is accelerating?
Yes, 1g is typically no
problem for labs.
Will B-field change the
dusty-plasma
dynamics?
Yes, but not much.
Terrella Lab ( NSSTC)
Levitated Dusty Plasma w/Magnets
The Dust Trap
• Arc discharge on 3μ SiO2 dust grains charges them
negative. Probable charge state on dust is –10,000
e/grain.
• They are trapped in a positive space-charge region
adjacent to ring current. The RC is formed by -400V
DC glow discharge on NIB magnet, streaming
electrons ionize the air, maintain the RC. Phasespace mismatch of streaming electrons and trapped
ions produces the space charge. Highly anisotropic
B-field contributes as well.
Langmuir Probe mapping
Discharging Dust
●
●
Won’t
negatively
charged dust
discharge from
thermionic
emission? And
won’t 100nm
dust have huge
corona discharge
current?
Yes, but not as
much as one
might think.
Discharge vs Dust Size
Photoelectrons vs. size
550 AU
Power, Mass, Acceleration
●
●
Acceleration = DV / 5yr= 0.002 m/s
The following values are scaled from Chapline’s
Am242*-fueled rocket. We have not done a separate
neutronic analysis to get the appropriate volumes for
LiH moderator and U235 dust.
–
–
–
–
–
10m x 0.5m radius, with 30cm moderator = 5.4 ton
Co-Sm magnets 2cm thick w/Al windings = 1 ton
Graphite superstructure, radiators, liquid Na = 1.6 ton
Assuming that the payload is 1 ton, then total=9 tons
For a trip to 550AU, the fuel is then .02*9=.18 tons
●
350 Megawatt reactor (Nerva was 4.08 GW) ~3mg/s
●
0.5Ly Oort Cloud5.6 GW consuming 50mg/s
Nuclear Pollution?
●
●
Since radioactive fission fragments are emitted from the
rocket, how dangerous is this for the Earth?
From the two missions analyzed, we calculated how long
each rocket is withing 10 Re of the earth, and how much
fuel is burned during this time.
–
–
●
550 AU mission = 720 g U235 = 3 moles
0.5 Lightyr mission=3.7 kg U235 = 15 moles
We modelled the transport through the radiation belts,
ionosphere & stratosphere and decay lifetimes of 60 decay
products. Short-halflife products decay before reaching the
surface of earth. Long-halflife products produce almost no
radioactivity. We list radioactive products that make it to
Earth from 10 moles U235, both by number and curies.
Modelled Pollution
from 10moles U235/P239
●
By moles (total radioactivity ~10% of U235)
Rb87
– Sr90
– Cs135
– Cs137
– Nd144
–
●
.1
.2
.3
.3
.05
= 1 uCu
=1800 Cu
= 4 mCu
=3600 Cu
= .01 nCu
By Curies fast diff
Sr90
– Ru108*
– Cs137
– Ce144
– Pm147*
–
1800
204
3600
1900
2300
slow diffusion
1800
110
3600
770
930
Cosmic Ray production
C14 = 266 Cu/yr
550AU Mission Concept
350MW Fission Fragment Rocket
Conclusions
●
●
●
●
An interstellar probe is still a challenge with a nuclear
fission-fragment rocket, but 550AU gravitational lens or 1
Lyr Oort Cloud missions are eminently feasible.
We chose these missions to illustrate how close the fission
fragment rocket comes to the stuff of science fiction but
using the materials found already at hand.
For example, 550 AU is very promising. At 350MW, the
rocket is still 1/10 of Nerva power, and could accomplish
an even shorter mission than 10yr (or bigger payload than
1 ton.) Nor is pollution a real problem.
Therefore high DV missions are enabled by a promising
high-efficiency nuclear technology.
4Lightyear Alpha Centauri
Found-ET-must-go-now-scenario
●
●
●
Acceleration = DV / 25yr= 0.06 m/s
The following values are scaled from Chapline’s
Am242*-fueled rocket. We have not done a separate
neutronic analysis to get the appropriate volumes for
LiH moderator and U235 dust.
–
10m x 0.5m radius, with 30cm moderator = 5.4 ton
–
Co-Sm magnets 2cm thick w/Al windings = 1 ton
–
Graphite superstructure, radiators, liquid Na = 1.6 ton
–
Assuming that the payload is 1 ton, then total=9 tons
–
For a trip to Alpha Centauri, the fuel is then 24*9=240 tons
208 Gigawatt reactor (Nerva was 4.08 GW) ~1.8g/s