The Fission Fragment Nuclear Rocket Robert Sheldon and

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] – d. V = Vexhaust* log( final mass / initial mass) Material Isp Limitation Solid fuel LH 2/LOX Nuclear Thermal Gas Core Nuclear MHD Ion 200 -250 350 -450 825 -925 ~2, 000 < 5, 000 < 10, 000 fuel-starved efficiency-starved energy-starved Fission Fragment ~1, 500, 000 fuel-starved Matter-Antimatter Photons ~10, 000 fuel-starved 30, 000 - both-starved

Mission to the Gravitational Lens at 550 AU 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 = 550 AU / (5 yr) = DV / 5 yr=. 0027 m/s ● So DV = 425, 000 m/s ● Isp (m/s/10) Mrocket / Mpayload 1, 500, 000 1. 029 1, 000 1. 04 500, 000 1. 09 MORAL of Story: 100, 000 1. 5 DV ~ V_exhaust 10, 000 70. 6 450 1. 2 e 41 ●

Mf/Mi Comparison Missions Mission Grav. Lens Oort Cloud Alpha. Centauri Technology 550 au in 10 y 0. 5 Ly in 30 y 4 Ly in 50 y LH 2/LOX 450 s ISP 1. 2 e 41 exp(2222) exp(10666) Xe Ion 10, 000 s ISP 72 2. 69 e 43 2. 9 e 208 Fission Frag 1, 500, 000 s 1. 029 1. 95 24 Fusion Frag 2, 000 s 1. 021 1. 65 11

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 (LH 2/LOX 350 s) =H 2 O @14 k. K Nuclear thermal: Heat+propellant (Nerva 800 s) =H 2 @2 k. K Gas Core Nuclear: High heat+H 2 (2000 s) = H @50 k. K Plasmas with “magnetic walls” & nonthermal acceleration – – – MHD Engines use magnetic fields to produce a 1 -D magnetofluid nozzle for a gain of about 3 X. H~100 V ISP~2000 Ion Engines achievie non-thermal velocities from k. V electric field acceleration. Xe at 10 k. V = 100 V/nuc ISP ~ 10, 000 Fission-Fragment achieves non-thermal Velocity from MV nuclear forces. 2 Me. V/nucl ISP ~ 1, 500, 000

NERVA nuclear thermal circa 1968 1. 5 GW Pu 239 reactor cooled with GH 2 run for >30 minutes, stopped and restarted without incident at Jackass Flats Nevada test site. One version made 4. 08 GW for 12 minutes. Held the record almost 30 years for the highest power nuclear reactor on Earth. By comparison, the largest hydropower dam is 12 GW. – – Mass (dry) = 34 ton Diameter = 10. 5 m Thrust = 867 k. N in vacuum ISP ~ 820 second at 1. 2 GW 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 U 235 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 U 235 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 U 235 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 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 – R 0|-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 Me. V) 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 U 235 dust Moderator is lightweight Li. H Multipole permanent magnets on sides contain fragments

Dust suspension FAQs ● ● Can the dust be suspended while the rocket is accelerating? Yes, 1 g 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 Si. O 2 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 -400 V 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 100 nm 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 / 5 yr= 0. 002 m/s The following values are scaled from Chapline’s Am 242*-fueled rocket. We have not done a separate neutronic analysis to get the appropriate volumes for Li. H moderator and U 235 dust. – – – 10 m x 0. 5 m radius, with 30 cm moderator = 5. 4 ton Co-Sm magnets 2 cm 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 550 AU, the fuel is then. 02*9=. 18 tons ● 350 Megawatt reactor (Nerva was 4. 08 GW) ~3 mg/s ● 0. 5 Ly Oort Cloud 5. 6 GW consuming 50 mg/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 U 235 = 3 moles 0. 5 Lightyr mission=3. 7 kg U 235 = 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 U 235, both by number and curies.

Modelled Pollution from 10 moles U 235/P 239 ● By moles (total radioactivity ~10% of U 235) Rb 87 – Sr 90 – Cs 135 – Cs 137 – Nd 144 – ● . 1. 2. 3. 3. 05 = 1 u. Cu =1800 Cu = 4 m. Cu =3600 Cu =. 01 n. Cu By Curies fast diff Sr 90 – Ru 108* – Cs 137 – Ce 144 – Pm 147* – 1800 204 3600 1900 2300 slow diffusion 1800 110 3600 770 930 Cosmic Ray production C 14 = 266 Cu/yr

550 AU Mission Concept 350 MW Fission Fragment Rocket

Conclusions ● ● An interstellar probe is still a challenge with a nuclear fission-fragment rocket, but 550 AU 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 350 MW, the rocket is still 1/10 of Nerva power, and could accomplish an even shorter mission than 10 yr (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.

4 Lightyear Alpha Centauri Found-ET-must-go-now-scenario ● ● Acceleration = DV / 25 yr= 0. 06 m/s The following values are scaled from Chapline’s Am 242*-fueled rocket. We have not done a separate neutronic analysis to get the appropriate volumes for Li. H moderator and U 235 dust. – – – ● 10 m x 0. 5 m radius, with 30 cm moderator = 5. 4 ton Co-Sm magnets 2 cm 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. 8 g/s