Introduction: The driver problem

Direct energy conversion uses energy from ions, electrons, photons (x rays) or even neutrons produced by fusion or fission reactions to create electrical energy.  This completely circumvents the limitations of heat engines and thermodynamic cycles, allowing for efficient, high powered electrical energy production at reasonable temperatures and low masses.  This energy is used to power the fusion driver, that initiate the fusion process and for the more energy intensive secondary systems of the starship.  

For a fusion starship, in particular for a fast high performance robotic probe such as Icarus, the mass of a thermal system required to convert the radiative energy into hot gas or steam to drive turbo generators to power the ignition system is prohibitive.  Direct energy conversion systems are an absolute requirement. Even then, the current state of the art for direct energy conversion is not very advanced and this may be a major obstacle to the development of these vehicles.

The design of the fusion driver, the system that delivers the energy that makes the fusion reaction happen, has also proven to be a challenge that has yet to be resolved. 

System sizing

The power and energy requirements to initiate fusion reactions are considerable.  The following table lists the average driver power requirements for the fusion vehicles described on this website, as well as the original Daedalus:

The Daedalus and l’Espérance instantaneous ignition powers are much higher, in the order of 1x1016 W, but are applied for only a very short time, a few microseconds.  For Daedalus, each pellet ignition event requires  about 2,5 GJ of energy.   L’Espérance has similar power requirements due to the lower frequency of explosions, despite being on average much less powerful than Daedalus.

The Firefly Ignition power is close to the present electrical power generation capacity of North America.  Initial start up requires 0,15 GJ of energy over a period of 50 microseconds to achieve stability of the Z-pinch().  

For all starship designs in this report the initial power is provided by an energy storage system charged up by a conventional fission reactor.

Once the drive system is powered up and stable, the fission reactor is switched out and the energy comes directly from the energy recovery system.

Why thermal systems are not appropriate for fusion starships

Thermal systems are too heavy.  The best mass ratio ever achieved for a thermal system was 157 kW per kg for the Space Shuttle main turbopumps().  For the 74 GW driver required for L’Espérance, this corresponds to a mass of 147 tonnes.  And this is without the radiators or turbo alternators and other electrical systems that would multiply the mass by a few orders of magnitude to tens of thousands of tonnes.  For the 1 100 GW of Firefly, the masses would be … unreasonable.

Energy delivery

The driver problem

The system used to initiate nuclear fusion is called the driver, a term inherited from nuclear weapons research. The gain is the ratio between the power of the engine and the power of the driver.  A gain of 1 means all the power from the engine is used to power the driver, and the system has no output except thermal or radiative losses.  Thermonuclear weapons have gains of more than one but that is partly due to the increased efficiency of the fission reaction. Creating a system that does not destroy the driver with the reaction is the goal of fusion research.  However, this goal has proven difficult to reach.

In the National Ignition Facility (NIF) for example, the energy reaching the capsule is about 5% of the input energy, and 95% of the laser driver energy is lost as heat().  So, in the NIF original design, delivering 150 kJ to the target required 3 MJ of laser energy.  It was expected this would release up to 20 MJ of fusion energy, for a gain of about 7.  However, the lasers themselves are highly inefficient, with  less than 1% efficiency.  Therefore they required over 400 MJ from the power capacitors to produce 3 MJ of light.  The gain would have to go up by at least 20 times, to nearly 140 (400MJ/3MJ), to actually produce excess energy.  Diode-pumped solid-state lasers (DPSSLs) might eventually achieve an efficiency of 50%().  In such a case 6 MJ of stored energy energy might produce 20 MJ of fusion power per shot.  NIF expected to improve capsule design and laser conversion efficiency to up to 150 MJ per shot.  This would have provided a gain of about 25.  However, since the energy for the driver would come from a thermal system extracting energy from the reaction with an efficiency of about 35-40%, the actual gain would have been about 10.  It does not seem too much of a stretch to hope for the gains of 30, 50 and 67 required for Daedalus or l’Espérance.

In late 2022, the NIF experiment produced a gain of about 1,5 for the fusion reaction.  Due to the above mention inefficiencies in the driver, the overall gain was about 0,002(check).  Thus the design of a workable fusion driver remain the major obstacle to fusion energy, and eventually fusion starships.  

Electron gun drivers

The Daedalus team chose electron guns as the energy delivery system for the engine.  These already had high efficiency in the seventies  and the Daedalus team expected them to get better in the future.  Present systems have efficiencies of over 95%(ee) .  Unfortunately, theoretical and experimental work by Benford() showed that the electron flow from the electric guns would not remain sufficiently coherent to deliver the required energy to the target.  Additional research might yield more positive results but a recent review by Benford() did not find anything new on the subject, and proposed the same negative conclusion.

Lasers

Lasers were the energy delivery system envisioned for ICF in the first papers describing the concept().  Lasers remain the most likely drivers, although there have been many questions raised about their efficiency() and complexity.  Ideally, x-ray lasers would be the most appropriate type of lasers, as they would replicate the x-ray stream from fission explosions that successfully ignited fusion pellets in the Centurion experiments().  The NIF facility ultimately created x-rays from more conventional laser sources, but at the cost of a very low driver efficiency.  There does not yet appear to be a development path for high efficiency x-ray lasers.  Indeed, despite some research done for the Strategic Defence Initiative (SDI) using nuclear pumped lasers(), the idea remains largely theoretical.  And the SDI design destroyed the driver at every shot, making the concept unusable for propulsion.

Plasma guns

The Zeus team proposed the use of neutral plasma jets as the energy delivery systems.  This would essentially be slugs of deuterium and hydrogen driven onto a magnetically confined fuel target, creating a dense “liner” that would compress the target to fusion conditions and also serve to absorb neutrons from the fusion reaction.  Energy would be required to accelerate short jets to high velocity (+100 km/s), and the gain would be the difference between the jet energy and the fusion energy.  At this time, the concept suffers from the excessive mass of the material contained in the plasma jets.  The overall burnup fraction (fuel target + plasma jet) is lower than 1%, and the final ejection velocity too low for interstellar propulsion.  An unsolved problem is that as the design is optimized for higher burn up fractions, the liner gets thinner and eventually is no longer thick enough to absorb the neutrons from the fusion reaction. However, it remains an excellent solution for travel in the solar system.()

Magnetic compression

Magnetic self compression from a strong electric current through a conducting plasma is the driver for the Z-pinch Icarus Firefly.  One of the first fusion designs, the magnetic compression pinch instability has proven to be an almost insurmountable challenge, although the Z-pinch proposes a possible solution to this problem, shear flow stabilization, described in the chapter on Icarus Firefly.

Energy recovery

Direct energy conversion 

The original Daedalus study included a direct energy conversion system. It extracted energy from the nuclear fusion pulse through magnetic induction and stored it electrostatically.  The energy was then channeled to power the electron guns used to compress the propellant pellets to create a new fusion pulse, and the cycle was repeated.  

The general study of direct energy recovery from energetic ions has been researched since the seventies(1).  Successful trials have been made, with efficiencies reaching very high levels even at the prototype  stage.  But the technology has since lain dormant.  The problem is that although the efficiency of energy recovery from ionic fission products is high, the easiest fusion reaction to achieve, deuterium+tritium, produces most of its energy in the form of very high energy neutrons (75%), with only 25% of the energy going to the ions. Even at a high efficiency of 90%, the overall system efficiency is only 20% for purely ionic energy extraction. While for a liquid cooled fusion reactor, with a shielding system converting both the neutron and ion energies into heat,  steam turbines driving turbo-generators will have system efficiencies of 35 to 40%.  Thus providing better overall efficiency while using only existing and thoroughly well understood technology.

These systems work by slowing heavy ions, transforming their kinetic energy into electrical voltage, magnetic field force or electromagnetic waves and driving a current that can store the captured energy in an energy storage system.

Some direct energy conversion systems have found use in communication and scientific research under the name of “Depressed Collectors”.  They recover energy from electrons.  These are found in amplifiers (TWT) and gyrotrons, where they act as electron scavenging devices that reduce the power requirements for these systems by recuperating part of the emitted electrons, lowering power requirements and thermal dissipation problems(3).

Direct energy conversion from X-rays is also undergoing research, in particular for energy recovery from boron fusion reactor experiments(). 

At this time, there are no technologies for direct energy recovery from neutrons, although nuclear pumped lasers() could be considered as a kind of energy recovery system.

Magnetic induction

The compression of the magnetic field by the expansion on the fusion products in a pellet explosion induces a current in a conductor.  This design is usually paired with Inertial Confinement Fusion systems.  In a typical ICF cycle, much of the energy from the reaction gets stored temporarily in the compressed magnetic field. The rebounding of the magnetic field returns most of the energy to the fusion products and redirects them backwards to produce thrust.  But some of the energy can be removed from the magnetic field both as it compresses and rebounds, stored in the electrical power system in the form of electrical and magnetic fields and then used to power the next ignition cycle.  This is conceptually similar to a 4 stroke car engine cycle, with the energy recovery system playing the role of the alternator and battery.

This is the system that was proposed for Daedalus:  Two conductive rings were inserted into the magnetic field of the drive nozzle.  As the magnetic field was compressed and then expanded by the expansion and departure of the hot plasma, the magnetic field lines crossed the conductors and induced a current.  Electrostatic energy was stored electrostatically in vacuum transmission lines at 250 MVolts and an electric field gradient of 109 V/m.   The Daedalus study authors reviewed their design in 1986 and concluded that electrostatic storage system faced serious technical challenges, and proposed the use of inductive storage as an alternative(5).  The energy would be stored in a large electrical current flow, rather than in a large voltage potential. Use of superconductors made this solution particularly attractive, but the problem of creating switches capable of operating under the required loads was seen as a major technical obstacle for which no ready solution was available at the time of the review (1986). However, the development of Thyristor arrays for HVDC (high Voltage Direct Current)  lines for long distance electrical power may have created a solution to this problem.

Magnetic induction is not reserved to ICF fusion.  Continuous flow fusion system such as Icarus Firefly could achieve a similar effect by changing the mass flow rate of fuel. This would vary the plasma pressure and create an expansion compression cycle in the nozzle, therefore inducing a current with the changing magnetic field.

Travelling wave converter

This system was designed for the 14,7 MeV ions from fusion reactions. A Gyrotron guides fusion products into a microwave cavity with a powerful (10 Tesla or more) magnetic field.  The field slows the particles, that shed their energy as microwave radiation in the 150 MHz range.  This radiation is converted into high voltage DC output through rectennas.  The maximum projected efficiency is 90%().  Similar results are proposed by Tri-Alpha Energy() as the Inverse Cyclotron Converter, that acts, as it names implies, as an inverted cyclotron, slowing down ions in a spiral path to extract energy through interaction with a magnetic field.  

Converting kinetic energy to electrical energy

Venetian blind type converter

This system uses a series of plates, in an arrangement similar to venetian blinds. Ions are separated from electrons by a grid, and decelerated by the electric field between the blinds.  The ions are collected on the plates, that serve as electrodes, creating electrical currents and high voltages.  Tests were done in the seventies, reaching high efficiencies().  However, the system has many potential drawbacks, including erosion, high temperature operation and difficulties interacting with very high energy ions().  The system cannot be used with the high energy fusion products of He3/D reactions.

X-ray Photoelectric converter

This is interesting for fusion system which produce large amounts of x-rays, such as Firefly, or Boron 11 fusion systems.  However, except for aneutronic fusion reactions, the systems will also be exposed to a heavy neutron flux that may damage the materials or overheat them.  This is not applicable to Daedalus or l’Espérance as the Bremsstrahlung radiation is expected to be mostly absorbed in the pellet.

Nuclear pumped lasers

The Ghost design, introduced for the 2013 Icarus Interstellar design contest, used nuclear pumped lasers to bypass the energy recovery step altogether, creating the laser drivers for the fusion reactions directly from the neutron output.  The idea was that the neutron flux from the drive would enter a tube lined with uranium 235. Fission fragments would be created in a plasma in the tube, and excite the plasma in such a way that it created the inverse population required for the creation of a laser effect in the x-ray wavelength.  Unfortunately,  this failed for the Ghost team due to the high mass of the system and because the number of neutrons was insufficient for the laser effect.(check why exactly). Work done for the Strategic Defense Initiative program in the 1980s showed that the nuclear pumped approach might be possible, but also showed that focusing problems (there are no x-ray mirrors or lenses) as well as lower than expected power levels were significant hurdles for the development of this technology().

Energy storage

Superconducting magnets

Superconductors are available that can store 0,036 MJ/kg().  For the 3 GJ required for the Daedalus or the L’Espérance driver, this represents a mass of about 80 tonnes.  Using high tensile carbon materials, power densities of 0,15 MJ/kg should be achievable().  The energy storage system would then weight about 20 tonnes.

Although 3 GJ might seem like a lot of energy, it is only 834 kWh, or the energy content of a few high end electric car batteries.  The Large Hadron Collider at CERN already store up to 10 GJ of energy. The problem is not the amount of energy, but the rate of charging and discharging required to power the driver, or to extract power from an ICF explosion, that requires added steps and equipment, and therefore increases the overall mass.

τ = L/R.

τ= the current decay rate

L= inductance (xx)

R= electrical resistance (Ω)

 of the resulting isolated circuit, with a discharge time constant of  (cern).   

Capacitors

Capacitors are an alternative storage system for the energy required to start up the drives of all the versions of Icarus.  They can deliver the short power burst required to initiate the fusion reaction, which becomes self sustaining after the initial surge.  For Icarus Firefly, 0,15 GJ are required.  Current activated-carbon ultracapacitors are capable of storing up to 10 Wh (36 kJ) per kg, while graphene ultracapacitors have been constructed in the lab to achieve 64 Wh (230 kJ) per kg [29]. Given these energy densities, the charge for startup could be stored in redundant capacitor banks of about 30 metric tonnes, similar to the mass of superconducting magnets serving the same purpose.

An item that would require careful design and some advancement in the state of the art is the rate of current discharge.  Although the capacitors can store the required energy, they cannot, as yet, deliver the power at the required rate for all the designs. Present rates of discharge are limited to about one order of magnitude below what would be required for the starships.  So without improvement in rate of discharge the minimum mass for the capacitors would be in the order of 300+ tonnes.

Digital quantum batteries(dd) may offer an order of magnitude improvement in storage capacity, and should be studied further, as they could also solve the rate of discharge problem.

Capacitor equations

Energy handling

Energy needs to be transported between the energy recovery, the energy storage and the energy delivery systems.

Thyristor arrays (high power switches)

Thyristor arrays handling 6,4 GWatts at 800 kV have been developed for HVDC power lines used for long distance power transmission.  Individual thyristors with a maximum voltage of 8600V at 4000A, 150mm in diameter are currently available(6). 300mm thyristors should be possible(dd). These arrays should be scalable up to the 1100 GW required for Firefly, and even up to the 3500 GW required for L’Espérance.  On the Earth’s surface thyristor arrays are housed in large buildings with extensive cooling systems.  Vacuum application in starship propulsion will still require cooling but may be more compact and less massive overall.  Thyristors operating at 150C are standard up to 350C is possible for silicon carbide Thyristors(). 

150mm thyristors and wafer (ABB)

Typical thyristor array.  By dividing large voltages and currents into many lines in parallel with smaller voltage steps in series, the array can serve as a switch for the large power levels required for starship systems.  The main issues will be mass and cooling.

Conductors and Superconductors

👐

Direct energy conversion  Metric  Required  Achieved  TRL

Switches  Mass (kg/kW)  xx(xx) 1  10    (1)

Superconductors Mass (kg/MJ)  Current (A/cm2)  Magnetic (T)

Cooling systems Mass (kg/kW)

References

1  Papers on Direct energy Conversion

2  Daedalus papers

3  NASA, Multistage Depressed Collector With Efficiency of 90 to 94 Percent for Operation of a Dual-Mode TravelingWave Tube in the Linear Region Peter Ramins and Thomas A. Fox  ,1980

4  http://en.wikipedia.org/wiki/Direct_energy_conversion

5  Project Daedalus Reviewed, A.BOND and A.R. MARTIN*, JBIS, 1 986.

6  http://www.powerguru.org/six-inch-thyristors-for-uhvdc-transmission/

bb  Vobecky, J. "Future trends in high power devices." Microelectronics Proceedings (MIEL), 2010 27th International Conference on. IEEE, 2010.

cc  https://www.osa-opn.org/home/articles/volume_19/issue_5/features/the_history_of_the_x-ray_laser/#.UX3l-spUK0h

dd  Crowl, Adam, Project Icarus Capacitors for Icarus, Phase III Study, April 2011.

ee  https://en.wikipedia.org/wiki/Electron-beam_processing  Wikipedia

Next up : Power systems