Presentations

A new methodology for achieving the formation and compressional heating of a Field Reversed Configuration (FRC) plasmoid has been investigated both theoretically and experimentally at the University of Washington and MSNW [1]. The essential components of the Compact FRC Fusion Reactor (CFR2) staged compression method for achieving high fusion gain are a combination of high ion temperature, a steady FRC translation at subsonic velocity (100-150 km/s), along with small incremental radius reductions employing adiabatic and isentropic radial compression only. This method does appear to preclude the approach previously proposed by the author where supersonic merging was employed for compression and heating. This approach, however, avoids the catastrophic flux loss that occurs during the simultaneous acceleration and compression phase, as well as the adverse aspects of flux compression of the resultant merged FRCs needed to reach fusion temperatures. Also, with the CFR2 being a single-ended device, it makes possible a travelling fusion burn minimizing first wall thermal loading issues. With no need for flux compression in the burn chamber, a much more robust first wall can be employed providing for a close-fitting flux conserver at the maximum confining fields. It also provides a straightforward way to recover some of the remnant energy from the FRC in a remote chamber at termination.This approach, based on previous FRC empirical scaling, is expected to achieve fusion gain as large as 10. For the FRC, the fusion gain, G ~ j_p×B_e² where j_p is the FRC poloidal flux and B_e the confining axial magnetic field. G is essentially independent of the FRC radial scale making feasible a small, compact reactor. The necessary j_p (> 60 mWb) is achieved by employing a large formation chamber (~ 0.8 m radius) combined with the requisite axial magnetic field reversal time (E_q ~ 20 kV/m) for generating a high initial temperature (T_i > 1 keV) FRC. By employing the dynamic formation procedure [2], the FRC is accelerated to ~100-150 km/s as it is formed. This subsonic velocity is maintained as the FRC is translated through a series of cylindrical coils of decreasing radius but of sufficient length (> L_FRC) and number (~ 10) to produce essentially an isentropic and adiabatic radial wall-compression of the equilibrium FRC. The final stage is a 12 cm diameter, 3-6 m long confinement and burn chamber with a vacuum field of 7-9 T produced by external solenoidal coils inside the flux conserving cylinder. The vacuum field is compressed by the FRC on insertion to 35 T resulting in fusion gain conditions for the 2-5 ms transit. The large ratio of FRC to inner wall radius (0.85) substantially lowers the FRC edge pressure thereby greatly reducing both particle and thermal losses. The resulting D-T fusion yield for the 3.5 MJ FRC is 20-40 MJ/pulse. The final ejection and expansion of the FRC into a large, low field mirror chamber provides a mechanism for FRC energy recovery. The experimental justification and the physics basis for the entire process from formation through burn of the Compact FRC Fusion Reactor (CFR2) concept is presented.

References:

[1] John Slough 2025 Nucl. Fusion 65 106019

[2] Votroubek G., Slough J.T., Andreason S. and Pihl C. 2008 J. Fusion Energ 27 123

Slide Notes

Slide 1

Today I will be talking about what I believe to be the best chance to achieving fusion ignition and gain via a small scale FRC fusion reactor that I’m calling the Compact FRC Fusion Reactor or CFR2 for short . I will begin by giving a little bit of a background and motivation for why I believe this to be the most promising, and possible only way forward for fusion with the FRC.

Slide 2

In answer to the question why we are all here today, I believe it’s because we all see the potential of an FRC based fusion reactor. I list several of the FRC attributes here. I have marked in green the ones that are distinguishing features of the compact FRC fusion reactor or CFR 2. To me, the most significant advantage is #7 that the FRC can be formed and heated remotely in a far less demanding environment than the burn chamber. As will be seen for the CFR2, the necessary Tritium breeding is optimized by maximizing the blanket coverage. 

Slide 3

Despite these advantages. FRC based fusion faces many substantial challenges for viability. FRC’s typically start out life as a cold, partially ionized, gas and require turbulent dissipation of the trapped poloidal flux to raise the ion temperature sufficient to gain stability. The naturally high equilibrium beta results in a strong gradient across the separatrix. This leads to lower hybrid drift driven edge transport with a restrictive confinement scaling. The FRC suffers from several instabilities in the MHD regime therefore the FRC always requires stabilizing large orbit ions – from formation to burn. As will be seen, satisfying both these stability and transport concerns significantly restricts the operational space for attaining fusion gain. As will also be seen, there appears to be no feasible operational space for an FRC reactor using so-called advanced fusion fuels such as D – Helium 3 or P – Boron 11, which does not bode well for the near future as these are currently the only large-scale efforts being pursued today.

Slide 4

As I am making a very strong indictment of fusion using advanced fuels, it is worth seeing in more detail why I believe this to be the case. As can be seen in the graph, the ion temperature needs to increase by roughly a factor 7 for the D-helium 3 reaction to have the same reaction cross-section as D-T. It’s even worse for P-Boron 11 which, in deference to our hosts, I will not analyze here. Let’s now consider the consequence of this increase in temperature on fusion energy generation. We will we will assume here that the external conditions are the same between the D-T and D-Helium 3 FRC reactor plasmas. In other words, they are confined by the same magnetic field and have the same length and radius and obey the same particle confinement scaling. The ratio of the fusion energy produced by each FRC is found in the equation at the bottom of the slide. The first term indicates that the fusion products from D-Helium 3 are actually somewhat larger than that from D-T. 

For the same confining magnetic field, pressure balance tells us that the plasma density must drop by 1/7  to accommodate the higher temperatures needed to reach the same reaction rate. The second term indicates the consequence of that fact. This term also contains the fraction of the deuterium, tritium and helium 3 in each of the FRC plasmas. In the third term, we That for DT the optimum value of 50% can be applied for each of the two species as sufficient tritium can be supplied from the fusion blanket. This is not the case. For D-helium 3 fusion, the only helium 3 is that produced by the D-D side reaction which has a much smaller reaction cross-section. The maximum self sustained supply of helium 3 at 70 KeV temperature is roughly 12%, and even this relies on removing the tritium from the other D-D side reaction, storing it and waiting for it to decay into helium 3. With a half-life of 12.3 years, this takes several decades. Finally, there is the effect from the lower hybrid drift confinement scaling which has a density to the 1/2 dependence with a further reduction in fusion gain. The net result is that D-T fusion yield is 300 times greater than D Helium 3, which puts D Helium 3 based fusion well beyond even considering.

Slide 5

I mentioned that the transport scaling for particle confinement is based on the lower hybrid drift scaling first observed in the LANL FRX experiments which goes like Rs squared over rho I. Using radial pressure balance, this can be restated as scaling as n to the one-half times Rs squared. Notice that the particle diffusion coefficient  then only depends on n. The empirical scaling from LSX and other FRC experiments show the LHD like scaling as well as an additional, non-dimensional dependencies on the elongation denoted by epsilon and the  ratio of the FRC separatrix radius to coil radius, x sub s.  This is also consistent with experiments at TAE that did not involve adding beams and electrically biasing with plasma injection on the open field lines. While these techniques may improve the confinement, they’re not likely to be easy to accomplish in the reactor scenario for CFR2, or for any reactor for that matter.

Slide 6

There is one major exception to the simple LHD scaling shown on the previous slide. It has to do with the non-dimensional dependence on x sub s. As the graph on the left shows, flux compression significantly reduces x sub s and the consequence of that being there is a very high separatrix density. This is the case for almost all prior FRC experiments. This is bad news in that it causes the LHD edge turbulence to create a very large transport effecting the bulk FRC plasma. It is critical to get the edge density and temperature low at the closed field boundary to the open diverter flux. Not achieving this will extremely limit the ability of the FRC to achieve or sustain the necessary fusion parameters and transport for fusion gain. It is clear from these two graphs that. If one can operate at high x sub s, one can achieve the necessary isolation of the bulk FRC plasma from the turbulent transport conditions in the scrape-off layer.  The advantage of wall compression is that the large x sub s achieved initially at large coil radius can be maintained all the way to the small radius of the reactor thus preserving this key transport advantage throughout the entire compression and burn phases.

Slide 7

Let’s now consider the fusion gain scaling based on past FRC confinement, but with the appropriate high x sub s scaling as this is what we’ll achieve with CFR2. As was seen earlier, the fusion cross section in the range of 7 to 13 keV, which is the target range for CFR2, is given by a dependence that scales with the ion temperature to the five halves power as shown. Substituting the confinement scaling as well as this fusion cross section dependence, yields an expression for the fusion energy consisting of a product of density and ion temperature taken to the five halves power.  Radial pressure balance between the external magnetic field pressure and plasma pressure allows for the density and ion temperature dependence to be replaced with a magnetic field dependence that goes like B to the fifth power. Using the usual expression for the FRC energy density, that being 3 halves times average beta times the null density times temperature can also be put in terms of just Be and constants.  The fusion gain is just the ratio of the fusion energy to the FRC energy.  When one divides one by the other, we find the following expression for G in black. For the rigid rotor profile, the poloidal flux depends only on x sub s times the separatrix radius squared times the external magnetic field. Substituting this into the gain equation results in the magenta expression that only depends on the magnetic field squared and the poloidal flux plus the non-dimensional variables. This is what I mean when I say that the fusion gain is independent of any explicit dependence on size. There is no inherent disadvantage for a small reactor. It certainly is a strong advantage when it comes to large B field as the smaller coil have a much smaller hoop force to contend with. The flux can be high since it came from a coil set with an area almost 200 times larger than the reactor coil. The graph on the left displays the gain curves for a given poloidal flux and magnetic field. The expected operating region for the CFR2 is also indicated on this plot, which is based on the results from the MHD analysis and a 1D plasma  code. It is also clear from this plot that there no realistic way to attain ignition parameters for the range of x sub s and epsilon values typically obtained in past and current experiments.

Slide 8

It turns out, as will be shown later, that for the CFR 2 flux confinement is crucial. The results from past experiments typically show a flux confinement that is somewhat less than that of particle confinement, but scales roughly in the same manner. This is should not be surprising as the FRC currents are diamagnetic and thus flow where the gradient is steepest and the resistivity is anomalous and high. As the contours in the lower graph shows, that even at an x sub s of 0.6 the theta currents still flow primarily at the edge. At an x sub s of 0.85, however, it can be seen that the theta currents broaden, and the peak is now at the magnetic null. It is anticipated that this shift could significantly reduce the flux loss rate making operation at high x sub s a high priority. 

Slide 9

This brings us to the limitations imposed by the kinetic stability criteria for FRC’s. As can be seen there are several ways in which one can express the stability criteria. But I will use the one that was first derived to explain the results on TRX and is the most explicitly kinetic in character. This is the first expression in blue. The MHD Hall stability can be expressed as a function of this if you prefer that measure. Both experiment and MHD theory, as well as hybrid kinetic theory, see  increased stability threshold as s over epsilon is reduced. Empirically, from the LSX data, S over epsilon less than or equal to 1 defined the area where the FRC still showed good confinement.

Slide 10

There is one last point to discuss before detailing the CFR2 concept. This has to do with the stability scaling under adiabatic compression. There are two methods for compressing the FRC, flux compression and wall compression. As you can see on the left of this slide. that with wall compression, assuming that the FRC flux loss is minor, x sub s remains constant as the FRC radius is decreased whereas for flux compression the coil radius remains constant as the plasma radius is decreased. The stability of the FRC actually improves as the FRC is wall compressed primarily due to the FRC length decreasing more slowly than the radius resulting an increase in epsilon. The opposite is true for flux compression  criterion which actually becomes more unstable after the average beta term approaches unity and its reduction effect becomes negligible.

Slide 11

Given the requirements for fusion gain, along with the restriction placed by the stability criteria, my first attempt at a reactor concept that could achieve fusion gain was based on the supersonic merging and compression of FRC’s. It relied on mostly wall compression along with kinetic heating from the supersonic merging of two FRCs to get to fusion conditions. With the results from experiments along with 2D resistive MHD calculations, the basis for the concept was established.  VC funding was sought and I formed Helion Energy as a spin off from MSNW with the Trenta device based on this concept. Since then, with further analysis and consideration of the concept, particularly in the area concerning the FRC stability as it is compressed, made it clear to me that there was an intrinsic problem with this approach. First, a fairly modest initial ion temperature was needed to be able to attain supersonic FRCs capable of the non-isentropic heating desired at merging.  Second, the axial compression needed to perform the wall compression during acceleration led to a very small elongation such that the stability parameter was greater than one throughout the entire process. This caused anomalous flux loss during the entire compression phase leading to a very small fraction of the FRC flux making it into the compression chamber – far from what is needed for fusion gain.

Slide 12

The flux loss is likely driven by Kevin Helmholtz instability as well as Raleigh Taylor instability due to the aggressive axial compression required to drive the FRC down the conical accelerator. It is also possibly driven by tilt instability. From the 3D calculations on the right-hand side of this slide, the closed magnetic field lines just inside the separatrix are found to open up from even small angular displacement of the FRC. It was found in 2D MHD modeling the results of LSX that the initial flux loss during the 3D turbulent FRC formation where s over epsilon is greater than 1 could be replicated in the code by assuming what I call an S resistivity that describes the observed flux loss. When this is applied to the MHD calculations for the supersonic merging version and compression method it replicates the flux loss that can be inferred from the results from Trenta MHD modeling.

Slide 13

The cure for the issues encountered the initial concept is the following. Start with a large radius chamber, at least 80 cm, that can trap sufficient flux (80 mWb) at low fill pressure so that a high initial ion temperature (< 1keV) can be achieved at high x sub s (0.85) and sufficient length (4 m) to have s over epsilon less than 1.  Avoid the co and interchange instabilities with subsonic FRC motion, employing small radial compressions spread over many (10) stages to achieve the desired wall compression. No merging is required so the device can be single ended allowing of a traveling fusion burn minimizing reactor wall loading as well as a separate FRC Energy Recovery Chamber after burn.

Slide 14

The 2D MHD calculation was carried out where the S resistivity was used while S over epsilon was greater than one,  and then the lower hyper drift resistivity thereafter. The pressure contours and field line plots for the first five stages of that calculation are on the left. On the right we have from the calculation temporal plots for key parameters  One can see that the FRC velocity is subsonic oscillating somewhere around 140 km/sec. The initial ion temperature for a 1 mTorr fill, was around 11 hundred eV. The initial trapped flux settles out around 100 million or whatever. However, the decay rate observed in the code was only 500 microseconds. It was found that this was essentially a code resistivity. The expected value based on empirical scaling data would be closer to 8 milliseconds.  

Slide 15

To look at the full 10 stage compression with the proper empirical flux and particle decay rates, a 1D code was constructed. The values of the FRC fusion scaling parameters at the beginning of each stage of the CFR 2 are shown in the two plots. Given the high initial ion temperatures, the final temperatures would be greater than desired for the optimum fusion gain. This can be easily corrected. One introduces pellet fueling on CFR 2 at roughly stage 6 where the s over epsilon value has now reached suitably low value that the addition of more plasma with the consequent higher density and lower temperature. This process, however, does not cause the s over epsilon value to increase, but inhibits it from further decrease. The parameters that are not influenced by pellet fueling are plotted on the left. It can be seen in the plots on the right that the energy in the plasma does not change with pellet fueling as expected. The other parameters are changed in a manner that optimize to the desired reactor values. It can be seen that the total inventory is roughly three times that of what it would have been with no fueling. The fueling here would be much simpler than what occurs on a tokamak, as the injection speed is the speed of the FRC, which is many times what can be achieved with a pellet injector. All that’s required here is to just simply drop the pellets at the proper axial and radial location.

Slide 16

The basic layout and operation of the burn stage for CFR 2 is shown in this slide. It should be noted that, even though the FRC experiences a 35 Tesla external field, the flux conserving wall experiences the pressure of a magnetic field of only 25 Tesla. This is due to the fact that the initial Bias field is already present at FRC injection so that the wall only experiences the field from the currents induce by the FRC. The region of at the wall that experiences the maximum field is also only a small fraction of the total FRC length.

Slide 17

An important consideration when one has a reactor with the large x sub s value desired here is what happens to the fusion alphas that are created inside the FRC. As they have a much higher energy than the thermal energy of the D and T ions. Fortunately, there is a very high field in the small gap between the FRC and flux conserving wall. A 3-D calculation of the trajectory of a 3.5 MV alpha particle was done for the CFR2. The alphas were emitted with a spatially uniform and random trajectory with the number weighted by the pressure profile. This result was that most of the alphas were emitted very close to the null. With the 35 Tesla external field, none of the fusion alphas intercepted the reactor wall. The bias field however is tapered off downstream in the reactor to keep the FRC x sub s at 0.85. The external field thus slowly decreases as the FRC decays. At 27 Tesla, about 5% of the alphas did reach the wall. It was decided that the bias field for the rest of the reactor chamber must provide an external flux equal to the 27 Tesla at x sub of 0.85 to minimize the interaction. This meant at that point the x sub s of the FRC slowly decreased for the rest of the transit in the burn chamber to a final value of 0.75 at termination.

Slide 19

The final element of stage 10 involves the ejection of the FRC into the expansion chamber. This chamber and coils are at a scale similar in size to the formation Stage 1 coil, but shorter in length, with a set of mirror coils to both stop and trap the ejected FRC. The expansion chamber to burn chamber radius ratio is much larger at 14 to 1. The reason for such a large ratio is that it provides for greater expansion which lowers the bias magnetic field required as well as the plasma energy density. Since the FRC will be allowed to decay in the ERC, the lower plasma energy will help mitigate damage to where the diverter terminates with on the ERC wall end wall. For the efficiency analysis of the FRC energy recovery, it will be assumed that the pre-existing bias field B zero is provided by a high temperature superconducting coil to avoid any ohmic losses from a pulse system on this scale. This is possible here given the minimal fusion output and relatively low bias field requirement. It Is on the order of two Tesla. The induced currents in the flux preserving liner coils are driven by the compression of the bias flux from the FRC insertion and expansion. The liner coils are positioned radially inside the and along the axial and radial extent of each high temperature superconducting coil. A schematic of the one of these coils. Can be found in Figure 35 along with the recovery elements to be explained shortly. The final step is to recover the work done in increasing the magnetic field potential energy inside the conducting liner by FRC compression. This is equal to the total current induced in the liner coils times the flux inside the liner coil where the total flux produced by the initial bias field. The vacuum flux inside the liner remains constant throughout the process. On this time scale, these liners continue to act as flux conservers. The liner current however increases from zero to a maximum with the FRC achieving equilibrium inside the ERC. From Amperes law we can deduce the efficiency in terms of the work done on the B field and the injected initial FRC energy and is given by the formula shown. It is clear that the maximum efficiency is here is less than 40%.

Slide 20

I leave you to read the summary here while I take questions. There’s one point I’d like to make though. It may well be that the CFR2 is the only possible scenario for the FRC. The reason to think this stems from the need to satisfy both the criteria for stability and the criteria for a fusion gain of at least 10. Assuming optimal values for ion temperature, deuterium-tritium ratio, epsilon and x sub s we have the. Be ^2 times Phi must equal at least 100. From the stability criterion, assuming a rigid rotor profile, for s divided by epsilon to be less than one, 0.378 times the product of the reactor coil radius and external magnetic field must be lees than one. This implies that the FRC flux must always be greater than 14.3 times the radius of the confining  coil squared. Note that the flux is in units of Webers so that for coil radius of a half meter, the FRC flux must equal or greater than 3.6 Webers. This is well beyond anything I can imagination. I’ll leave it at that.

A new methodology for the compressional heating of an FRC plasmoid has been studied both theoretically and experimentally that, based on previous FRC empirical scaling, should achieve fusion ignition. For the FRC, fusion gain ~ j_p×B_e² where j_p is the FRC poloidal flux and B_e the confining axial magnetic field. The gain is essentially independent of the FRC scale, making feasible a compact reactor with a highest surface to volume ratio. The necessary reactor level j_p (>50mWb) is achieved by formation in a chamber of sufficient radius (~0.8m). An inductive technique is then employed to accelerate an array of thin, lithium metal bands radially inward to compress the FRC. These bands act as a sabot and merge in a non-destructive manner at the entrance to a cylindrical, 10cm by 3m burn chamber with 6 T vacuum field. The FRC is injected increasing the field to 30 T for the 2 ms transit. High x_s (~0.9) substantially lowers the FRC edge pressure significantly reducing thermal and particle losses. The fusion burn produces 20MJ/pulse. The eventual FRC ejection and expansion and into a large chamber allows recovery of ~90% of the FRC energy. All Compact FRC Fusion Reactor elements will be covered.

Slide 1
Today I will be talking about what I believe to be a new approach to achieving ignition in a small scale FRC fusion reactor that I’m calling CFR 2. I will begin by giving a little bit of a background and motivation for why I believe this to be the most promising way forward for fusion with the FRC.

Slide 2
The key to this concept is to use wall compression to reach high temperature and density rather than the usual flux compression typically found in most FRC experiments. Let me be clear at the outset, this is not Magneto Inertial Fusion, or MFE fusion where extremely high magnetic fields are desired and required. In this case, the fields that are generated by the wall compression result in confining fields on the order of 30 Tesla or less well below the yield strength of the magnet materials. The wall in this case is an assembly of large radius, very thin, submillimeter thick bands made of either lithium or aluminum that are driven in radially and axially by magnetic induction from coils external to the vacuum chamber. The only purpose here for the wall compression is to bring the FRC to fusion conditions, and to then translate the FRC into a stationary, relatively low-field, cylindrical burn chamber. This alleviates the need for the super high magnetic fields as the fusion burn is now a more conventional one where the FRC confinement time that dictates how long the burn lasts, not the dwell time of the imploding liner. In this way, the fusion burn is remote from the compression chamber and is in an ideal breeding geometry. The reactor being an open linear system, The management of the plasma facing components becomes much simplified with a readily replaceable tube as the first wall.
The converging metal bands provides direct, high efficiency adiabatic heating all the way to fusion conditions with 85 to 90% conversion of electrical to plasma thermal energy. The small compact fusion burn maximizes the surface to volume ratio, allowing for much higher fusion power densities. By starting with a large high flux and relatively high temperature FRC, one can maintain kinetic stability from formation through fusion burn. Even assuming lower hybrid drift transport, it was found that the CFR2 system was sufficient to reach fusion gains greater than 5. It is important to note that the magnetic field compression by inductively driven bands in this scheme is non-destructive. The lithium bands remain solid throughout the compression, producing no contaminants and doing no structural damage. After compression they fall harmlessly away allowing for reuse on future pulses.
This inductive band compression has been performed and validated at scale in our laboratory where B field compression way beyond the necessary 30 Tesla to as high as 200 Tesla was demonstrated.

Slide 3
All the steps necessary for the compact FRC fusion reactor to achieve fusion with the FRC are shown in this figure. Details of each step will be given later, but a quick overview now will be useful in understanding where I am going. The first step is to form the FRC in a large chamber in the conventional way using a field reverse theta pinch to form a high flux, high temperature FRC. This formation chamber is actually not shown in this figure for clarity. What is shown is the compression chamber that this FRC is translated into. That is where the thin bands are positioned at the inner chamber wall, near under the driver coils. The translated plasma is shown with the target parameters desired.
Step 2. Is the radial injection of DT pellets for fueling during the FRC compression. This is done for two reasons. First, it allows the density to be brought up to sufficient levels for maximum density during fusion burn. And two, it provides for a way to reduce flux and particle loss as the pellets are injected radially inward at a velocity matched to the band inward velocity and thereby produce new plasma at the null. This process will flatten the magnetic and density profiles and thereby reduce flux loss and particle transport. It is also a way to reduce tritium loss, as the tritium only enters the FRC in a confined way.
Step 3, which occurs almost simultaneous with Step 2 is the wall compression of the FRC by the inward radial and axial acceleration of the lithium bands. In the calculations made with a hybrid circuit and plasma 1D code, it was found that the velocity of the inward motion of the bands reaches a maximum of roughly 2 1/2 kilometers a second, less than the sound speed in the Lithium bands and far less than the FRC so that there are no shockwaves and the FRC compression is totally adiabatic. The band launch angle and speed is adjusted for each band to arrive at the proper location near the entrance to the burn chamber at the same time.
In Step 4, the lithium bands, now thickened and merged, form a shell at the burn chamber entrance. With the field significantly higher inside the lithium shell, the FRC is ejected into the much lower field of the reactor chamber which in on the order of six Tesla. This field rapidly rises due to the compression from the FRC itself to somewhere near 30 Tesla. It slowly drifts toward the downstream reactor exit spending most of the time at the reactor midplane. – All of the FRC’s motion is controlled by small changes in the magnetic field along the way.
The 5th step is the final stage. The FRC is ejected from the reactor chamber and expands into the energy recovery chamber, which is on the order of the size of the compression chamber, so that the expansion is quite large. In this way, much of the thermal energy remaining in the FRC can be converted back into electricity by doing work on the low magnetic field that exists in this chamber prior to insertion.
I will now give a brief description of the physics involved and point out the unique advantages CFR 2 concept, particularly in regard to the fusion burn.

Slide 4
First, I would like to discuss the FRC transport scaling I will be using. It is based on the lower hybrid drift scaling first observed in the LANL FRX experiments which goes like Rs squared over rho I. Using radial pressure balance, this can be restated as scaling as n to the one-half times Rs squared. Notice that the particle diffusion coefficient then only depends on n. The empirical scaling from LSX and other FRC experiments show additional non dimensional dependencies on elongation and x sub s. as well as the LHD scaling. This is also consistent with experiments at TAE that did not involve adding beams or massive plasma injection on the open field lines. While these things can improve the confinement, they’re not likely to be easy to do in a reactor scenario in CFR2 ,or any reactor for that matter.

Slide 5
There is one major exception to the simple LHD scaling shown on the previous slide. It has to do with the non-dimensional dependence on X sub S. As the graph on the left shows, flux compression significantly reduces X sub S and the consequence of that there is a very high separatrix density. This is the case for almost all prior FRC experiments. This is bad news in that it causes the LHD edge turbulence to create a very large transport effecting the bulk FRC plasma. It is critical to get the edge density and temperature low at the closed field boundary to the open diverter flux, as any tokamak scientist would tell you. Not achieving this will extremely limit the ability of the FRC to achieve or sustain the necessary fusion parameters and transport for fuson gain. It is clear from these two graphs that. If one can operate at high X sub S one can achieve the necessary isolation of the bulk FRC plasma from the turbulent transport conditions in the scrape-off layer. In order. The advantage of wall compression is that the large X sub S achieved initially at large radius can be maintained all the way to the small radius of the reactor thus preserving this key transport advantage throughout the entire compression and burn phases.

Slide 6
Let’s now consider the fusion gain scaling based on past FRC confinement, but with the appropriate high X sub S scaling as that is what we’ll achieve with CFR 2. The fusion cross section in the range of seven to 13 keV which is the target range for CFR2 is given by a dependence that scales with the ion temperature to the five halves power as shown. Substituting the confinement scaling as well as this fusion cross section dependence, yields an expression for the fusion energy consisting of a product of density and ion Temperature taken to the five halves power. Radial pressure balance between the external magnetic field pressure and plasma pressure allows for the density and ion temperature dependence to be replaced with a magnetic field dependence that goes like B to the fifth power. Using the usual expression for the FRC energy density, that being 3 halves times average beta times the null density times temperature can also be put in terms of just Band constants. The fusion gain is just the ratio of the fusion energy to the FRC energy. When one divides one by the other, we find the following expression for G in black. For the rigid rotor profile, the poloidal flux depends only on X sub X time the separtrix radius squared times the external magnetic field. Substituting this into the gain equation results in the magenta expression that only depends on the magnetic field squared and the poloidal flux plus the non-dimensional variables. This is what I mean when I say that the fusion gain is independent of any explicit dependence on size. There is no inherent disadvantage for a small reactor. It certainly is a strong advantage when it comes to large B field as the smaller coil have a much smaller hoop force to contend with. The flux can be high since it came from a coil set with an area almost 200 times larger than the reactor coil. The graph on the left displays the gain curves for a given poloidal flux and magnetic field. The expected operating point for the CFR 2 Is indicated the symbol on this plot, which is based on the results from from the MHD analysis and the 1D circuit with plasma code. While no current experiment can be put on this graph, the recent Helium Energy Trenta experiment almost makes the plot. Its data point lies just outside the lower left-hand corner as indicated. As it is a low X sub S flux compressed FRC, it is unlikely to ever reach breakeven indicated by the dashed black curve, let alone get to the gain of five curve in the upper right corner, which would be needed for even the most optimistic reactor scenario. To get there would take a tremendous increase in magnetic field as well as over an order of magnitude increase in poloidal flux.

Slide 7
It is time to talk about the five steps characterizing the operation of the CFR2. The first step is FRC formation and translation into a compression chamber. As you can see in this figure, the 2D resistive MHD code is used to explore how this would work. A source chamber roughly the same size as the compression chamber employs a dynamic reverse field technique that both creates and accelerates the FRC axially out of the chamber. The FRC then expands into the compression chamber in the usual way, but at high X sub S. On the left are pressure contours, on the right are the ion temperature contours. In the large radius source. Over 1.6 meters in diameter the FRC can be formed at a fairly low very low fill pressure of 1 mTorr. This assures a high initial ion temperature for stability. The large chamber also provides for a large poloidal flux. In this case the final flux just before band compression is initiated is greater than 80 milliwebers.

Slide 8
Step 2 is the pellet fueling during the band compression phase of the FRC. Deuterium and Tritium pellets are injected at six ports about the midplane for azimuthal symmetry. As you can see in the picture sequence, these pellets are rapidly vaporized and ionized by the hot FRC plasma electrons and then spread out to increase the density by over a factor of 3 from what the FRC started with. The ablation and ionization of the pellets occur primarily at the magnetic null region of the FRC. As mentioned, by being buried deep inside the FRC limits both the particle loss and flux loss during the compression phase. This is important as we want to minimize the particle loss during the 300 microseconds it takes to fully compress the FRC plasma.

Slide 9
The third stage is the wall compression of the FRC with radial and axially converging lithium bands. The driver circuit parameters and design of the Li band driver system was then optimized by running a parametric analysis using the 1D code, that integrated the behavior of the driver circuits. It also incorporated 3D magnetic field effects, and band Joule heating. Some key results from the 1D Li band code for the optimized design point for fusion gain can be found in the graph on the lower left of the slide. A plot of key parameters over the final microseconds before translation into the reactor chamber can be found in the upper righthand side of the slide. The internal magnetic field confining and the FRC does not rise significantly until near the end of the compression at small radius. It can be seen that a high magnetic field and a long dwell time of the shell at peak compression is sufficient to execute the FRC translation into the reactor.

Slide 10
The experimental validation of the inductive compression of the metal bands was carried out in the Plasma Dynamics Laboratory at the University of Washington in collaboration with MSNW staff. The experimental setup is shown in this rather busy slide on the lefthand side with the vacuum test chamber in the upper left and a Cad drawing of the inside of the vacuum chamber in the lower left. The chamber was equipped with two driver coils and the ability to test two Aluminum bands simultaneously for convergence tests . For taking the fast-framing camera shadow pictures on the right, however, only one band was driven radially inward for clarity as to the dynamics involved. The Aluminum band was vaporized even with reduced power. The windows were protected by a one-inch thick plate of polycarbonate so that pictures could be obtained without damage to the vacuum vessel. What is seen in the pictures is the compression sequence during the latter part of the collapse, as the aperture for experimental images from the end ports was limited to14 cm radius. It is clear though is that the experiment reproduced the same timing and velocity behavior as seen in the. ANSYS 3D explicit dynamics calculations. This is also reflected in the graph below the pictures. The 1D code prediction are also included on the same radius verses time plot as the ANSYS calculations and experiment with all agreeing. The buckling seen in the pictures was not unexpected as the initial band radius to thickness ratio is as large as 500. While this was not a major issue it is not desirable and can be avoided by pleating the bands around the periphery at roughly the mode number so that the bands fold up into a uniform mass with no internal stresses.

Slide 11
The fourth step is lithium shell docking at the entrance to the reactor burn chamber, which is followed by the ejection of the FRC into the chamber for fusion burn. As was seen in the 1D calculations, the Lithium bands converging at the entrance to the burn chamber bring the interna field up to about 33 Tesla. The FRC then readily moves into the flux conserving chamber which has only a vacuum field of about 6 tesla. As the FRC moves in it compresses the vacuum field eventually up to 30 Tesla when it fully enters the burn chamber. Magnetic field compression continues to raise the field in the shell as the back pressure slowing the shell inward motion dramatically falls in magnitude with the FRC exit.

Slide 12
It is worth mentioning some of the unique aspects of the CFR2 with regard to reactor operation. Since it is a pulsed experiment, there is no need for high field superconducting magnets here that must surround both the chamber and a large breeding blanket. Nor is there any need for large neutral beam port pathways, ECRH waveguides, divertor channels, fueling lines and all the other plasma startup and sustainment systems as these lie outside the reactor region. for CFR2.
The breeding blanket, therefore, can be unconstrained without the need for significant internal structures and obstacles. And can therefore achieve a higher tritium breeding ratio. For instance, with Flybe as the breeding medium, the entire volume, even that surrounding the feed plates to the inner magnets can be used., and this alone could achieve a breeding ratio of one point 27. The 6 Tesla vacuum field desired inside the reactor chamber can be readily produced using ultra high strength aluminum beryllium coils. This material is 80% Beryllium by number density, so it can also function as a multiplier to produce a breeding ratio as high as 2. The use of high yield strength materials such as Beryllium Copper for the first wall, along with the small diameter, allows for the high magnetic field operation desired for the CFR2 while still maintaining large margin of safety even at field of 30 tesla.

Slide 13
Stage 5 deals with the termination of the FRC burn and the recovery of the remnant FRC energy which would also include the fusion alpha heating during the burn. At the end of the FRC burn cycle, as it emerges from the reactor chamber, it expands rapidly into the energy recovery chamber. Here the FRC kinetic and thermal energy can be readily converted into electricity. This is done in much the same way as it has been suggested for pellet based inertial fusion devices. The way it works is a preexisting low level axial magnetic field fills the chamber as is shown in the 2D MHD calculation on the left. This field is rapidly compressed by the rapid insertion of the large energetic plasma. Here the CFR2 has the advantage over ICF in that the plasma is a stable, magnetized plasmoid, not just an explosion of hot plasma as in ICF. At peak magnetic compression, the large surge in the current supplying the magnetic circuit due to the rapid field rise, is electronically diverted to storage capacitors thus capturing the work energy done by the plasma expansion. As indicated in the pressure volume diagram, this process is equivalent to the diesel cycle. Plugging in the various volumes and energies for the various stages of the CFR2 shown at the right of the diagram, one can determine the efficiency for the CFR2. It is found that the efficiency for recovering the final FRC energy would be 84%.

Slide 14
So in summary then, let me iterate why employing the inductively driven wall compression of an FRC to achieve a compact high-power-density fusion reactor is the most promising path for FRC fusion. One, It reduces by orders of magnitude the scale and complexity involved in the development of a fusion energy reactor regardless of concept . Two, It provides for a vastly lower cost, risk and a much shorter timescale for implementation. Three, it employs a linear, simply-connected topology that solves long term material exposure with easy replacement, breeding issues due to blanket coverage as well as exhaust plasma in diverters.
For confinement better than projected: Lower electrical energy recycling for more productive and lower cost electricity. Advanced fuel cycles such as self-sustained 3He catalyzed D-D are unrealistic as the energy confinement would need to be over 300 times that of DT
For confinement lower than projected: The CFR2 would be of immediate utility to both fission nuclear energy production in the near term as well as fusion plant startup in the long term. It could also be used as a hybrid reactor for transmutation of the vast quantities of nuclear waste that have been accumulated over decades from fission plants.

Slide Notes

Slide 1

Fusion energy has been invoked for many many years as the answer to rapid space travel. Of course as the joke goes, the realization of commercial fusion is just 20 years away – and always will be.  In this talk I will try to make the case that it is actually possible to use nuclear fusion as the propulsive energy source in what I believe to be a practical way to employ fusion energy for space propulsion. I have dubbed the approach as the fusion driven rocket, or FDR for short. . It is based on solid experimental and theoretical work that I will  try to cover here. It would be impossible to make a. Convincing argument in just 15 minutes. But if this talk piques your interest, you will find a more thorough discussion in the paper and references.

Slide 2

Just like the commercial fusion witticism, one could argue that manned exploration of mars is just 20 years away and always will be.  This slide helps one understand why this is so. Essentially, it takes too long and costs too much. The drivers for these two related problems are indicated here. Because the projected times-scale for traversing the vast distance with conventional propulsion systems there is a considerable cancer risk from radiation exposure. There are other possibly less lethal concerns such as mental fatigue, bone and muscle loss, plus the increased risk of a critical failure due to the long durations. The high cost is driven by the tremendous complexity required with up to a dozen massive launches to adequately fuel provide the equipment and provisions needed for a successful mission. The solution to these problems are clear – a much shorter trip time requires high specific power as well as an appropriately high exhaust velocity or ISP if you like.

Slide 3

This slide indicates in a graphic way, just how significant  the risk is for long duration missions such as the Mars mission for humans. It is believed that the fusion driven rocket can reduce  the transit duration enough to reduce the risk below the safety threshold set for such missions. By the way, I believe there is a method that could be used to ameliorate this problem with active magnetic shielding, and as an advertisement, I include the reference for the paper that I gave on this this topic at this conference a few years ago if you’re interested.

Slide 4

To show the importance of high specific power and high exhaust velocity. I present a quick and dirty scaling argument. Accelerating the spaceship mass over at time T requires a power given simply by the top expression assuming the characteristic exhaust velocity given by the second expression. And one can then define a specific power given by the 3rd expression, where the inverse is the specific mass. The total trip time to go a distance L then is given by the final equation. Which shows, unfortunately, the rather weak  dependence on the specific mass. The equation also indicates the need for high Isp if it is to match the characteristic velocity. The requirement for the specific power even for even the 90 day Mars transit is well beyond what can likely be accomplished by either conventional chemical propulsion systems of nuclear thermal or electric propulsion.

Slide 5

It is now time to give a description of what I mean by the fusion driven rocket. Here you see on the left an artist rendition of what the final spacecraft  would look like. You will have to wait for a more detailed description of the various elements until the end of the talk. On the right. Is a schematic of the processes that occur within the thruster. The  basics of how the rocket operates will be described in the next several viewgraphs. The key to achieving a high specific power is this scheme is that virtually all the fusion energy, plasma radiation as well as the plasma energy itself will be directly deposited into the lithium propellant. The lithium propellant starts out as large radius, very thin hoops, or bands if you will, that lie very close to the thruster wall and directly under conventional copper coils. These coils are external to the thruster and are activated to inductively drive the lithium bands Inward, both radially and axially, compressing a pre-existing plasma. The large magnetic flux inside the thruster from these external coils protects The thruster and spacecraft from both the plasma and the nuclear processes that will occur in that plasma, in other words, the nuclear fusion. The isolation provided by the magnetic field and the high efficiency of thermal conversion in the magnetic nozzle allow for the radiator mass to be considerably reduced. The solid propellant essentially eliminates tankage mass. Both help to further reduce the specific mass of the system. Pulsed operation provides for a way to have variable power without sacrificing thrust or ISP?

Slide 6

I will now go over the steps involved in the process of generating the large fusion energy gain desired for the fusion driven rocket. First step, an FRC plasmoid is generated inside the thruster by external RF antennas. More will be said of this later. This plasmoid is now fueled radially by injecting frozen pellets of deuterium and tritium. This may sound a bit wild, but it is actually a common technique in tokamak fusion to refuel the plasma. The bands of lithium are now inductively driven inward to high velocity by the external magnetic coils. The converging bands compress the plasmoid to fusion conditions ending in step D. The thick shell formed by the bands now capture the fusion product energy as well as that of the plasma and radiation. The immense energy release, a factor of 300 times that input in to the FRC plasmoid, vaporizes and ionizes the lithium shell, creating a bolus of plasma at the center of the strong, divergent magnetic field that is created by both the driver coil and nozzle coils. The thermal energy of the lithium plasma is then efficiently converted into directional thrust and high ISP as it transits and expands out of the magnetic nozzle.

Slide 7

Step one is the generation of the FRC plasmoid using an external RF frequency rotating magnetic field. I will go over this slide quickly as it’s a somewhat complicated process, but basically it comes down to generating an internal rotating, transverse magnetic field generated by external saddle shaped coils as shown. This is done with a rotation frequency that lies between the ion cyclotron frequency and the electron cyclotron frequency. So in this way the electrons are magnetized to this field and rotate with it while the ions do not. Hence current. That current generates an internal magnetic field that when sufficiently large, reverses the magnetic field and forms a closed magnetic configuration  with a confined plasma that we call a Field Reversed Configuration or the FRC.

Slide 8

Step 2 is the pellet fueling during the band compression phase of the FRC. Deuterium and Tritium pellets are injected at six ports about the midplane for azimuthal symmetry. As you can see in the picture sequence, these re are eventually ionized and spread out to increase the density by over a factor of 10 from what the FRC started with. The ablation and ionization of the pellets occur primarily at the null region of the FRC which  limits both the particle loss and flux loss during the compression phase.

Slide 9

The third step is the lithium band compression of the FRC to high magnetic field, which is expected to be in the range of 500 Tesla. This is to achieve the high fusion gain that requires the megabar pressures such a field produces.  A good amount of theoretical and experimental work has been done on this step. In other labs fields as high as 720 Tesla have been produced albeit at a smaller scale. What I am showing here is the ANSYSs explicit dynamic calculations for the 3D compression employing a set of three bands. While this code will produce the most accurate picture of what it should occur in the actual thruster as well as experiments, A 1D code was also constructed to accurately include the full interaction between the driver electronic circuits and the bands. Both codes agreed on the subsequent dynamics as will be seen in the next few slides.

Slide 10

The experimental validation of the inductive compression of the metal bands was carried out in the Plasma Dynamics Laboratory at the University of Washington in collaboration with MSNW staff. The experimental setup is shown in this rather busy slide on the lefthand side with the vacuum test chamber in the upper left and a Cad drawing  of the inside of the vacuum chamber in the lower left. The chamber was equipped with two driver coils and the ability to test  two Aluminum bands simultaneously for convergence tests . For taking the fast-framing camera shadow pictures on the right, however, only one band was driven radially inward for clarity as to the dynamics involved. The Aluminum band was vaporized even with reduced power. The windows were protected by a one-inch thick plate of polycarbonate so that pictures could be obtained without damage to the vacuum vessel. What is seen in the pictures is the compression sequence during the latter part of the collapse, as the aperture for experimental images from the end ports was limited to14 cm radius. It is clear though is that the experiment reproduced the same timing and velocity behavior as seen in the. ANSYS 3D explicit dynamics calculations. This is also reflected in the graph below the pictures. The 1D code prediction are also included on the same radius verses time plot as the ANSYS calculations and experiment with all agreeing. The buckling seen in the pictures was not unexpected as the initial band radius to thickness ratio is as large as 500. While this was not a major issue it is not desirable and can be avoided  by pleating the bands around the periphery at roughly the mode number so that the bands fold up into a uniform mass with no internal stresses.

Slide 11

The driver circuit parameters and design of the Li band driver system was then optimized by running a parametric analysis using the 1D code, that integrated the behavior of the driver circuits. It also incorporated 3D magnetic field effects, and band Joule heating. A summary of the various parameters assumed for the starting conditions and some key results from the 1D Li band code for the optimized design point for fusion gain can be found on the left of the slide. Plots of key parameters as a function of time can be found in the middle row with the righthand row being a blow-up of the final microseconds when the magic happens. From the summary the internal flux inside the band remains essentially constant during the band drive as well as the implosion indicating that the Lithium band is an excellent flux conserver on this timescale. The internal magnetic field confining and the FRC does not rise significantly until near the end of the compression at small radius. For the details of the fusion gain calculation, I refer you to the paper. The result is shown here in red. It can be seen that a high magnetic field and a long dwell time, which is roughly the outer radius of the band at peak compression divided by the band velocity, are essential. The expected gain is somewhere between 220 and 880 based on the details of the expansion physics and burn up fraction of tritium. A conservative value of 300 was assumed here and for the mission analysis. With this gain, the jet velocity with the expected conversion efficiency of around 90% would be 34 kilometers a second.

Slide 12

Given the large fusion energy yield per pulse (~ 600 MJ) it is worth a brief description as to how the thermal energy released by the fusion event is translated into directed energy and in a manner that achieves high conversion efficiency. It is possible to simulate the plasma behavior in the divergent field of the magnetic nozzle as illustrated in the  time sequenced plots with the 2D resistive MHD code. Once the Lithium shell has been vaporized and sufficiently ionized, the calculation is initialized to be at a point early in the expansion phase of the plasma. At this time the plasma is assumed to have essentially no net axial velocity and is expanding outward in the divergent nozzle magnetic field at the characteristic thermal speed of the Lithium plasma ~ 30 km/sec. The fusion heated Lithium plasma initially expands out in all directions, but in the transverse direction to the magnetic field it must sweep the nozzle magnetic field along with it. The result is the plasma compresses the magnetic field as the field is also prohibited from penetrating through the electrically conductive thruster wall. This compressional work is accomplished by giving up plasma thermal energy. The transverse motion of the plasma rapidly decreases due to both this thermal energy loss and the increasing magnetic field back pressure. There is however no such deceleration parallel to the magnetic field. In fact, the opposite occurs. With the now much larger compression magnetic field near the wall, the axial force on the plasma is greatly increased by the enhanced radial component of this field, and thereby aiding the rapid conversion of thermal to axial plasma velocity. It is worth noting that it also has the benefit of creating a strong magnetic barrier to any plasma-wall interaction. The rapid conversion of plasma thermal energy to directional energy is attested by the plots of the plasma thermal energy and plasma axial velocity as a function of time found the gragh on the left. It can be seen that the conversion process starts before the plasma has expanded to the thruster wall. At roughly 40 microseconds  the plasma’s transverse motion has been halted and is in pressure balance with the compressed magnetic field. This represents the time when the external magnetic at the wall reaches its peak value of 3.5 T. While not a small field, it is well within the structural capability of many materials. The vacuum magnetic field needs to be no more than 10% of this field while still providing more than sufficient buffer distance at the wall during the plasma passage. The roughly 9% of unconverted thermal energy however does not have to be written off as a frozen flow loss. A large fraction can be recaptured into the energy storage system. It must have the ability for the wall current to flow into an auxiliary circuit and supply all the energy required to operate the fusion thruster. Again, I refer you to the paper for more details on how this would work

Slide 13

Various Mars mission scenarios were examined based on the fusion conditions derived from the 1D analysis presented earlier and validated by past experiments with inductively driven bands and experimental results on RMF FRC formation and lifetime.  A list of the mission goals that were targeted to be attained by the Fusion Driven Rocket can be found in the Table on the left. The Copernicus orbit analysis software was used to determine the maximum Mars payload, assuming a limit of a 150 ton initial launch to LEO and a 30 ton return mass. Constrained by the ΔV budget of a 210-day round trip and a 50-day stay, it was found that a 50 ton Mars lander was the maximum Mars payload. This mass was deemed reasonable considering a full Mars orbit insertion using the main FDR engine, which would minimize the thermal protection structural mass of the entry, descent, and landing systems. The dry mass breakdown is shown in the Table on the right. Foremost was a payload mass large enough to carry the Mars Ascent/descent Vehicle (MAV). For this study the payload to Mars would be primarily a Mars Entry Vehicle of 50 t. This vehicle would consist of the entry, descent, and landing system, the surface habitat for the 50 day stay, and finally the ascent module to rendezvous with the orbiting FDR spacecraft for Earth return. It is worth noting that a Crew Exploration Vehicle (CEV) or Earth re-entry vehicle is not included in the mass budget. One of the mission assumptions was to have FDR as a reusable spacecraft. Therefore, the entire spacecraft will be in circular Earth orbit where a rendezvous with a CEV could be made.

Slide 14

The final mission architecture, therefore, consisted of a Trans-Mars Injection maneuver with a ΔV of 7.6 km/s over a 4.5-day period. The spacecraft would then coast 60 days before performing a Mars Orbit Insertion burn to be captured into a Martian orbit. This burn would last a total of 6 days and use 60 t of propellant to achieve the ΔV of 13.8 km/s required. This would result in a total outbound trip time of 70.5 days. The 52 ton Mars DAV with crew and supplies would descend to the surface and remain on the surface of Mars for 50 days. The DAV would then ascend and rendezvous with the FDR spacecraft delivering the crew, but remain in Marian orbit to be used by future missions. The spacecraft, including the Mars Transit-habitat and the Fusion Driven Rocket, would leave orbit to return to Earth. The initial mass of the returning spacecraft to Earth would be 68 tons with a dry mass of 30 tons. The Trans-Earth Insertion (TEI) maneuver to escape the gravity of Mars and set a trajectory back to Earth requires a ΔV of 16.2 km/s, requiring 26 tons of propellant over 2.6 days. Again, the spacecraft coasts for 86 days before performing a final Earth orbit insertion maneuver to place the spacecraft back into its original circular 400 km orbit. With a burn time of 1.2 days, the total inbound transit time is 95 days, resulting in a total mission time of 210 days. It can be seen that the lithium fuel requirement exceeds the one launch criteria of 150 t. It actually makes more sense to bring the fuel up as a separate launch to keep it under the 150 t limit. It reduces risk and makes both launches considerably cheaper as the heavy lift is just fuel cargo. Ignoring for the moment that one should not assume more than one miracle at a time, there is a method for significantly reducing trip time and increasing payload with no significant changes to the spacecraft operation as already outlined. It invokes a aerobraking technique that does not suffer from the mass penalty that was seen for solid shields. It is referred to as the Plasma MagnetoShell. The concept is based on achieving sufficient drag from ambient atmosphere, but at much higher altitude using a large-scale magnetized plasma “parachute” tethered to the spacecraft. Experiments at MSNW funded by the NIAC program have demonstrated the feasibility of this concept for just this kind of rapid, high delta V braking required for MOI and EOI. Given the many unknowns and substantial challenges to successfully develop the FDR, the Plasma MagnetoShell looks to be a “gimmie” in comparison. A reference to this work can be found in the paper.  Also, an ultrafast 90 day mission with a 22 day stay is also described in the paper.

Slide 15

I’ll finish by showing a little movie that we made back a few years ago after completing the initial design of the FDR. The concept has been improved and the mission analysis has been updated quite a bit since then, but it still gives a fair representation of the essential elements of the fusion driven rocket concept.

Slide 1

Deep space exploration and development have been stymied due to the enormous energy demands of rapid transit over such vast distances. Past missions have been carried out by relatively small spacecraft, launched on the few occasions when rare planetary alignments provided for the gravitational boosting necessary to perform the trip on a somewhat reasonable timescale. To create a more efficient propulsion system for in-space transportation one must look to the NEP and NTP systems that are now being developed, and which show tremendous potential to facilitate the human exploration of Mars. They are, however, unlikely to meet the challenge of deep space without significant advances in efficiency and the ability to use in-situ resources. The ultimate goal of the work to be described here is the development of a planetary orbiter spacecraft that can operate on the gases found in virtually every planetary atmosphere and is designed in such a way that it can also harvest the planetary atmospheres to obtain the fuel required for the return. It is based on experimental results employing a robust plasma creation and sustainment methodology referred to as a Rotating Magnetic Field (RMF) generated Field Reversed Configuration (FRC). I will now make a short digression to provide a little bit more background on this important critical aspect of the Harvesting Planetary Orbiter.

Slide 2

Essential to the groundbreaking in-space propulsion technology being described here was the demonstration of critical performance criteria for the proposed propulsion system to be used on the Harvesting Planetary Orbiter. This was accomplished in a series of experiments performed at MSNW. These experiments employed a robust plasma creation and sustainment methodology referred to as a Rotating Magnetic Field (RMF) generated Field Reversed Configuration or FRC. Most importantly, the tests were executed in a manner pertinent to the very environment that would be encountered in planetary atmospheres, and in particular, the gas giants. The methodology was actually developed as a startup process for a nuclear fusion concept that employs light gases similar to the Jovian planetary atmospheres, namely Hydrogen and Helium isotopes as the nuclear fuel. In the experiments, a Hydrogenic gas was partially ionized and accelerated out of a Magnetized Cascaded Arc Source with directed velocities of 35 km/s that is characteristic of the conditions that a Harvesting Planetary Orbiter (HarPO) would confront in the atmospheres of the Jovian planets. A schematic of the experimental layout is depicted in this slide. I will now give a brief review of what an FRC is as well as how a Rotating Magnetic Field, or RMF can generate this is rather unique. and adaptable plasmoid that has shown to have many important applications from space propulsion to nuclear fusion. 

Slide 3

The field-reversed configuration (FRC) is a toroidally symmetric plasmoid that has the geometrically simplest, most compact, and highest power density of all currently studied magnetic confinement schemes [1]. The most successful FRCs in terms of the Lawson parameter for fusion have been obtained in a modified low-compression theta-pinch initiated with the axial magnetic field reversed – hence its name. The basic configuration and key equilibrium parameters can be found on this slide. This formation method will be referred to as pulsed formation. Even though the results from the largest FRC experiment to date indicated confinement scaling adequate for fusion [5], it was also clear that the pulsed formation method could not provide the required confining closed flux into the reactor regime. There was a method that has been investigated experimentally for years that generates the toroidal currents directly by means of a transverse rotating magnetic field (RMF). The early experiments were limited to currents of only a few kiloamperes, so that the confining axial magnetic field, Be, generated was no larger than the rotating magnetic field. The effects of the RMF on the FRC plasma confinement were also unknown as the plasmas generated in these experiments were partially ionized and in contact with the vessel wall, with electron temperatures less than 20 eV.

Slide 4

Starting with the Star Thrust Experiment (STX) [7] at MSNW and the University of Washington, the RMF fields were applied to a cylindrical vacuum chamber with the proper hygiene required for the containment of a hot plasma.  A schematic of the cross section of the plasma column and RMF lines of force is shown in the figure at the upper left. . The coil set of axial conductors employed to generate the Rotating Magnetic field are also shown. The FRC was also formed in a flux conserving coil and operated in a manner that prevented FRC plasma contact with the vacuum vessel wall. The experiments were of sufficient size; a half meter diameter and over two meters in length, and with sufficient RMF power, produced an FRC with magnetic field was over 5 times larger than the RMF field used to generate it . This device is shown at the lower left. These two orthogonal sets, carrying sinusoidal currents phased 90 degrees apart, produce an m=1 rotating magnetic field of constant amplitude as indicated in the end-on schematic in the upper right.  The frequency of the RMF is such as to lie between the electron cyclotron frequency and the ion cyclotron frequency. In this regime, only the electrons can respond to the rotating field. The ions, due to their greater inertial mass, cannot respond sufficiently to the movement of the rotating field and are left unaffected by the RMF. The electrons, however, are magnetized and flow synchronously with the rotating magnetic field thus producing a large azimuthal current. Using this technique, RMF FRC’s. have been produced in gases ranging from hydrogen, which can be seen in the plasma on the lower left, to xenon as we can see in the pictures on the lower right. The end on picture clearly shows the doughnut like shape of the FRC radial profile.

Slide 5

Now back to the critical experimental  results that were key for the Harvesting planetary orbiter feasibility. A picture of the central section of the experiment is shown here. The transverse Rotating Magnetic Field is generated in this case by two pairs of phased coils at right angles. Using a quasi-Helmholtz pair to generate each of the horizontal and vertical components of the RMF made for a more uniform transverse field, but a rather complicated looking coil set. In these experiments the transverse RMF fully ionized the gas, and in the process generated large azimuthal currents which produced an FRC that confined and heated the ions. The ion heating is believed to be a consequence of the RMF induced flows to establish and maintain the field-reversed magnetic field profile which apparently can be a somewhat turbulent process. The helium needle injector with a fast puff valve is highlighted as transiently doping the plasma with a small amount of helium  provided an accurate way to measure the plasma ion temperature during the RMF plasma sustainment phase of the experiment.

Slide 6

A plot of the measured plasma temperatures during the equilibrium phase on STX-HF shown here confirms this ion heating. As mentioned, to obtain a direct measure of the ion temperature, a small jet of Helium gas was introduced into the plasma from a side port shown in the previous slide.in   After sufficient time to reach thermal equilibrium with the Hydrogen, the observed temperature from Doppler broadening measurements of a He II line confirmed the ion temperature inferred from radial pressure balance with the external magnetic field . This pressure balance temperature is also shown on the plot. The electron temperature measured by the Langmuir probes was more or less constant as it was clamped at roughly 10 eV by energy losses to the continuous process of ionization of the incoming neutral flow. It should be noted that the slow overall decrease in temperatures observed is due to the drop in RMF input power and thus axial magnetic field pressure, as the energy source for the RMF was an electrolytic capacitor bank which was slowly being discharged over the course of the experiment.

Slide 7

As in all FRC experiments, the stationary axial guide magnetic field confined the plasma within the chamber wall and conducted away the lost plasma at the end of the FRC forming an on-axis plasma jet and thereby converted plasma thermal energy to directional energy. A singular and particularly advantageous effect of the RMF electric fields is that it produces a radially inward flow of the plasma which dramatically reduced the plasma loss rate from the FRC. The transit time of the highly directed neutral flow through the FRC formation region observed with no RMF was only ~ 5 microseconds. The transverse RMF immediately entrained electrons to flow azimuthally and with the inwardly directed flow, rapidly built up the density and size of the FRC, but now with a plasma confinement time an order of magnitude larger.

The entire process can be observed from the end-on filtered plasma emission captured by a fast-framing camera. The images and setup can be found on the slide. The camera is positioned far enough away from the FRC that there is very little parallax. It is clear that the FRC is isolated from any contact with the formation chamber wall with the exhaust eventually intercepting the large chamber walls nearer the camera. The hard vacuum maintained in this chamber (~ 10^-9 Torr) assures that a spacelike environment is maintained throughout the experimental run. The large mirror ratio axially across the formation chamber assures that the plasma flows only in the downstream direction

Slide 8

All of these advantageous features observed in the experiments make for a uniquely efficient and revolutionary electric propulsion system, and the initial design of the HarPO spacecraft was made to assure that they would be maintained in a scaled-up version. The figure illustrates both the elements of the HarPO as an EP thruster operation and as a harvester. The ability for the same device to be able to switch from harvester mode to propulsion mode is done in the following manner. Internal segmented conical deflector redirects incoming flow to an array of pumps. Deflector is retracted to convert HarPO into a high-power electromagnetic thruster with only the incoming flow as propellant. It should be noted that the HarPO concept has actually several aspects that hold promise for an electric propulsion system that would have critical advantages over traditional systems utilizing Xenon. For one, the RMF FRC based thruster is a fully electromagnetic and electrodeless propulsion technology. RMF FRCs have been successfully formed and maintained at powers ranging from 1 kW to 10 MW. Due to the electrodeless nature of the thruster, FRC plasmas have been produced in a myriad of gases from Hydrogen through Xenon. The RMF-formed FRC technology has been demonstrated to be very lightweight as it is essentially comprised of nothing more than a hollow-tube dielectric chamber surrounded by a set of cylindrical coils capable of generating a few hundred Gauss axial field, along with a pair of saddle shaped antenna coils arranged radially on top to generate the RMF as shown on the slide. The only circuitry required is a steady, low voltage supply for the axial coils. The RMF requires  a sinusoidal current oscillation for each antenna. This can be accomplished by a capacitor connected to the RMF antenna and IGBT switch to drive the L-C circuit. The specific mass for the preliminary flight design is anticipated to be less than 1 kg/kW for thruster and PPU combined. And finally, there is the magnetic isolation of the FRC plasma from the thruster walls and spacecraft during operation while also providing efficient thermal conversion of the plasma exhaust provides much higher efficiencies and specific power. It is these attributes, combined with the low-drag, maximal intake design that provides for the dual role as a propellant harvester

Slide 9

The HarPO system needs to efficiently process the incoming atmosphere into the necessary thrust and Isp for drag makeup. With an ion temperature similar to that observed in the STX experiments, only a fraction of the gas ingested during the orbital transit would be required as the projected Isp operating on a Hydrogen propellant is 12 thousand seconds. The plasma chamber radius, which is the same as the spacecraft intake aperture, need be no larger than that employed in past RMF FRC experiments with a forty centimeter radius. Several 2D Magneto-Hydrodynamic (MHD) calculations for both STX and the larger RMF FRC experiment have been performed that accurately reproduced the experimental results. This same code was used for numerical calculation of the HarPO thruster system. The results from these calculations are shown in the figure here. It can be seen that as plasma is lost from the FRC, it flows down the magnetic gradient along the separatrix onto the open field near the axis and in the process forms a highly directed plasma jet far from the HarPO walls thereby maximizing both thrust and Isp.

It should be noted that this is a 2D MHD fluid calculation. The effect of the magnetic mirror reflection of the plasma ions and electrons moving in the upstream direction cannot be resolved as this effect is due to the individual particle behavior in velocity space. Also, the influence of a relatively strong directional neutral particle flow was also not part of the calculation. These two important effects will be incorporated in future analytical efforts in optimizing the HarPO both as a thruster and as a harvester with the addition of a kinetic particle tracker which can capture the effects of both the ambient magnetic and electric fields on the individual particle motion.

Slide 10

The preliminary analysis and design of a prototype HarPO has been completed based on the drag, power processing, energy and fuel-storage requirements that would be encountered in the Jovian atmospheres. At this point the operational scenario would be as follows. The basic mode of transportation to/from the planetary systems will employ a NEP reactor along the lines of that proposed for the JIMO mission. It is believed that the HarPO, as indicated in the MHD analysis, should be capable of operating on Hydrogenic fuel at a steady power of 300-400 kW. This is made possible by the negligible plasma radiative (fully stripped ions) and conduction losses (thruster walls magnetically isolated from plasma). A significantly larger reactor power is thus envisioned (0.5 to 2 MWe), allowing for several HarPOs to be mounted to the main spacecraft. At arrival, the HarPOs would be sequentially launched on an elliptical orbit around the planetary body such that the apogee (and main spacecraft orbit) is at an altitude for negligible drag, and the perigee of the HarPO is at the proper altitude for optimal propellant harvesting as shown in this figure. During most of the orbit the HarPO is in the harvesting configuration where a segmented conical deflector is engaged to direct the flow into the annular collection plenum where it is pumped and stored. For drag make-up, the deflector is retracted so that all of the atmospheric flow into the maw of the HarPO can be utilized as propellant There is enough energy in the HarPO battery array to power the HarPO for drag makeup for the entire orbit. On return to apogee, it docks to the NEP spaceship, unloads the propellant, and recharges its batteries for the next orbital run. The propellant collection and impulse required for drag makeup for an orbit at Neptune is found in this figure as well.

Slide 11

Starting with this preliminary analysis, there was a series of studies that have been performed to address the significant unknowns in terms of the physics, required technologies, and mission architecture.in order to have confidence that the HarPO concept is a viable one. A preliminary Direct Simulation Monte Carlo calculation with parameters similar to what would be encountered by a HarPO in the Jovian atmospheres. The results of the calculation assuming a Knudsen number of 50, a Mach 10 flow and a completely diffusive interaction with the conical surface can be seen in this figure. It can be observed that virtually all particles diffuse out radially which would be into the pump-out chamber region of the HarPO spacecraft. The drag coefficient was calculated to be 2.05 indicating nearly complete flow stagnation, which is the desired result. Future calculations will quantify the collection efficiency as well as examine a range of capture geometries to optimize harvesting.

Slide 12

To get a sense of the amount of propellant that must be harvested by the HarPO for a typical Jovian planetary mission, it is useful to make a “back of the envelope” calculation for a specific interplanetary transit. A good representative case would be a sample return mission from Jupiter for a spacecraft similar to the NEP powered JIMO spacecraft. The dry mass for this spacecraft was to be roughly 24,000 kg. As the reactor module made up only 25% of this mass, and a large fraction of the remainder was comprised of the spacecraft module itself, the possible differences in nuclear power requirements and thruster masses between the JIMO and a HarPO mission should not be that significant a factor in this zero-order analysis.

For simplicity, it will be assumed that a Hohmann transfer orbit will be employed on the return, for HarPO and that aerobraking in the Earth’s atmosphere will be employed at Earth. The delta V budget for the Hohmann transfer for a Jupiter-Earth mission is 6.3 km/sec with a transit time of 2.7 years. It will be assumed that the HarPO spacecraft thrusters will operate on hydrogen achieving the same ion thermal temperatures as found in the subscale experiments at MSNW which would produce an exhaust velocity of 107 km/sec which is considerably higher than what was assumed for the combination of Ion and Hall thrusters to be employed on the original JIMO mission. . The transfer will thus require a propellant mass at departure from Jupiter equal to 6 % of the spacecraft dry mass, or 1,440 kg of Hydrogen. A similar analysis to that for Neptune was performed for a HarPO operating in the Jovian atmosphere. Due to the higher orbital velocity at Jupiter, more propellant was consumed in drag makeup so that roughly a net of 1 kg of Hydrogen was obtained on each HarPO orbit. The time to complete one orbit at Jupiter is 3.0 hours. A conservative 50% duty cycle for HarPO propellant  collecting will be assumed with depositing propellant and battery recharging making up the remainder. The propellant collection rate would be 1 kg/hr for the array of 6 HarPO thrusters. The time to refuel for the return transit mission would thus be 60 days. Given the 985-day return transit time, this is a very reasonable collection period, and would likely be increased to provide more exploratory time at Jupiter as well as to fuel an even shorter return trip time.

Slide 13

A successful development of HarPO would radically change in-space transportation in several ways. It would enable deep space missions to all the Jovian planets. With essentially unlimited amounts of fuel, the exploration of the Jupiter’s Moons, and by “planet hopping”, those of Saturn through Neptune could be further explored on the same mission. Importantly, the ability to fuel a return to Earth opens up the possibility of a much greater trove of both small and large specimens from the many moons that populate these planets, not least of which is the return of the very valuable spacecraft and scientific hardware deployed for the mission. The reuse of these assets would substantially reduce the cost of further missions to both the inner and outer planets.

Slide 1

For planetary missions one could argue that the challenge of orbit insertion at the target planetary body is as critical and potentially as fuel consumptive as launch. Aerocapture represents a fuel-free method of deceleration which significantly reduces the mass of an interplanetary spacecraft as a substantial fraction of the spacecraft mass is propellant needed for the orbit insertion. Plasma MagnetoShell Aerocapture is an innovative approach that has been studied at MSNW that would enable low-cost, low-risk aerocapture for a range of Earth-return  and deep space missions. This talk will describe the concept , its experimental demonstration as well as how it could be implemented on a future manned mars mission.

Slide 2

As can be seen in the figure, the Plasma MagnetoShell acts like a plasma parachute. It has the advantage that, due to the much larger scale of the plasma, the MagnetoShell drag can be much greater than aerodrag from a conventional aeroshell. A major advantage to the plasma aeroshell. Is that the drag can be adjusted in a matter of seconds by simply changing the scale of the dipole magnetic field. It has the same Delta V savings as inflatables, but without any heating concern as it is only the dipole plasma that is heated from intercepting  incoming atmospheric neutrals. The plasma itself is effectively massless – less than a gram. The dipole coil and tethers do,  of course, have some mass, but far less than a solid aeroshell. The dipole magnetic fields required. ae low –  only in a few 100 gauss at most. The power requirements are thus  low and there is no need for high temperature superconducting coils. The challenges, of course, are that the system is at a very low TRL at this point, and that further testing of the full-scale system may be quite challenging. due to the large scale of the plasma.

Slide 3

The PMS concept conceived, analyzed, developed, and experimentally tested  is a possible answer to the limitations of conventional aerobraking with a solid shield.  The method involves the deployment of a coil set that would produce a dipole magnetic field that is then filled with a plasma to provide the necessary impediment to the atmospheric flow past the spacecraft and thereby produce the desired drag for braking. The energy dissipation required to slow the spacecraft would be deposited directly into the plasma ions, and then by thermal equilibration, to the electrons. The energized electrons will then maintain the plasma density by ionization of the incoming neutral particle flow. The dissipation of spacecraft directed energy shows up as an increase in plasma thermal energy which is eventually carried away by diffusion and loss  at the magnetospheric periphery in an isotropic manner. It should be noted that the Plasma MagnetoShell can extend out to distances far greater than the meter scale of the aeroshell coils making even more effective as a drag device. 

Slide 4

The spacecraft deceleration is achieved through neutral-ion collisions, but primarily through charge exchange interactions between the magnetized thermal ions in the Plasma MagnetoShell and the atmospheric neutrals as this has a larger cross section as the graphs show. In the charge exchange case, the fast and directed neutrals in the atmosphere are losing an electron to the ambient ions and are thus converted into ions, but now with a large velocity component in the direction of flow. These new ions are, however, magnetized in the dipole field due to their newly gained charge. Their motion is deflected by the dipole field and becomes trapped. Their momentum is imparted to the dipole coils through the field line bending. The coils in turn convey the force to the spacecraft via the tether producing the desired drag. In the charge exchange process, the thermal ions are converted into neutrals and leave the magnetosphere with a random direction reflecting their thermal distribution prior to charge exchange. The black box picture would then show thermally cold but highly directed neutrals entering the box on one side, and warm neutrals leaving the box in all directions. The box thus experiences a large reactive force on it, the drag force. 

Slide 5

It should be noted that the plasma MagnetoShell can operate at higher altitude and thus much lower densities than the traditionally conceived aerocapture systems due to the MagnetoShell’s collection radius being so much larger than the spacecraft itself. In fact, it must be so in order to minimize interference from the spacecraft on neutral particle collection. Operation in the high atmosphere is actually quite advantageous for several reasons. The frictional heat is deposited into the plasma rather than the spacecraft, which helps to sustain the PMS. The structural load on sensitive components like solar panels is kept to a safe level and it avoids risky low altitude trajectories. The plasma MagnetoShell is also a highly dynamic control surface that can be regulated through either the magnitude of the magnetic dipole field or the rate of plasma insertion into the dipole field. With the ability to modify the drag rapidly and precisely in varying atmospheric conditions, much larger braking forces can now be contemplated at low risk enabling very aggressive aerocapture maneuvers. As a result, the plasma MagnetoShell resolves the tradeoffs between current orbit insertion techniques. It offers the dynamic control of modern electric propulsion while providing the mission enabling mass and cost benefits of aerocapture. It should also be noted that the plasma MagnetoShell can be deployed as a shield against solar radiation and solar flares if need be.

Slide 6

In order to obtain a quantitative estimate as to the breaking force that a plasma MagnetoShell system of a given size can provide, one needs to understand the key physics and scaling for the concept to work as desired.  The list of parameters would include the following. Collision length: The incoming neutral densities must be low enough so that charge exchange will occur within the PMS of scale L.

Ion Magnetization: The magnetic field and velocity must yield magnetized CEX ions (and source ions) with gyro-orbits less than the size of the PMS.                

Bohm confinement timescale: Unlike most astrophysical plasmas, this plasma is neutral collision dominated, leading to Bohm diffusion and was observed in experimental testing that will be discussed later.              

Thermal equilibrium: Thermal equilibrium controls plasma density and effective radius thus drag. Incoming particles result in hot ions which heat electrons. Ionization cools electrons, The relaxation rate between the ions and electrons is thus an important parameter

Plasma Density: The plasma densities must be high enough that the PMS is collisional during entry, otherwise neutral particles are unimpeded. Using the empirical cross section for charge exchange for a given atomic species and electron temperature, the charge exchange reaction rate equation can be calculated for each radial position averaged over a chord length.

Slide 7

In order to determine the equilibrium plasma and spatial distribution of a PMS under a given set of incoming planetary atmospheric flow conditions, a 2-D transient model was developed. This model uses a cylindrical axisymmetric approximation. This cylindrical approximation has an aspect ratio of 3 and assumes uniform chord properties. This assumption is believed to be accurate enough as the typical mean free path will likely be a large fraction of the PMS scale length as only a few collisions will happen along a given cylindrical chord. Assuming uniform azimuthal plasma conditions, the radial profiles of ion and electron densities and temperatures can be calculated. A uniform incoming neutral flow distribution is assumed.  Two more assumptions are made in the model; the plasma has little effect on the magnetic field, so a fixed magnetic field profile is assumed, and collisional cross sections use several empirical models to determine the inter-species interaction of neutrals, ions and electrons.

The model also uses the atomic species profile most appropriate for a given planetary atmosphere. For Earth, a Nitrogen atom is assumed. For Mars, an Argon atom, while for Neptune, Hydrogen is assumed. As the mean free path is large, a non-Maxwellian, non-equilibrium plasma condition is assumed in which the plasma ions and electrons do not automatically relax to equilibrium states and equilibrium temperature distributions. The additional equations for all the parametric relation needed to complete the analysis can be found in the paper.

Slide 8

As the set of equations that characterize the processes in the PMS are time dependent, the time evolution of the PMS does not necessarily guaranty that there will be a stable equilibrium if the feedback from environmental conditions is not well posed. Fortunately, this was not the case. Several features of the equilibrium are as expected. First, given sufficient time, ion temperature will equilibrate reflecting the average directed energy of the incoming flow. For typical aerobraking applications, incoming neutrals have directional energies 5 to 30 eV. Electrons equilibrate to a temperature somewhat lower, depending on the specific ionization rate. Ion density will equilibrate to a fixed value based on the incoming neutral ionization rate and the diffusion rate. For incoming flow energy that is less than a typical ionization energy (12 to 20 eV) particle density will decay at the diffusion rate. When the density is sustained by limited ionization at flow energies greater than the ionization energy, significant increase in ionization occurs, resulting in an expansion of the PMS until a matching radial equilibrium condition is met in which ionization rate and diffusion rate are equal. Density distributions evolve over the time scales of the confinement time, which for these devices and typical operating conditions will be many minutes or longer

Slide 9

While the 2D analytic model was useful for obtaining the equilibrium and optimization of the Plasma MagnetoShell, a particle-based calculation was used to accurately measure the drag from the flow of neutral atmospheric particles. This slide shows the results from these calculations  The Y axis reflects the ratio of the drag intercepted by the plasma MagnetoShell normalized to the drag from the particles colliding with the dipole coil alone.  The ratio is also stated in terms of the variables that  determine these drag forces. The X axis is the ratio of the gyroradius based on neutral flow energy to that of the coil radius.  Th is a critical parameter as it reflects the effective magnetization of the charge exchanged atmospheric neutral.  As can be seem, the  PMS drag could be increased  to several thousand times larger than that of the dipole coil itself. These values, however, are obtained with the largest magnetic dipole fields as the rest of the variables that characterize the drag magnitude scaling are determined by the location and velocity of the spacecraft. The particle tracing figures inset in the figure indicate what is going on here. As one turns down the dipole field magnitude, the drag is eventually produced only by trajectories very near the coil itself. This is because the dipole magnetic field vector is primarily aligned with the flow when the dipole axis is orientated in the same direction as the flow. It is however only the transverse magnetic field component to the flow  that deflects the highly directed ions. Unlike gravity, the Lorentz force acting on these ions is a non-central force, so that the magnetic field can only deflect the ion, not slow it down. Clearly, while a transverse (to the flow) orientation of the dipole axis  would present to the flow the largest transverse deflecting field for a given magnetic field energy, almost any significant axis tilt would significantly reduce the pass-through ions . This calculation would however require a full 3D code which was beyond the scope of this study.  

Slide 10

And in addition to the transverse to the flow orientation of the dipole axis. It was found that the addition of two coaxial dipole rings at significantly larger radius from the dipole and antenna set shown in the figure can considerably enhance the plasma uniformity and density surrounding the main, inner dipole. The field requirement for these additional “outrigger coils” is quite small as the falloff in field for even the high beta dipole is still 1/r2 so the that the confining field from these coils need be no greater than that from the main dipole field at that radius. In addition, the outriggers provide for a straightforward and rapid means to change the drag force to maintain the proper deceleration in changing atmospheric conditions without attempting to accomplish this through varying the current drive input power and plasma density, which could prove to be slower and possibly more complex than desirable. This optimal version of the PMS was demonstrated in experiments as shown in the figures on the right. A Rotating Magnetic Field formed by an orthogonal pair of hoops with phased oscillating currents was  used to generate and sustain the Argon plasma seen here. The success in these experiments made this design the basis for the PMS to be employed  in the validation  tests performed in the large vacuum chamber at MSNW. 

Slide 11

Several experimental tests were made of the PMS concept to better understand the relevant physics as well as to develop techniques for obtaining the proper plasma density and methods for sustainment. Given the optimal sizes for space, it was clear that the experimental program would have to be a subscale test. Fortunately, MSNW had a large enough vacuum chamber and pumping speed to carry out a meaningful test. It was also equipped with a thrust stand and an array of power and diagnostic feedthroughs As was previously discussed, a prototype PMS device had been shown to be a suitable a candidate PMS testing in the large chamber on a thrust stand. The subscale experimental testing presented several challenges. In order to have the CEX/collisional mean free path (mfp) to be on the order of the much smaller PMS, the plasma density must be proportionally higher. Another challenge was obtaining a suitable neutral flow at the characteristic speeds one would encounter on reentry into a planetary atmosphere. These challenges were successfully met. An arcjet was employed to produce a supersonic neutral flow with only a small ionized component. A pulsed, fast gas puff system was also employed on the arcjet as well as directly into the mini–PMS In this manner the background density of  less than 10-5 Torr was maintained to prevent attenuation of the drag and maintain a collisional environment similar to what would be encountered in the upper atmospheres. The impulse from the Argon flow from the arcjet could be readily measured on the MSNW torsional thrust stand, which has been shown to be able to resolve 100 micro-Newton forces. Calibration was accomplished at atmosphere with a standard DC, weight calibration system. In vacuum, calibration is accomplished with a standardized, AFRL-calibrated cold gas thruster. The arcjet neutral flow and density were measured by a pair of Langmuir probes along with a time-resolved fast ion pressure gauge.   

Slide 12

The Plasma MagnetoShell was tested with Argon gas under a variety of operating and gas flow conditions. By measuring the increase in the deflection of the torsional thrust stand, the impulse could be measured. By comparing the average exposure time of the stand to the various impulses, the approximate average drag could be determined. It should be noted that displacement of the stand from operation of the electromagnetic and gas feed systems was observed to be negligible.  The Plasma MagnetoShell demonstrated a dramatic increase in effective drag over the interaction case with arcjet heated neutral flow and the PMS magnetic dipole active, but no RMF plasma. With the addition of the RMF plasma to the PMS, over a thousandfold increase in drag was observed thereby demonstrating charge exchange and neutral entrainment, as well as validating the PMS concept.

Slide 13

It is worth mentioning that it is possible to calculate the power requirements for the sustainment of the PMS plasma. One can calculate the power needed by looking at the azimuthal current density driven by the RMF. It is simply the integral of eta  J squared over the plasma volume. From flux conservation one can calculate the current density. Performing the integration over the PMS volume one can determine the scaling for the RMF power system. It can be seen that the power increases with the PMS radius, however the drag will increase as the radius squared. This is good news as it means that the larger the PMS, the more efficient it will be in producing drag. 

Slide 14

A fair amount of effort was expended in examining  the scaling and optimization of the plasma MagnetoShell component hardware. For example, the scaling dependencies for both electrical power and magnet mass are shown for the generation of a 250 Gauss dipole coil in the figure on the left. To perform the optimization requires knowledge of the orbital trajectory, the analysis of which was carried out for two missions: a Neptune orbiter and a manned Mars orbital insertion. The complete system design of both missions including PMS coil, tether, power processing, and battery mass were scaled and specified. The orbital simulation code Copernicus was used to determine optimal trajectory as well as optimize the magneto shell size as a function of altitude. The results for several key PMS elements is shown on the right for the two cases.

Slide 15

It is also worth mentioning at this point how the experiment compares to the actual requirements for a manned Mars insertion. That is shown in this table. As can be seen,  several of the parameters achieved in the experiments are quite similar to what would be needed for the manned Mars insertion. Of course, due to the large difference in size,  parameters such as radius, dynamic pressure, deceleration force will be quite different.  It can be seen that the power requirements are still require reasonable, and miniscule compared to what would be required for braking with conventional thrusters.

Slide 16

The results from the Copernicus calculation of the Martian aerocapture are shown here  for a 60 ton spacecraft inserted via a simple Hohmann transfer. The atmospheric model used came from the standard Mars atmospheric climate database version 5.0.  An at the outset a large Plasma MagnetoShell is required to perform the initial insertion and capture. In this simulation, a 10-meter radius PMS is employed. The initial capture is performed over the course of five minutes followed by a long coast period of 8 days between aerobraking maneuvers. This allows for recalculation of orbital trajectories and adjusting for the dynamic atmosphere. An 18 orbit aerobraking maneuver is then performed to circularize and lower the Mars orbit. The results for the Martian orbital braking maneuver on the right indicate a peak braking deceleration of 0.6 g’s with a dynamic pressure of 130 N/m2.  The calculated ion temperature for this braking maneuver is also shown.

Slide 17

In conclusion then, it has been shown that the Plasma MagnetoShell aerocapture concept holds great promise for providing huge mission mass, time and cost savings. It would reduce hardware mass by over an order of magnitude from conventional aerocapture structures. The drag is controllable in real time allowing for much more aggressive braking maneuvers. With no spacecraft or aeroshell heating concerns it can come in “hot” thereby saving transit time as well as fuel.  As a non-interfering add-on, it can be easily integrated with existing spacecraft as well as other entry technologies.

Slide 1

In this talk I will be discussing the critical issue of astronaut radiation exposure to Galactic Cosmic Radiation, or GCR on long interplanetary missions. It is believed that  we have discovered a suitable magnetic field topology for this shield which we refer to as a Dipolar Toroidal Magnetosphere (DTM). The shielding characteristics and advantages of this approach will also be described and compared to past efforts. 

Slide 2

It is worth a look at the actual doses one experiences for typical activities on Earth and in near space to get a sense of how damaging interplanetary travel is in terms of the exposure. You can see that the estimated Mars trip is orders of magnitude more radiation than has been experienced in space up to now. I should point out the green box on the lower right corner. A REM takes into account the effect of radiation on humans and is the same as a rad of x or gamma rays . 100 rads are equal to one joule per kilogram which in SI units is a gray. 

Slide 3

From the previous slide, it is clear that Gaactic Cosmic Rays, or GCR, are clearly one of the most, if not the most, significant biological risk for human exploration. The GCR are comprised of particles born and accelerated to relativistic speeds in supernova remnants. They include almost all the elements in the periodic table, and are bare nuclei fully stripped of electrons. They’re very high energy. Most are  100s of MeV to GeV. There are continuous and essentially isotropic. Coming from an all directions.

Slide 4

As can be seen in the graph at the left. The energies of the GCR extend well beyond GeV range, however flux falloff is exponential for these high energy particles. And while higher mass particles can do more damage, they are much fewer in number. Still, as the graph on the right indicates, their contribution to the total radiation dose in not negligible. Unlike material shielding. A great advantage of employing a magnetic shield is that a magnetic shield has the same deflection effect on particles of the same charge to mass ratio, which is essentially a constant for all the GCR nucleons. So that if you can deflect Helium GCSR away, you can deflect iron GCR away with the same magnetic field strength. It should be noted that while it would be impossible to shield against all GCR , the concern is with. particles of in the range of several hundred MeV to a few GeV, and deflecting these would reduce exposure to an acceptable level.

Slide 5

It is also worth noting that the Earth’s magnetosphere offers evidence of some protection against GCR from magnetic fields. Although the field strength of the magnetosphere is only a fraction of the Gauss. It is quite large – several earth radii, and it will be shown that its the product of the field strength and scale length of the magnetospheric field that is important. This product is significantly increased during a solar maximum due to the increased pressure of the solar wind and it can be seen that the actual GCR reaching the Earth is diminished. However. given the weakness of the field, it can only significantly limit the lower energy GCR reaching the Earth. The actual shielding at the surface of the Earth comes from the atmosphere itself. This type of shielding is generally referred to as passive material shielding.

Slide 6

Since passive material shielding works on Earth, why not use it in space? As can be seen by the above graph, in order to effectively shield against. GCR using a passive material shield, the amount of material is so great it would require at least five heavy lift launches to bring sufficient shielding material into low Earth orbit. The biggest issue with material shelling Is the fact that the interaction between the very energetic GCR and the material Is such that many daughter particles are created when the GCR particle strikes a material particle. These daughter particles are still in the MeV range and therefore actually increase the radiation dose. The shielding mass must now be ever thicker in order to stop and absorb daughter and all subsequent generations until the GCR energy is fully dissipated.

Slide 7

As mentioned earlier, it is the product of the magnetic field and distance that is needed to deflect particles. It should be noted that, unlike the electric force, the magnetic force is a non-central force. In other words, it induces a motion that is perpendicular to the magnetic field direction. So you cannot stop a GCR particle with a magnetic field. You can only deflect it. When the GCR particle interacts with the magnetic field, it begins to be deflected. In a spatially constant B field. The particle would actually orbit the magnetic field with a radius as stated here. For shielding, we are only interested in a deflection of less than 90 degrees as it moves through the field. It turns out that when you look at the deflection of a GeV per nucleon particle, it requires something on the order of a 10 Tesla meter field. It should be noted that the structure of the field is very important as an improperly chosen geometry for the field can just as easily deflect a GCR into the spacecraft as well as away from it.

Slide 8

Given the fact that magnetic fields could be useful in shielding spacecraft from the GCR, several magnetic field-based concepts were developed with a wide range of magnetic topologies. The two most promising concepts at the time, as well as the most developed in terms of both the physics and the required technologies for implementation are shown here. The European Space Agency developed a high current density toroid that surrounded the ESA proposed Columbus habitat  as you can see on the righthand side of the slide. The other was the result of a NASA Innovative Advanced Concept study where six long and large inflatable coils were to surround the habitat and a seventh coil providing a reverse directed field to buck out the return flux from the outer six solenoids.  I will refer to these concepts as NIAC 6-1 and ESA Toroid. Clearly both suffer from a serios lack of magnetic shielding from GCR coming in from the axisymmetric polar regions. There are even bigger issues for these concepts that rely on closed fields, that is deflecting fields that are primarily only inside the field generating coils .

Slide 9

A plot of the magnetic field vectors in the symmetry plane for the ESA toroid on the left, and a cross section of the NIAC 6 – 1 solenoids is shown on the right. It can be clearly seen that the magnetic fields outside the coils are quite low and that there is no really significant field along the polar axis, and what is there provides no deflection as it is aligned with the direction of the incoming GCR..

Slide 10

There is, however, an even bigger problem for both concepts as is illustrated in these bar graphs. Since neither the ESA nor the NIAC 6 – 1 solenoids have any significant magnetic field on the spaceward side of the coils, these systems are completely vulnerable to all GSR, even the lowest energy ones that would be deflected by the shield. They strike the both magnets and supports structures which are much larger than the habitat they’re trying to protect. This interaction with these elements lead to the creation of daughter particles as described earlier in the discussion of passive. material shielding. As you can see. that the primary GCR, exposure,  denoted by the blue regions in the bar graphs, reaching the Columbus habitat in this case. is significantly reduced by the magnetic field-distance product- at least for the GCR entering from the barrel region of the toroid. The daughter particle creation, however, is so significant that it actually intensifies the exposure. Increasing the magnetic field continues to diminish the primary GCR as higher energy GCR are deflected , but there is little improvement as neutrons and gamma rays are. Essentially unaffected by magnetic fields and only decrease by limiting even further the entry of higher energy primary GCR. To accomplish this requires High Temperature Superconducting technology well beyond current technologies.

Slide 11

Fortunately, there is a solution to the magnetic shielding of spacecraft that avoids the problems that were encountered by the ESA and NIAC concepts.  It is referred to as the dipolar toroidal magnetosphere, or DTM. Here the habitat is shaped as a toroid ala 2001: a space odyssey if you like. Here also the HTSC currents flow along the outer surface of the toroid in the toroidal  direction. This is opposed to the usual magnetic toroid winding, such as the ESA toroid, where the currents flow in the poloidal direction. The high temperature superconducting magnets are placed external to the habitat and can thus be supported by the spacecraft itself. By adjusting the magnitude of the toroidal currents that surround the habitat one can make the magnetic fields  inside the habitat to be negligible and at the same time provide a deflecting magnetic field that is perpendicular to all incoming GCR regardless of direction.

Slide 12

In order to evaluate the Dipolar Toroidal Magnetosphere as a shield, a computer code was written that would allow for tracking relativistic particle motion from charged particles (GSR) as they enter the field of the DTM isotropically from all directions. This was done at a scale that would be appropriate for a potential interplanetary spacecraft. As can be seen, having the currents flow in just the right way on the surface of the toroid, it is possible to eliminate all the magnetic field inside the spacecraft while at the same time creating a large magnetic field external to the spacecraft and importantly, to do it in a manner that provides equal protection for the HTSC magnets as well. The next slide will show linear plots of the magnetic field along the two dashed lines and denoted by figure A and figure B on the slide.

Slide 13

As the toroid aspect ratio, that is the diameter of the toroid compared to the cross section of the toroid becomes large. The magnitude of the toroidal currents as one goes from inside to the outboard side of the torus more uniformly. For a typical aspect ratio of 10, a large current must flow along the inner wall as shown in figure B to keep the flux inside the spacecraft at 0.

Slide 14

The results from the initial testing of the DTM were very encouraging. As you can see in the graph, there is absolutely no penetration into the spacecraft from the one Gev particles. 1500 proton particles were used in this numerical calculation emitted from a cylindrical boundary at 30 meters. The apparent lack of particle traces one sees at small radius just reflects the fact that the volume goes like R-squared and so in a 2D R-Z plot it looks as though there is a dearth of particles when in fact it is just the r weighting.

Slide 15

It has been difficult to determine just how good a given concept is at shielding since each concept has previously been evaluated for both the fields and technologies required in ways unique to the particulars of the concept. It was decided that in order to make comparisons, a set of criteria needed to be developed that would allow for the unbiased evaluation of all concepts across the board based on normalized parameters common to all concepts. For the effectiveness of the shielding, the best  parameter for evaluation of shielding efficacy was the total HTSC mass required to stop at least ¾ of the GCR of a given characteristic energy per unit volume of habitat.  Without specifying the HTSC material, the amount of mass will be proportional to the sum of all the currents required times conductor length. This number divided by habitat volume for that particular concept was used and denoted by the symbol, capital omega. The second criterion has been discussed. And is clearly a key criterium. Amount of total mass, scale of structures and complexity of assembly  must also be considered.  As the cost of access to space is still a determining factor in concept feasibility. Finally, there may be other advantages for the shield outside GCR protection, given the likely high cost of the shielding systems. What other uses for the shield might be? Might there be besides crew protection that would help justify the investment in shield deployment. Not surprisingly. The DTM has great advantage in all four criteria. Over. The other two concepts. As mentioned already. The issue with secondaries is. mitigated by the fact that the. magnetic field of the DTM shields both the magnets and the spacecraft. In terms of launch and space assembly, the DTM Magnets can be incorporated as part of the structural  elements  of the spacecraft itself. There are in fact other applications for the DTM besides shielding. that will not be discussed here due to time constraints, but are presented in the paper that describe using the DTM for aerobraking in planetary atmospheres and thereby significantly reduce the fuel and mass requirements for these missions. 

Slide 16

I show this slide just to indicate what the analytic analysis of the dipolar toroidal magnetosphere yields as a function of the magnetic spatial and GCR energy parameters. The results from the detailed numerical calculations follow this scaling closely and were actually  somewhat more favorable than the criteria shown here. 

Slide 17

A quick slide just to show. The various scales and geometries of the three concepts that were used in the GCR shielding effectiveness numerical studies.

Slide 18

Again, a quick slide just to show the direction of the current flow for generating of the magnetic fields employed in the three concepts.

Slide 19

This slide to show modulus be along the primary axis for the three concepts. It is clear from this that. The DTM has by far the largest fields external to the habitat. The key to avoiding. Secondary particles.

Slide 20

Shown here are the spherical surfaces where the GCR particles are randomly and Isotropically emitted for characterizing the shielding effectiveness of the three concepts which can be seen inside the semi-transparent spherical. launch surfaces.

Slide 21

The Lorentz 3D code from Integrated Engineering was used to accurately determine the GCR particle trajectories. These calculations were performed using thousands of test particles. What’s shown here is the ray tracing for just a few, showing in this instance how they are deflected by the magnetic field from the DTM.

Slide 22

For the shielding effectiveness comparison, 500 MEV protons were employed which characterized the peak of the GSR spectrum which was 200 to 500 MeV. The fraction, f, that reached the habitat compared with the number that were launched and struck the habitat with no magnetic shielding. This is plotted as a function of the HTSC mass per habitat volume It is clear from this plot. That the DTM is overwhelmingly more effective. Neither the 6 – 1 solenoids or the ESA toroid could reduce the GCR to zero regardless of the current magnitude. This is primarily due to the fact that there is unshielded areas at each end on the polar axis. In fact, the ESA toroid plateaued before reaching a the ¾ reduction criterium.

Slide 23

With the comparison done and the winner determined, I will now discuss only the results of the DTM and how it would perform using the full spectrum of the GCR. What is plotted on the lefthand side is the magnitude of the GCR proton flux. It is plotted now on a linear scale rather than in the log-log scale shown earlier, as it is far more realistic in representing the bulk of the GCR and how much is actually found at a given energy per nucleon. Recall the analytic condition shown earlier for the full deflection of the DTM. The result from the numerical calculation and experimental results which I will not discuss here but can be found in the paper, the constant of proportionality, C sub-B, has a value of roughly one. The plot on the left and be replotted in terms of the relativistic parameters gamma and beta. Which is shown on the right. Here it is assumed that the nucleon mass is that of the proton or 0.94 Gev.

Slide 24

The magnetic shielding results employing the DTM for the biggest and principle, bad actors, that being hydrogen, carbon and iron, can be found in these three plots. It can be seen for a highly reasonable on-axis central magnetic field strength of 0.64 Tesla, a very significant attenuation of the GCR can be obtained for all three elements. In the case of GCR protons, which have the highest flux of any GCR element, an increase of magnetic field to just one Tesla on-axis reduced the proton GCR to a negligible amount over the region of most concern.

Slide 25

In summary, a magnetic shielding configuration was discovered that can deflect the vast majority of the GCR. It is referred to as a Dipolar Toroidal Magnetosphere., or DTM The DTM shields both habitat and magnetic coils, significantly reducing secondary particle irradiation hazard. The DTM magnetic shielding can be mounted on the spacecraft reducing structural mass and complexity. A low-cost method for evaluating the shielding effectiveness on Earth was realized with a subscale DTM test using a surrogate ion beam in a large vacuum chamber.

Slide 1

In this talk I will be discussing the possibility of a refueling orbital navigator, or RON. The two critical issues for in-space operations that would be addressed and potentially solved by the refueling orbital navigator concept: those being Active Debris Removal or ADR and the In-space maneuverability, The latter representing the possibility of providing mobility and relocation options for current space assets as well as in space repair and reconstruction of disabled assets.

Slide 2

There are many issues associated with ADR. Most important is the economics of such a mission. Due to the high ΔV requirements, conventional chemical or propellant based systems become implausible. To remove the significant number of debris needed to stabilize the space environment, a reusable system is necessary. This demands a specialized spacecraft that can maneuver between many target objects, capturing each and transferring them into terminal orbits. Such an Orbital Transfer Vehicle (OTV) requires the ability to perform hundreds of km/s in ΔV over its mission lifetime. However, even with the most fuel-efficient, high-Isp device, total propellant mass scales prohibitively for long, multi object deorbiting missions. As a solution to this problem, a number of propellantless technologies have been suggested for ADR. Due to the lack of success thus far, and the fact that extensive tests in the actual space environment still need to be performed, it appears that these large structures are not the best near-term solution. Moreover, long range solutions, such as lasers, can easily yield significant amounts of additional debris if target conditions and attitude are not well known. The alternative is to employ an air breathing propulsion system. Such a system will require (1) a low drag spacecraft, (2) a high collection efficiency scoop, and (3) a high Isp, high efficiency thruster. How these three criteria can be met with the RON spacecraft will now be discussed.

Slide 3

There are many challenges associated with in-space maneuverability due to the high Delta V requirements. The Refueling Orbital Navigator (RON) spacecraft accomplishes this in the following way: it employs a highly elliptical collection orbit, or ECO, to obtain fuel by scooping up and storing the molecular nitrogen and atomic oxygen encountered during the low altitude perigee periods of its orbit. As RON can thrust at apogee, it can achieve the extensive orbit lowering needed for ADR. Additionally, RON can thrust at perigee to provide drag compensation for low perigee refueling, stable non-Keplerian orbits, or rapid phase changes. As such, the RON concept can also find application to other planetary missions.

The ability to capitalize on In Situ Resource Utilization (ISRU) for electric propulsion on the next generation exploration class missions would have a direct impact on potential planetary mission feasibility, scale, and cost. A refuellable orbital navigator employing electric propulsion would enable detailed studies of the planetary bodies from Venus out to Neptune. For the ADR mission, chosen for the focus of this study due to its extreme importance, the requirements for the RON are modest. RON requires only 5 kW of on-board solar power as energy collected during the higher altitude portions of its elliptical orbit can be stored for higher power operation later. Operating in this manner RON would generate 1.2 N of thrust with a specific impulse, Isp, of 2500 sec. for ADR. During a 10-year mission life, RON would process 5500 kg of propellant to ferry 80 spacecraft, perform 850 degrees of plane change, with over 100 km/s of delta-V, all with a single spacecraft launch, and requiring no additional onboard propellant. The ADR mission and how RON would enable an affordable method for reducing orbital debris will be outlined

Slide 4

This is clearly not the first attempt  to build, design and fly an air breathing spacecraft capable of sustained propulsion in the upper atmosphere. A major reason, and a big difference between the RON spacecraft operation and previous airbreathing concepts is the Elliptical Collection Orbit, or ECO, employed on RON. For the usual airbreathing spacecraft concept, a circular orbit was assumed. This demands a propulsion system that can operate in a steady manner to overcome the constant drag of the propulsion system and solar array One finds that even for a fairly efficient Electric propulsion system, over 30 kW of propulsive power would be required. When one includes the size of the solar array required, the drag is significantly increased. More drag means more power, and more power means more solar array mass and thus more drag – this ends in an ever-increasing upward spiral. The way to break this cycle is the ECO.

Slide 5

To realize the revolutionary impact a Refueling Orbital Navigator could represent, consider this analogy. Up to now, objects in near earth space are like trees. You put them in a certain place and that’s where they stay until they die. What is being proposed here is more like an animal, perhaps a cat. It can move from place to place to do a wide variety of essential tasks. But this freedom of movement requires a lot more energy. So, like a cat, it needs to spend a fair amount of time doing nothing but recovering its strength while expending little energy. When it comes time to move, it devours its fuel voraciously. The ECO provides for just such a life cycle. As discussed, to overcome drag the power required for the propulsion system needed to 30 kW, 50% efficient at an Isp 0f 2000 seconds which is within the range of Electric Propulsion (EP) thrusters already developed for space application. In order to stay in orbit, the spacecraft thrust must be equal to or greater than the drag. Equating the two expressions for these key parameters shown at the top left, one can solve for the highest atmospheric mass density the craft can fly which is shown on the top right. Here V is the orbital velocity of the spacecraft, and A is the area of the particle collection scoop, which is assumed to be the largest obstruction to the flow.  It is not the area of the solar array as the 30 kW propulsive power is only required near perigee when the spacecraft confronts significant atmospheric densities, as can be seen in the graph at the lower right. This allows for the solar panels to be reduced  by 90% of what would be required to stay at a constant orbital altitude at 150 kilometers. After leaving perigee where it scooped far more fuel than is needed at high altitude, the spacecraft can rest and recharge its batteries. Just like the cat. 

Slide 6

I show this slide to indicate the kind of environment in which the Refueling Orbital Navigator will operate. The graph on the right shows the exponential fall-off in the number density of particles as one goes higher in altitude and explains the short time when the fuel accumulation occurs during the lowest altitudes. The gas composition also changes considerably as one goes further up in the atmosphere. At the target altitude of 150 kilometers, the primary constituents are roughly 60 percent molecular nitrogen, 30 percent atomic oxygen, and the remainder being small amounts of argon, helium and hydrogen. Ionized gases are negligible at less than 1 %. Unlike other planetary atmospheres, the Earth atmosphere heats up from the absorption of solar energy in the UV spectrum by oxygen species. At the target perigee altitude of 150 km it Increases dramatically. These temperatures are still negligible compared to the stagnation temperature one would reach from stopping the flow of particles entering scoop at the orbital speed of 8 km/s. This corresponds to a stagnation temperature of 90,000 degrees Kelvin.

Slide 7

Due to the extremely low number density, the thermal energy content is still quite low, and the spacecraft itself is not significantly heated by the fuel captured. The heat flux from the various thermal sources can be found in the graph on the left.  Compared to heating from solar rays the flow stagnation is the largest component at low altitudes. The waste heat from operation of the electric propulsion system, even at high efficiency. Dwarf the scoop heating. The graph on the right shows that at the altitudes that will be encountered by the RON, radiative cooling will be sufficient to keep the fuel capture system at reasonable temperatures.

Slide 8

It should be noted that the interaction of the atmosphere with the RON spacecraft is always in the regime of molecular flow. This simplifies the calculation of drag and the interaction of the scoop with the atmosphere as to the collection efficiency as well as the drag coefficient as the two are related. The most accurate calculations required the use of the Direct Simulation Monte Carlo code originated by Bird in the 1990s to describe the interaction of gas particles with material structures in space. Due to time constraints the analysis of the scoop performance, as well as experimental testing of a prototype scoop will not be covered in this talk, however both are fully discussed in the paper. To sum up those results, collecting all of the gas entering the scoop in the molecular regime, where one can essentially ignore all particle interactions with other particles compared to the interactions with solid surfaces such as the scoop, would lead to a drag coefficient of 2. This represents a completely inelastic interaction (i.e. stagnation) of the flow with the spacecraft. For less than 100% capture, the reflected particles not captured will increase the drag with the maximum being a drag coefficient of 4, the condition where all particles are reflected elastically as in striking a flat plate. Experimentally, the measured drag coefficient from a supersonic flowing gas in the molecular regime for the conical scoop array shown on the right was 2.2.    

Slide 9

Detailed calculations of the operational space for the RON were conducted with regard to several key parameters such as the spacecraft drag coefficient, thruster Isp, perigee altitude, and scoop capture efficiency. The following graphs will illustrate how the variation in these parameters affect the prime concern of the refuel time as this dictates mission feasibility beyond just drag makeup. A good one to start is the effect of altitude on thruster Isp as reflected in refuel time. This graph makes it pretty clear that dipping to lower altitude at constant (max) power leads to a reduced Isp as more thrust is needed to overcome increased drag and eventually results in a much longer refueling times. One can see why 150 km  makes for a good compromise that doesn’t demand a particularly high Isp compared to what can be achieved from current EP thrusters. The 10 square meter scoop with a diameter of just over 31/2 meters was chosen as a reasonable size that can readily be accommodated inside the fairing of a large rocket. A drag coefficient of 2 was also assumed.

Slide 10

These two figures indicate the range of refueling times based on the drag coefficient and the capture fraction. For the assumed drag coefficient of 2 and an altitude of 150 km, the minimum capture percentage is roughly 50%. Below that, RON is less able to take in enough ambient gas to perform the necessary drag makeup. It is also shown that at an altitude of 160 km, it would be advantageous to have a higher Isp of 3500 seconds. It is worth noting here that for high capture percentages, refueling times level off and are not reduced significantly with increasing Isp. The opposite is true at lower capture percentages. At 50% capture efficiency and 2000 second design point, it would take 220 days to refuel, while a 3500 s design would only take 68 days, a reduction of 69%. In fact, a high Isp design would allow for operation at lower capture percentage of 30 to 40%, if need be, but at the price of increased mission time. In this way though, drag can be made up while maximum mass efficiency can be maintained. These plots clearly illustrate why optimizing scoop design for minimum drag and maximum capture efficiency are critical for reducing mission time.

Slide 11

As of now, there is no clear winner as to the best EP thruster for the RON as most of the current thrusters have been developed to operate on Xenon. There is in principle no reason that the leading EP candidates – the Hall Effect Thruster, or (HET), and the RF based Ion Engine, both of which have been developed to TRL 9, could not be adapted to operate reliably and efficiently on atmospheric propellants. There have been recent efforts to do just that. As the studies are somewhat preliminary, it is difficult to get much in the way of performance data from these devices on atmospheric propellants. Operating a standard HET on atmospheric propellants was performed by a collaborative effort between ESA and Alta. The HET was operated on mixture of N2 and O2. A picture of the thruster operating on the mixture is shown at the left. The test results for both the HET and RIT thrusters showed a substantial deterioration of performance with respect to Xe. This is no doubt due to the low ionization efficiency of the atmospheric propellants. However, the results did indicate that both devices were capable of continuous and stable operation on atmospheric propellants. As can be seen in the figure on the right for the operation of the HET on the N2-O2 mixture, the thrust was a strong function of power. The maximum thrust, Th = 30 mN, was obtained at roughly 1 kW input with a flow rate of 2.7 mg/s. This was roughly half of that of the Xenon performance. If one assumes roughly half the efficiency of the Xenon results, one has that the Isp would be the target 2000 seconds. Since the target thruster power for the RON is 30 kW, these results are far from the operating conditions required for RON. Ionization efficiency and thruster performance quite often get better as the power is scaled up as attested even by the data shown here. It could very well be that the target of 50% efficiency operating on atmospheric propellants can be achieved at sufficiently high Isp for optimal performance of RON at the power levels desired. There are other candidate EP thrusters such as the Pulsed Inductive Thruster and the MAP thruster that naturally operate on lower mass propellants and could also be employed on RON . The advantages of these less developed thrusters are discussed in the paper.

Slide 12

As of January 2018, an estimated 8,100 tons of space debris had accumulated in low Earth orbit consisting of spent rocket bodies, mission-related debris, and collision fragments. With regard to ADR, the largest concentrations of high mass debris are found in the inclination band of 81 to 83 degrees and in the altitude range of 800 to 1300 km as shown in the figure. These large concentrations of high mass debris indicate that these regions of space should be the first targeted to reduce the overall collision probability in the space environment. A characteristic mission employing the RON to do just that was carried out.

Slide 13

As an example of the characteristic RON mission, the retrieval and disposal of a 1 metric ton Russian Cosmos class satellite at 950 km was studied. Each retrieval and disposal mission can be divided into the four stages shown on this slide. Stage 1: RON will rendezvous with the target. During this stage RON will assess the dynamics of the debris and the best way to interact with it. More will be said of this later as well as in the paper. Stage 2: RON will ferry the target to an elliptical termination orbit using onboard fuel. Once there it will release its cargo and start firing at perigee to perform drag make-up. For a 1 MT target and an estimated 500 kg RON, this ferrying maneuver will require 15 kg of propellant using a 2,500 sec Isp thruster. At 1.5 N thrust, burning at 10% of the orbit period, RON will take approximately 9 days to ferry to the necessary terminal orbit. Stage 3: RON will refuel for several orbits using ambient propellant to counteract drag and store fuel for future use. Reorientation maneuvers can also be done from this terminal orbit. Inclination changes and right ascension plane changes require high ΔV, and therefore have a high propellant requirement, and this is why RON is so well suited for ADR. Stage 4: RON will exit the terminal orbit and rendezvous with the next target. As the next target is at a similar altitude, another 223 m/s of ΔV will be required. It is estimated that to move RON into a proper orbit and new position will require 3.4 kg of propellant over 2 days

Slide 14

Putting it all together, the final prototype design for the Refueling Orbital Navigator was arrived at and is shown in the figure above. The total atmospheric mass impinging on the scoop collector array during each elliptical collection orbit Is 88 grams. Of this, 18 grams are reflected leaving 70 grams collected which is compressed and stored in the storage tank reflecting the 80% efficiency of the scoop. Of the 70 grams collected. 29 grams must be supplied to the EP thruster for drag make up. This leaves 41 grams stored from each cycle to be used to perform the spacecraft maneuvers for ADR or whatever mission is desired.

Slide 15

An initial mass estimate was formulated by including all the traditional spacecraft systems and by estimating the mission-specific systems. Some of the unique components include the atmosphere-collecting system and a target capture mechanism. The solar arrays and batteries were sized by requiring roughly 40 MJ to run the 30-kW thruster, while including extra energy to operate the other subsystems. An extra 25% of the total mass was added as margin for this early stage of the design.

Slide 16

Once a quick and effective orbital transfer method has been devised, there still remains an important technological challenge to overcome: the transmission of momentum from the removal system to the space debris in order to carry out the deorbiting maneuver. Several active debris removal concepts have also been proposed in debris mitigation including harpoons, magnetic darts, robotic arms, adhesives, scoops, interface probes, applied fields, nets, and plasma beams. The most obvious way is to dock the removal system with the target before the deorbiting starts. This operation can, however, be technologically complex and very risky. Space debris are noncooperative objects generally characterized by a problematic attitude motion (tumbling motion, flat-spin rotation, large amplitude oscillations, etc.) making docking dangerous and/or unfeasible. During the initial survey conducted of active debris removal methods, there was one particular approach having special merit. It is referred to as the Ion Beam Shepard (IBS) [25]. It is unique in avoiding many of the problems previously described, but it is especially advantageous when one considers the fact that with RON one has an essentially unlimited fuel resource for this task. The Ion Beam Shepard (IBS) is aptly named as it is what it says it is. The IBS employs a highly collimated, high-velocity ion beam, produced on board the spacecraft orbiting in close proximity to a target debris object, which is directed against the target to modify its orbit and/or attitude with no need for docking.

Slide 17

In closing, it should be mentioned that another important application of thew Refueling Orbital Navigator is Exploration and ISRU Utilization  on other Planets. Virtually all planets, as well as several planetary moons of interest, have a substantial atmosphere. Therefore, it goes without saying that a spacecraft that could be sent to these bodies, that had the ability roam virtually anywhere over the surface of the planet taking data, sending out probes, collecting samples, etc. and limited in time only by the survivability of the spacecraft hardware, would be a game-changing advance. A discussion of this exciting prospect of RON deployment can also be found in the paper.

Slide 18

As a summary, I repeat the introductory slide for your perusal as talk time has run out. Thank you.

Slide 1

In this talk I will be describing what I think most people would consider to be a somewhat of a radical, and possibly crazy notion. I will be discussing the use of fissile materials such as Uranium 235 and Plutonium 239 to directly heat melt, vaporize and ionize a solid lithium propellent. This is not a bomb scheme like project Orion. The fissile  fuel  employed is at a much smaller quantity than one would use for either an explosive or a thermal reactor for that matter. The idea here is to achieve a high ISP as well as a high specific power adequate enough to have a manned spacecraft travel to distant planets in an efficient and cost-effective manner. This has not been accomplished up to now for many reasons which will also be discussed. The main purpose here is to show how a reasonable, scientifically validated  approach to using fissile material can lead to a breakthrough in space propulsion.

Slide 2

Fission-based power sources have been used in space as early as the Voyager missions. However, the use of nuclear power for manned space exploration has yet to be accomplished. Reactor-based (thermal or electric) propulsion efforts have been stymied by enormous launch costs and extremely long mission times, which endanger the health of the astronauts, and increase the risk of failure due to the need for complex space rendezvous and assembly. Long and complicated missions increase mission and human health risks dramatically. Also, having missions that go beyond a presidential term in office risks a lack of governmental support, as well as public interest and tolerance for the increased cost. The solution is a propulsion method that can dramatically shorten the trip time as well as reduce the amount of mass, and thus cost, required to be launched into space. These requirements can be met with a compact, high efficiency, single-launch spacecraft such as that projected for the Fission-Powered Pulsed Plasma Propulsion concept (F4P) to be discussed here.

Slide 3

In order to cut trip time for a direct mission to Mars, it is worth considering what governs the time it takes to travel directly to Mars. Accelerating the spaceship of mass M over a time Tau implies a power given by the formula shown where one travels at a characteristic velocity that depends on alpha where alpha is the specific power which is the jet power divided by the spacecraft mass. It is also the inverse of the specific mass if you prefer that unit of measure. Using these simple approximations, the trip time to go a distance L can be expressed by the formula shown above in terms of the specific mass and the distance of the trip. As you can see, the dependence of trip time to cover a fixed distance is only weakly dependent on the specific mass. The trip time can only be made better by a dramatic reduction in the specific mass as it scales very weakly on this variable to the one third power.  The very favorable trip scenario shown in the figure is for a 90-day transit to and from Mars, where each transit length is roughly about 1 1/2 astronomical units, that is the distance from the Earth to the sun. This mission would require a specific power greater than 400 Watts/kg. If one considers an even shorter time of 30 days for the Mars transit, while it does shorten the trip distance because of the higher velocity required for the spacecraft and can therefore catch up to Mars more quickly, it requires a specific power greater than 2.5 kW/kg. It is believed that with Fission-Powered Pulsed Plasma Propulsion these target specific powers can be achieved.

Slide 4

The total mass for the Mars mission is dependent on many factors, the primary ones being mission duration, length of stay, and type of propulsion employed. These are nicely summed up in this figure taken from the NASA DRA-5 Addendum #2. It is evident from this analysis that long duration missions using either thermal or nuclear-electric propulsion can reduce the total mass in LEO to something that is less than the entire space station. It is also evident that these missions will be of long duration and will thus require the most protection from Galactic Cosmic Radiation as this is the greatest threat to the health of astronauts on long duration missions.

Slide 5

With propellant storage and reactor mass being significant issues, only minor reductions in trip time and in-space mass and assembly are predicted. As an example of the huge amount of material required when employing either nuclear electric or nuclear thermal propulsion, I show you a slide for the precursor cargo missions required to bring up enough material into space and for the follow-on crewed mission. You can see that this results in a very complicated and a very costly, time-consuming undertaking. While nuclear electric can achieve the much higher Isp desired as Electric Plasma thrusters can be employed, the low conversion efficiency of the reactor as well middling efficiency of high-power EP thrusters creates a massive heat rejection problem as attested to by the large arrays required to shed the heat. 

Slide 6

As a consequence of these very long trip times,  the exposure of the astronauts to galactic cosmic radiation, or GCR Is very significant. In fact,  it can be a showstopper for manned interplanetary space travel due to the higher risk for GCR-induced cancers. Only for short duration missions, the ones predicted for the fission powered pulse plasma propulsion scheme allow for the avoidance of significant radiation damage to human tissue.

Slide 7

Si it is abundantly clear that the optimal Mars mission will not be a chemical-based or solar electric propulsion system, and a nuclear based system must do better than the current nuclear thermal and nuclear electric systems to avoid being put off as before due to time and cost . As an analysis tool it is useful to consider what attributes would comprise the ideal nuclear propulsion system. Such a system would satisfy the following criteria:

(1) all the nuclear energy is released into the propellant, not requiring conversion to electricity.

(2) The propellant requires no significant tankage mass by being a solid, and where

(3) the propellant is rapidly heated and accelerated to optimal exhaust velocity (> 20 km/s), with

(4) no significant physical interaction with the spacecraft thereby limiting thermal heat load, spacecraft damage, and radiator mass.

Such a system would drastically reduce the mass needed in LEO as well as operate at the highest alpha. It is also clear that nuclear electric fails on criteria 1 and 2 primarily due to the thermal radiator mass penalty due to the reactor and thruster inefficiencies. Other issues, besides the tankage for a gaseous propellant. A large one is the lack of thrusters in the megawatt class that don’t have significant wall erosion issues (criterion 3). Nuclear thermal fails on 2, 3 and 4. The Isp for even the light gases (H and He) is too low and creates serious tankage mass issues as well. There are also the problems associated with contact with the propellant and the heat exchanger walls (embrittlement and corrosion). It is time now to describe the solution that satisfies these criteria.

Slide 8

As mentioned, fission-based power sources have been used in space as early as the Voyager missions. Here, the immense energy of the nuclear reaction is to be harnessed in a more direct and powerful way to propel a large, manned spacecraft. The Fission Powered Pulsed Plasma Propulsion concept or F4P seeks to tap the tremendous energy gain from a nuclear fuel such as Uranium or Plutonium directly into the propellant, which would most likely will be Lithium metal. The process that will be used to perform this task has been evaluated through analytical and numerical modelling in both 1D and 3D calculations and is discussed in detail in the paper. At the bottom of this figure the process by which the F4P produces the large jet power required for rapid manned space travel is illustrated as a series of several distinct stages which will now be discussed in detail.

Slide 9

The process for the F4P space-based propulsion system proceeds in the five stages shown here as follows: Three large-radius Lithium metal hoops (a.k.a. liners) are sequentially drifted into the chamber along the axis of symmetry until they are positioned directly underneath the fast electromagnetic coils as illustrated in the upper lefthand corner. Stage (1): These hoops (or liners), are then rapidly driven radially and axially inward by magnetic compression from highly-coupled, pulsed electromagnetic coils. At the same time, a very small amount, perhaps a few grams of fissile material, either Uranium 235 or Plutonium 239 has been previously positioned inside a Beryllium cylinder as envisioned in the blowup in the lower lefthand corner. This small capsule is injected on-axis into the thruster chamber as depicted. The Beryllium cylinder has three roles to play. First, it acts as a sabot, then as a containment vessel, and finally as a fission neutron reflector for enhanced fission gain. The nuclear fuel is in the form of a coil filament and is attached at each end to the Be sabot forming a closed loop electrically. Unlike the illustration, the fuel form would be more of a pancake coil like a Bitter magnet to better fill the space and increase the fissile mass inside the sabot. Stage (2): The Lithium shells rapidly converge on the intersection point with the nuclear fuel. As they collapse inward radially and axially, they continually thicken to finally form an impenetrable blanket several centimeters thick. The kg mass Lithium shell having attained a terminal velocity of ~3 km/sec rapidly compresses the leakage magnetic flux inside the shell surrounding the sabot which in turn compresses the sabot and nuclear fuel. Stage 3: The large radial magnetic gradient across the sabot and fuel induces a huge current flow first in the Beryllium shell and then in the fissile fuel. Plutonium and Uranium both possess a high resistivity so that the fuel rapidly melts and then vaporizes. The gas becomes partially ionized due to the large electric fields associated with the enormous dB/dt now experienced by the fuel. The Beryllium shell  initially confines this ionized nuclear gas as it has a much lower resistivity and a higher melting point. Stage 4: The Lithium shell at this point has also been vaporized and ionized as it has a much lower boiling point and is easily ionized. The picture for the fuel and Lithium propellant is now one of an ionized gas being compressed by multi-megabar magnetic fields. It is worth recalling at this point a few key nuclear fission concepts such as critical mass. This is the mass of fissile material at STP required to sustain the fission reaction indefinitely. For masses higher than the critical mass, the chain reaction becomes a runaway, and if the mass is kept confined during this process, it is a bomb. This is of course not quite what we are after here. The idea here is to turn the fissile mass into at least a partially ionized gas and compress it to much a higher density, thereby considerably reducing the fissile mass required to obtain large energy gains. With nuclear reactivity scaling as the density squared, the compression of the nuclear plasma results in a large and rapid increase in the release of nuclear energy. The compressed density is now a factor of 11 times greater. The critical radius is thus smaller by the same amount, or less than a centimeter, and the critical mass is reduced by the square of this ratio, i.e., a factor of 125 or just 80 grams. The potential fission energy generated for even this small amount would be over 6,000 GJ! Clearly significantly less fissile material will suffice. A yield of at least 100 MJ per pulse is desired to provide the target jet power range of 10-30 MW, with a rep rate of several seconds. In order to limit the reactions to a tiny fractional “burn” of the fissile material, the amount of fissile fuel can be reduced by a factor of 20 or more. At 4 grams and a fractional burn of just three hundredths of a percent would yield 100 MJ of fission energy. Virtually all of the energy released (plasma thermal, radiant, shell-ohmic, and fission-fragment) is absorbed by the Lithium shell which is subsequently vaporized and ionized. Stage (4): Recall that it is this Li plasma that acts as the propellant in this scheme as its bulk mass is orders of magnitude larger than that of the fission fuel. Lithium is very advantageous as a propellant as it is a malleable solid for easy onboard storage. It has a very low atomic mass, and as an alkali metal, it is easily vaporized and ionized. Stage (5): The large external region filled with a divergent magnetic flux by the thruster coils now acts as a barrier and magnetic nozzle, which protects the spacecraft while expelling the Lithium and nuclear plasma. In this manner, both high exhaust velocity and thrust are achieved with only minor interaction with the spacecraft. The flux compression of the vacuum magnetic field by the expanding plasma plume generates a large back-emf on the thruster coils. This allows for a rapid recharging of the thruster capacitors [7,8]. With this system, the F4P could potentially be self-powered requiring no electrical input from the spacecraft solar panels.

Slide 10

The implosion of magnetically driven thin metal hoops (liners) to achieve extremely high magnetic fields and thus extremely high compressional pressures on electrically conducting targets has a long history. It was first accomplished in the groundbreaking experiments performed by Cnare in the early 1960s. Using this technique in experiments at MSNW, peak magnetic fields as large as 200 T were achieved with excellent symmetry and quality employing only a 540 kJ capacitor bank. More critically, it has been performed at the scale required for the full scale F4P.

Slide 11

It can be seen that there is very good agreement with the modelling. In order to carry out these tests, the device was operated at one quarter power (170 kJ capacitor bank energy) to prevent post implosion damage. A vacuum environment was essential to avoid generating shock waves and drag as these liners propagated inward radially. Even at ¼ power, the peak magnetic field reached over 200 T with a magnetic pressure PB = 0.16 megabar (2.4 million psi). The Aluminum liners melted and vaporized. Remarkably, there was no visible damage or Aluminum deposit on the outer Garolite vacuum wall. The Aluminum vapor jetted out along the system axis and terminated on 1 inch thick 12-inch square plates of polycarbonate, which showed considerable visible damage. The radial implosion must be highly inelastic, and the evacuation of the Al vapor appeared to be guided by the axial magnetic field acting like a shaped charge. The high energy liners were rapidly heated and melted by Ohmic currents flowing in the liner. The time is limited here, but the key results from these experiments are summarized in the paper.

Slide 12

The AEDS was used to characterize the liner dynamics in 3D. These calculations were carried out with the same material, dimensions and magnetic pressure time history as the liners employed in the experiments. The near identical results validate the use of the AEDS code for accurately predicting liner dynamics in designing future liner implosion systems. A 3D solver is necessary to analyze and test desired liner properties such as liner stability, convergence, both radial and axial, as well as internal stresses and buckling. The buckling seen both experimentally and in the AEDS simulations are expected due to the extremely large ratio of liner diameter to thickness (4,000!). As this ratio increases the mode number of buckles increases. It is best not to leave the potential buckling to underlying imperfections or lack of azimuthal symmetry when it would be only slightly more effort to form the liners with the locations for folding preset. This “pleating” will assure a symmetric folding as the liners converge radially inward with the added bonus of minimal internal stress buildup.

Slide 13

Shown here is the ΔV budget and spacecraft initial and final mass parameters for a 210-day round trip manned Mars mission taking into account all phases of the mission. This breaks down to roughly 80 days to get to Mars, a 30 day stay and a 100 day return. The Copernicus software was employed to develop the mission architecture starting around a single 134 metric ton launch to LEO. Again, due to time constraints, one can find further mission details in the paper.  

Slide 14

The F4P propulsion system has many components and a certain degree of complexity. This is due primarily to the fact that it is a three-element hybrid of a solar electric powered propulsion, electromagnetic plasma propulsion system, and a nuclear thermal system for propellant heating. Just to add another layer of complexity, it is a pulsed hybrid system to boot. I thought it  would be useful to have a power/energy flow diagram that Indicates how the whole thing comes together.  This is it. As a point design, the 200 kW of solar power is stored in the pulse capacitors and discharged roughly every 12 seconds. Of the 5 megajoules of stored, half shows up as liner kinetic energy ang the remaining magnetic energy is returned to the capacitors as part of the oscillating driver circuit.  The collapsing liners compress the nuclear fuel releasing over 500 megajoules of energy into the lithium liner/propellant producing a hot plasma whose thermal energy is converted to directed energy by expansion into the thruster magnetic nozzle. Thermal radiation, ionization, and other frozen flow  losses reduce the directed energy to 441 megajoule which with a rep rate 0f 12 seconds  produce a jet power of 36 Megawatts.

Slide 15

This slide outlines the F4P spacecraft with a schematic on the left, and a CAD rendering of the full spacecraft on the right. A saddle truss was employed to facilitate refueling and reuse of the F4P spacecraft. It also provides for the option of a propellant or equipment drop for an even faster 90-day round trip Marian mission (30-day transit with a 30-day stay). This mission would likely require fuel pre-deployment at Mars. The component design was carried out mindful of the diameter limits of future planned payload fairings. More details of the spacecraft design can be found in the paper.

Slide 16

The F4P Rocket mass budget is shown here. It is comprised of the main liner compression coils as well as coils to create a divergent magnetic field at the exit. The mass of these coils is estimated at 0.4 metric tons. A thermal management system will most likely consist of a working fluid that will be pumped through the high heat flux components such as magnetic coils and power electronics. The fluid will be radiated through space radiator panels. The two major spacecraft components relate to the lithium propellant. Since lithium is a solid at room temperature, a low mass propellant storage system can be employed with a minimal tank mass fraction. The propellant feed mechanism on the other hand will require more mass. It is assumed that the liner material will be completely fabricated to the required specification on earth and then loaded on to a roll. These rolls would then be used to form the liner hoops. The liners are guided mechanically and injected electromagnetically at low velocity into position under the driver coils prior to driver coil activation. Another important aspect of the support structure is a shock absorbing system that interfaces the F4P with the spacecraft structure. The total mass for the spacecraft structure is therefore given the largest mass budget of any subcomponent. More details, again, can be found in the paper.

Slide 17

In summary then, Fission-Powered Pulsed Plasma Propulsion (F4P) has been elucidated, and the key subsystems described and evaluated for feasibility. If successful, the most significant application for the proposed F4P concept would be to make routine; exploratory deep-space missions with a rapid Mars mission as the logical first step. The mission architecture and spacecraft design for a single launch, a 210-day manned Mars mission has been analyzed for a fission-based spacecraft. Trip times as short as 30 to 90 days were envisioned with payload mass fractions of 36-65%. The experimental and theoretical results from research at MSNW and others have validated the metal hoop compression process needed to produce the desired nuclear energy gain. The major unknown to be resolved is the functional dependence of the fission energy yield on the several key parameters of the compression process. This must be fully understood and tested to prove the viability of the F4P concept.