Introduction

It may be appropriate for the inaugural edition of the Journal of Electric Propulsion to offer a perspective from the earlier days of electric propulsion (EP). These remarks are based on personal recollections from the 1960’s and will necessarily be incomplete, if not parochial, and do not represent a comprehensive review, as already provided in several excellent books on electric propulsion. Others should be encouraged to offer their own personal memories, which are rarely, if ever, recorded in technical papers.

My first work in electric propulsion began in 1960 when I was attracted to the notion of accelerating neutral matter using nonuniform electric fields. At about this time, there was still concern that ion engines would have difficulty neutralizing the ion beam in space. Nonuniform electric fields applied to polar molecules, or any polarizable material, including metals, offered a way around this potential difficulty, and might be employed in a multi-stage arrangement as in a linac. It turned out later, that such a notion works only for pulsed fields, and neutralization of ion beams was not really a problem. In any event, this started me in the direction of high voltages, electrical discharges and EP. My first paying job, at age 16, was in the Electric Propulsion Lab of Prof. Robert G. Jahn at Princeton University the summer before my freshman year (1963).

Early history

The earliest mention [1] of using electromagnetism for space exploration occurs when Cyrano de Bergerac proposes to reach the Moon by hurling a magnet into the air while standing on an iron plate, repeating this (as in a pulsed thruster) to continue upward. A proper and comprehensive history of electric propulsion is provided by Ernst Stuhlinger in the first chapter of his superb book Ion Propulsion for Space Flight [2]. He covers the early considerations of rocket pioneers Robert H. Goddard and Hermann Oberth, especially in regard to electrostatic acceleration of ions and sprays. In Electrical Rocket Engines of Space Vehicles [3], S.D. Grishin and L.V. Leskov note that K.E. Tsiolkovskiy “in 1911 for the first time expressed the thought that with the help of electricity it is possible to give immense speed to particles ejected from a jet instrument”. They also provide a brief account of early electric propulsion activities in the Soviet Union. During the Cold War, there was only limited understanding of such activities, typically transmitted informally by V.V. Zhurin. (This changed dramatically in the 1990’s with the spectacular reports [4, 5] on the Stationary Plasma Thruster and other Hall thrusters.) Activities in China [6] were similarly almost unknown in the West, but included work on ablation-fed, pulsed plasma thrusters by An Shi-ming [7]. EP had some popular exposure in a Disney movie Mars and Beyond (1957) that included a fleet of six “cocktail umbrella” spacecraft spiraling into orbit around Mars. The umbrella was the radiator, the crew quarters were around the tip, and an array of ion engines was stationed on the shaft, which ended in a nuclear fission reactor. Ernst Stuhlinger was shown describing the system to Wernher von Braun (no audio, but presumably in German). I had retrieved this film for the AIAA Electric Propulsion Conference in 1985, and Edgar Choueiri did likewise for his excellent International Electric Propulsion Conference (IEPC) in 2005 in Princeton. This vision of large spacecraft for crewed voyages to Mars inspired and shaped much of electric propulsion thinking for subsequent decades.

Fig. 1 provides a taxonomy of electric propulsion techniques [8] c. 1995, most of which are discussed in terms of their operating principles in Physics of Electric Propulsion [9] by R. G. Jahn, ranging from colloid thrusters to ion engines and arcjets to traveling-wave notions. Some concepts did not last, while others disappeared, but have returned later, as more general appreciation of EP has occurred.

Fig. 1
figure 1

The taxonomy of electric propulsion (c. 1995) indicating the three basic categories of EP and some of the several individual techniques [8]

Full size image

A view in the 1960’s

The sixties were heady times, when NASA was funding many schemes that might advance opportunities for space exploration. Prof. Martin Summerfield remarked to me many years later that back then “any idea could get money, later any good idea would be funded, and now (c. 1985) even good ideas cannot find funding.” Indeed, anything that could make a spark might be proposed as the basis for an electric thruster. After all, that spark to a doorknob after walking across a wool rug in winter has temperatures much higher than in any chemical rocket engine. Conversion of high temperature, ionized gas into useful thrust might therefore be worth exploring, and continues to represent a challenge to more than one category of electric propulsion. (As a youngster, Robert H. Goddard wondered if such household electrical activity could help him rise above gravity, and he later returned to electric propulsion briefly as a research associate in physics at Princeton.)

The taxonomy of electric propulsion of Fig. 1 has three main categories: electrothermal, electromagnetic and electrostatic, among which there are (as Bob Jahn would phrase it) a “myriad” of variations. There was also a belief that the three categories would apply to three regimes of space flight mission. Electrothermal techniques at specific impulse values of several hundred seconds would be employed in earth orbit; electromagnetic thrusters, up to about 10,000 s, would take us to Mars; and electrostatic engines would be for exploration of the outer planets and beyond. No one thought that we would someday photograph Pluto with a spacecraft launched from the earth by chemical propulsion. This has been a victory for silicon, with vastly improved computers and light-weight electronics enabling compact diagnostics, communication links, signal processing, and gravity maneuvers. Indeed, this same progress with electronics in some sense impeded the entry of electric thrusters for primary propulsion roles because the previously expected need for megawatts of electrical power in space never developed.

Another curious impediment to the introduction of electric propulsion for space missions involved interactions with project managers responsible for selecting spacecraft thrusters. This included the very diversity of choices indicated in Fig. 1, with many different types of thruster clamoring for attention, but also offering some confusion. The other problem was that no manager wanted to be the first to fail in accomplishing the mission because of choosing an EP system. If a “conventional” thruster fails, it is just the luck of dealing with complex devices in space. If an electric thruster is selected and fails, however, then it might be taken as a “lack of sound engineering judgement”, ominous words not to be invited into one’s career.

So, while many research projects continued to explore electric propulsion concepts, the goal of getting EP applied in space languished. There were tests of an ion engine in a space environment (as SERT I [10] and II [11], by NASA Lewis Research Center), and the launch of an ablation-fed pulsed plasma thruster on a sub-orbital rocket by the Chinese [7]. A similar ablation-fed PPT [12] accomplished mission use with LES-6 (by Lincoln Labs). A more accepted application came with resistojets [13], including the Augmented Hydrazine Thruster (AHT), which added modestly to the Isp for station-keeping by resistively supplying heat to hydrazine thrusters. For some reason, these were not really celebrated as electric propulsion successes, at least in academia, perhaps because they lacked the delightfully intricate and clever physics of electromagnetism and flow. Such interesting EP concepts thus remained in the laboratory, and in a quandary regarding acceptance for space applications: EP could not be used until it was used.

Meanwhile, back in the lab, progress was made on several fronts. The notion of employing contact ionization (e.g., cesium on hot tungsten, porous plugs) as the source of ions for electrostatic propulsion, which had been promoted by E. Stuhlinger [2] at NASA Marshall Space Flight Center, gave way to electron-bombardment, discharge ionization [14] by H. Kaufman at NASA LeRC using mercury propellant, and continued to RF discharge techniques [15] at the University of Giessen (and, much later, microwave-driven sources [16] in Japan).

For plasma thrusters, early work on arcjets, which found application as high enthalpy sources for industrial needs and re-entry vehicle material testing, included successful development of a 30 kW radiation-cooled constricted arcjet that provided Isp = 1500 s with hydrogen and 1000 s with ammonia [17]. Work on arcjets extended into electromagnetic regimes, and initially involved inclusion of an axial magnetic field to spread the arc discharge heating more uniformly around the electrodes. But it was recognized also as a way to increase the impedance of the discharge. Such an impedance increase is associated with the rotation of the plasma in the axial magnetic field, and led to notions of converting the rotational motion into useful thrust by means of the so-called magnetic nozzle provided by the expanding axial field [18]. Thus, sub-categories of thrusters developed as self-field vs applied field arcjets. The latter were studied at Giannini Plasmadyne by A. Ducati, at Avco-Everett by R.R. John’s group, and at many other places around the world, including later at NASA LeRC by M. Mantenieks, R.M. Myers and J. Sovey. The self-field approach eventually became the focus of work at Princeton.

The initial activity at the Electric Propulsion Lab at Princeton University centered on the use of a large-radius, high speed (~ 40 km/s) implosion of a dynamic plasma discharge to create a high temperature pinch discharge on the axis [19] (Fig. 2). Plasma swept up by the imploding discharge would be expelled from the pinch column at high speed to provide the thruster exhaust. A related scheme [20] at Republic Aviation allowed the imploding discharge to be re-directed into an axial direction, rather than depending on the expulsion of plasma by the pinch; a similar approach was followed by A.I. Morosov, et al., in the Soviet Union. This involved adjusting the rz-shape of one electrode, allowing the plasma to exit out a central hole in the other electrode. The early Princeton experiments merely used two flat, circular electrodes. (Bob Jahn was a shock physicist, rather than a space cadet, and learned about large-radius, dynamic discharges while a faculty member at Caltech, sharing an office with George Vlases who was using an inverse-pinch discharge to create strong shocks.) In one set-up [21] at Princeton, operated by Rod Burton as a graduate student, a large hole was provided in one electrode allowing the plasma to escape into a bell jar (Fig. 3). In addition to the pinched plasma, the current pattern expanded out the hole forming a trumpet-shaped plume that persisted thanks to a more elongated (vs purely sinusoidal) current pulse.

Fig. 2
figure 2

Sequence of Kerr-cell photos of a large radius dynamic pinch discharge. The initial diameter is 20 cm and height 5 cm [19]. In the first three frames, a bar blocks the intense pinch light from bleeding through the Kerr cell

Full size image
Fig. 3
figure 3

Sequence of Kerr-cell photos of expulsion of pinched plasma from chamber (diameter 12.7 cm, length 5 cm). Note resemblance of plasma luminosity at 2.3 and 2.7 μsec to arcjet plume [21]

Full size image

The resemblance of this pattern to an arcjet flow led to a shift of the Lab’s effort toward studying high power, self-field magnetoplasmadynamic (MPD) thrusters by using a pulse-line to provide constant current to the discharge for hundreds of microseconds. This shift was encouraged by the discovery of an apparently higher efficiency regime of applied-field arcjet operation [22] by Adriano Ducati at Plasmadyne, when he substantially lowered the propellant input rate. (The thruster even operated when the mass flow was reduced to near zero, leading to concerns in future arcjet work with tank effects, e.g., refluxing mass.) Prof. Jahn returned from his annual summer consulting work at Plasmadyne to launch graduate student Kenn Clark on creating the first quasi-steady MPD thruster [23, 24] as in Fig. 4. This approach both enabled a university research effort at multi-megawatt power levels, but also offered the possibility for space applications to access higher efficiencies in repetitive, high power pulses, while operating at much lower average powers.

Fig. 4
figure 4

Kerr cell photo of cathode region of a quasi-steady 2.5 MW MPD arcjet, 100 μsec into a 17.5 kA, 200 μsec constant current pulse [24] The propellant is argon, but there is ablation of the Plexiglas insulator indicated by greenish light near the base of the cathode (1.9 cm diameter). Less than 10 % of the current actually flows within the bright central region, with the rest more uniformly distributed in the thrust chamber

Full size image

It should be noted that the Sixties was a time of amazing growth in computational capabilities, from a period in which some still believed that analog techniques could compete with digital computers to the appearance of large mainframe, digital computers on campus; (Princeton’s had 36 K of RAM!) For EP, such capabilities, however, mainly involved the development of zero-dimensional codes dealing with the nonlinearities of circuit behavior. Detailed modeling of complex MHD flows to describe the operations of MPD thrusters was deemed too difficult and was not within the research program at Princeton at that time. (Much later, such modeling was performed at R&D Associates, Inc. and The Ohio State University, adapting codes [25] borrowed from the Air Force.)

While the quasi-steady, self-field MPD thruster was great for training graduate students, it had little direct impact on the world of MPD arcjets, which continued in research programs at other laboratories. It was still difficult to find applications of electric thrusters for primary propulsion in the absence of megawatt space power supplies. An exceptional program was that at Los Alamos Scientific Laboratory where both nuclear thermal propulsion [26] and nuclear-powered, electric propulsion were pursued; the latter under Tom Stratton involved an applied-field MPD thruster using lithium as propellant [18]. It was envisioned that a nuclear reactor at 10 MWe would power such an arcjet to Mars as an alternative to nuclear thermal propulsion. The Apollo program, by its success in reaching the Moon, derailed this vision.

1970’s: the dark decade

After satisfying the US national goal of delivering astronauts to the Moon and returning them safely to earth, the impetus to continue further into the Universe with large missions diminished substantially and quickly. As the search for post-Apollo jobs became a concern, many folks listed helping reach the Moon on their resumes, but also deleted inclusion of their doctoral degrees, (lest it appear they would leave, if research funding returned). Aerospace engineers found new employment with high-energy lasers and high current power-line interrupters. Experience in electric propulsion led to applications for other needs for pulsed electrical power. For example in 1970, as a young Air Force officer, I introduced the use of large radius, imploding discharges of high-Z plasma to create very high power, short pulse sources of soft X-radiation [27] to simulate the effects of nuclear weapons. (This work at the Air Force Weapons Lab led eventually to the present Z-machine at Sandia National Labs.) The application of electric propulsion to a non-EP need was repaid many years later as the sophisticated numerical techniques from the nuclear weapons community were applied to electric thrusters, successfully modeling both self- and applied-field MPD thrusters [28, 29] and the ablation-fed PPT [30].

As with the cultural phenomena of the Sixties that actually extended into the seventies, the Dark Decade of the seventies continued into the eighties. Many programs disappeared. These included the applied-field arcjet work at Avco-Everett, and even the program on nuclear energy for rocket vehicle applications (NERVA) at Los Alamos, in which the nuclear-electric driven, applied-field MPD thruster [18] went down with the nuclear thermal propulsion effort. But, as with the Dark Ages, things were not completely dark. There were many plans for electric propulsion in both the US and Japan. These included applications for US Air Force missions [31], (much later expanded to several projects [32]), the NASA solar-electric propulsion system (SEPS) [33], and a space test [34] of an MPD thruster by Japan, under K. Kuriki. The Strategic Defense Initiative (aka, “Star Wars”) offered some impetus for advanced spacecraft and included the development and (eventual) space test of a 26 kW ammonia arcjet.

Perhaps the nadir of EP was epitomized by relegation of the electric propulsion sessions at the Joint Propulsion Conference (Cincinnati, 1984) to the basement of a fish restaurant across the parking lot from the main meeting. Still, EP continued to make some progress, even if efforts were supported merely to answer concerns of EP’s possible interference with communication or contamination of the spacecraft with thruster exhaust. (The latter was particularly odd because in comparison to chemical propulsion EP ejected much less mass at much higher directed speeds.) Note that the Dark Decade for EP described here was largely a US phenomenon, while substantial progress continued in the USSR, which would later greatly impact efforts in the West.

A false dawn and a new hope

In 1991, the Space Exploration Initiative (SEI) was announced in the US. This would include a return to the Moon, but was largely focused on a crewed mission to Mars. The baseline for the Mars mission was, of course, chemical propulsion. This committed the crew to a year-long journey to Mars, a very exciting entry and landing at Mars, and an even longer trip home. (One item that did not receive much attention was that no one knew the psychological effects of separating humans from the earth by vast distances, which was different than being able to look out the window at home from near-Earth orbit or even the Moon.) Clearly, the earlier dreams of nuclear thermal propulsion (NTP) and nuclear-electric propulsion (NEP) could do better by reducing trip times and costs. While SEI brought back many of the surviving pioneers of nuclear propulsion, environmental concerns had changed over twenty years, so the challenges of ground-testing NTP were deemed too expensive. With NTP set aside, NEP was for some odd reason taken down also, including laboratory work at JPL for which the name MPD thruster that had been associated with nuclear power for SEI had to be changed to Lorentz Force Accelerator (LFA) in order to survive.

While the “false dawn” of SEI had a significant effect on MPD thruster projects for primary propulsion, progress continued in this same time period on the use of EP for station-keeping in near-Earth orbit. The xenon ion propulsion system (XIPS) [35] developed by Hughes Research Lab found acceptance for use on communication satellites, thanks in part to close corporate connections. This led to recognition by others that EP could provide competitive commercial advantages. Such recognition helped to encourage the successful development and sale of kilowatt-class, hydrazine-fueled (thermal) arcjets [36] by Rocket Research (and its descendants).

Note that there are two main approaches for designing an electric propulsion system [8]. One starts with the mission, selects a thruster, and then searches for (or develops) a power supply; the alternative is to accept the power available on the mission, and then select the thruster. Earlier focus on EP for primary propulsion demanded high power in space (0.1–10 MWe), which was not available. But satellites, e.g., for communication, provide their own power sources to which smaller electric thrusters could be matched for station-keeping missions. (Later, to demonstrate technology in support of potential missions, solar-cell arrays were combined with Hall thrusters and ion engines in the design of consistent solar-electric propulsion systems, such as SMART-1 by ESA [37], and still later, the Solar Electric Propulsion System at NASA Glenn Research Center, c. 2015.)

Field of the future

For decades, electric propulsion had been known as “the field of the future, and always will be”. With the advent of station-keeping roles, the future arrived in the nineties [38]. This was helped immensely by the news and establishment of multiple connections with the EP efforts in the former Soviet Union. These efforts included very high power, lithium arcjets [5], and decades of development and flight operation of Hall thrusters. In particular, the performance of the Stationary Plasma Thruster [39] was quite encouraging. Hall thrusters (in other arrangements) had been touched on in the US [40], but not pursued as arcjets, MPD devices and even gridded ion thrusters seemed simpler embodiments. By way of comparison, one aspect offered an attractive feature of SPT vs MPD behavior. In a self-field MPD thruster, the back EMF voltage across the accelerated flow is parallel to that at the entry to the thrust chamber, where it is supported by the resistive voltage drop as propellant enters the chamber at low speed. This means that the thrust power is always proportional to the resistive power deposited in the propellant, so reducing the mass flow rate to obtain higher exhaust speed results in heating in excess of what can be absorbed into the ionization of the propellant, limiting the exhaust speed to values scaled by the Alfven critical speed [41, 42]. Attempts at higher speed lead to erosion and instabilities in the chamber. In the SPT, the voltage is developed in the axial direction by the Hall currents carried by circulating electrons, so high exhaust speeds, beyond ionization-based limits, can be obtained even with high atomic mass propellants, such as xenon; (other sources of erosion, of course, can exist.) Perhaps it was the ability of an electromagnetic thruster to obtain speeds much greater than Alfven critical speed (e.g., with xenon, krypton and argon) that led to some early confusion in which the SPT was identified in the West as a “closed-drift” ion thruster. (At the 1991 Electric Propulsion Conference, when we met in separate discussion sessions for ion engines and electromagnetic thrusters, I was able to verify with V. Kim that SPT indeed should be with the latter.) The ability to obtain high accelerating voltages without the need for the complex electrodes of electrostatic devices was another very attractive feature. Excellent interactions quickly occurred between Russian and Western groups, leading to rapid development of Hall thrusters for space applications, including joint ventures (e.g., STEX, the first western Hall Thruster flight in 1998, using a Tsniimash-style thruster via NASA LeRC to the Naval Research Lab, with a power conditioning unit from Rocket Research/Primex, led by Bill Smith), and commercialization (e.g., by Space Systems/Loral and Fakel Enterprises). Later efforts in the US involved new designs (e.g., by Busek Co., University of Michigan, and others).

Meanwhile, improvements in ion engine technology and demonstrations of satisfactory lifetimes in ground tests at JPL and NASA Glenn Research Center provided the basis for consideration of EP for primary propulsion. Although EP eventually missed an opportunity for the US mission to Mercury with Messenger, it did operate for voyages to the asteroids, including the NSTAR ion engine [43] for Deep Space-1, and Dawn [44]. Later, there were the fabulous Japanese missions of Hayabusa for sample returns [45], using engines with microwave generation of ions, and EP was selected by ESA/JAXA for the BepiColombo mission to Mercury. With confidence, therefore, the EP community was able to claim that “the future is now”. There was, however, still some lingering disdain by others.

For decades, the AIAA had sponsored a conference on electric propulsion that was held every year-and-a-half. Initially, these meetings were only in the US, but successful conferences were eventually held in Japan, Germany and Italy, co-sponsored by the local aerospace organizations. After the excellent 1991 meeting in Italy, the AIAA technical committee for EP selected Seattle as the next site, in order to capitalize on the proximity of Rocket Research for a visit to their electric propulsion activities. Unfortunately, AIAA Headquarters staff objected that the EP conference was too small to sponsor as a stand-alone meeting: “I’ve had larger cocktail parties in my home”; (perhaps her home was the size of a Cincinnati fish restaurant.) The EP community decided to proceed with the Seattle site, under the auspices of the Electric Rocket Propulsion Society, which was created for this purpose. We thereby had the first of over 25 years of the International Electric Propulsion Conference, which has served our community world-wide, as interest and activity in electric propulsion have grown dramatically. As chair of the AIAA Electric Propulsion Technical Committee and the first president of the new society, I helped to midwife this transition. More involved with the efforts needed for success, especially for the conference, were my co-founders: Joe Cassady, Frank Curran, Roger Myers and Bill Smith.

The S-curve

We might summarize the development of electric propulsion over the past 60 years by the so-called S-curve [46, 47] in Fig. 5. A typical curve of technology starts with a period in which basic notions are explored, gathering experience on what works and what doesn’t, and understanding the needs of future applications. A few concepts emerge that are deemed worthy of further development, and much progress is made toward practical and attractive systems. The curve then approaches a situation in which only incremental improvements are made on the basic concepts, representing refinements for particular applications. For EP, we started with the “myriad” electrical ways of accelerating matter to very high speeds. Some of these proved impractical, while others promised performance that did not match the needs of the market at the time. A few were able to go the distance toward applications, but encountered the Dark Decade, during which progress continued, but without substantial demand. With the growth of space assets, such as communication satellites, and recognition of the commercial advantages that high specific impulse could provide, (both as more time on station and reduced launch vehicle costs), electric propulsion has come into its own [48]. The upsurge in the EP curve starts c. 1991 with both this commercial recognition and with the revelation in the West of the Russian work on Hall thrusters. In terms of the two approaches to EP system design [8], satellites with significant electrical power offered the opportunity for EP to demonstrate its in-space capabilities. Success encouraged considerations of primary propulsion in the alternative design approach by which an EP technique is selected and space power is then designed to match. (Obviously, the overall spacecraft mission requirements contribute to the interplay between thruster and power supply selection and design.)

Fig. 5
figure 5

The S-curve for electric propulsion starts with exploration of a “myriad” of thruster concepts, survives the Dark Decade of the Seventies, and then approaches a mature plateau of practical applications [46, 47]. The sharp upswing c. 1991 corresponds to the recognition of the commercial advantage of EP for satellite station-keeping, and the news in the West of Hall thrusters from the former Soviet Union

Full size image

A few techniques, such as Hall thrusters and ion engines, have reached a mature state of development that can include primary propulsion roles, in addition to station-keeping larger spacecraft. Currently, as part of the Artemis project returning humans to the Moon, Hall thrusters are scheduled to lift the Habitation and Logistics Outpost/Power and Propulsion Element (HALO/PPE) from earth orbit to its station orbiting the Moon. Other techniques, from small arcjets to ablation-fed PPTs to colloid thrusters find applications on satellites of various sorts that make up a growing population surrounding the earth. The possible return of nuclear energy for rocket vehicle applications (as in the recent DRACO project [49] for nuclear thermal propulsion) may also lead to space-nuclear power for electric thrusters, ranging from arrays of ion or Hall thrusters to lithium-fueled arcjets (or even to a high power form of electrothermal thruster VASIMR [50], somehow claimed not to be electric propulsion, but “plasma propulsion”). Ground-test of very high power systems for primary propulsion continues to represent a major challenge because of the propellant throughput, both in terms of the necessary vacuum systems and also the total amount of propellant (e.g., xenon) needed for adequate life-tests. Operation of lower power modules is an accessible approach for ion and Hall thrusters, presuming acceptable statistical treatment for large arrays. Ground test of a high power, lithium arcjet is facilitated by the ability to condense the lithium exhaust as in the Los Alamos work [18].

The progress EP has made up its S-curve is a tribute to the dedication of many people over many decades. Early on there was the leadership of Ernst Stuhlinger, Harold Kaufman, Yvonne Brill, Bob Vondra and Bob Beattie. An absolute hero in the US was the late Dave Byers, who tirelessly worked both the technical and programmatic issues, particularly emphasizing the requirements of ground-testing electric thrusters for their long-duration missions. He mentored many other major contributors, including Frank Curran, John Brophy, Jack Stocky and Jay Polk at NASA, and Roger Myers and Bill Smith at Rocket Research (eventually Aerojet). With the emergence of the SPT, Len Caveny, then at the Strategic Defense Initiative Office, helped to start Alec Gallimore on his splendid career with Hall thrusters at the University of Michigan. Progress elsewhere included the incomparable efforts of Kyoichi Kuriki in Japan, Marianno Andrenucci and Giorgio Saccoccia in Italy, and Horst Loeb and Monika Auweter-Kurtz in Germany. Our Russian colleagues who successfully labored, long out of our view, included A. Bober, V. Kim, A.I. Morosov, G. Popov and V. Tikhonov. With the growth of the field, there are now literally hundreds of contributors, perhaps matching the thousands of satellites now in orbit.

A new future?

In rising toward the S-curve plateau of practical application, there will still be both continued improvements and new notions. These include better components, e.g., lower erosion electrode techniques, and the revival of older concepts with modern technologies, e.g., colloid thrusters. The latter benefit from the overall success of electric propulsion in general, but need to address the same standards of life-time and reliability that have allowed acceptance of mature techniques. When an S-curve is accomplished, it is time to move to a new S-curve. What might this be?

Looking ahead 60 years, an area of continued challenge is crewed exploration of the Solar System. This will require high specific impulse and high specific power. The former is always available, even extending to the speed of light in photon propulsion mentioned by Stuhlinger [2]. The latter, however, is much more difficult and will be limited by the specific power of radiators, which means we must create concepts that minimize waste heat. Furthermore, a crewed voyage might involve a spacecraft mass of upward of thousands of tonnes, including radiation shielding, (recalling the very large spacecraft of the Disney movie mentioned earlier), and needs to be accomplished with transit times of only a few months. These requirements translate into thruster powers of many megawatts. To operate far from the Sun, e.g., exploring the oceans of Enceladus, nuclear power is needed. This is akin to the NEP effort in NERVA at Los Alamos that had the goal of 10 MWe for its lithium arcjet, but ended in the early seventies before achieving that goal; (at Saturn, such power would require a solar array about 1.6 km on a side.)

Nearing the top of the present S-curve, there is continued interest in nuclear-electric propulsion, which would seem to be a straightforward connection of mature EP techniques to a nuclear fission source of adequately high specific power. The latter may, however, still be too difficult, which would leave us resigned to robotic vs crewed missions to the worlds of the outer planets because trip times would be too long for safe travel by humans. (We need specific powers in the limit of the radiators and very large total powers.) One potential approach recognizes that conventional rockets benefit from an “open” thermodynamic cycle vs the closed cycle of the power supply for “conventional” nuclear-electric propulsion. Is it possible to go directly from nuclear energy to high specific impulse flows far above values available with present concepts for nuclear thermal rockets?

For many years, there have been notions of using the high speed products of nuclear reactions, e.g., fission fragments, and alpha or beta-decays, directly for thrust. Unfortunately, the short ranges of such products in solid-densities, while fine for generating heat, otherwise demand very large areas of solid surfaces to make these particles accessible for providing thrust. Gas-core fission rocket concepts help, but are limited by competition among particle interactions, e.g., between fission reactions and Coulomb collisions, if we attempt to use magnetic fields to limit heat transfer from the necessarily very high temperature plasma for high specific impulse.

Nuclear fusion offers possibilities, but has proven very challenging over many decades seeking to obtain terrestrial power. Much of the problem with such a goal is the difficulty of competing with other, cheaper sources of power on earth (e.g., natural gas, or even solar or wind power). For space applications, particularly missions to the outer planets and beyond, such competition is not a concern. There remain, however, the detailed technical challenges of controlled fusion, especially dealing with the high flux of high-energy neutrons. Apart from insulator damage and induced radioactivity, that portion of fusion energy that is provided by neutrons represents a reduction in the overall system specific power, either because of the radiator mass needed to deal with the heat of neutron deposition, or the occasionally suggested expedient of simply allowing the neutrons to escape. For the former, we return to the situation of a closed-cycle power plant. With the latter, we lose much of the energy of neutron-rich fusion reactions, such as from D-T or catalyzed D-D, making a challenging task even more difficult.

Fusion based on so-called “advanced fuels”, such as D-He3, helps by a factor of about twenty-five, and is attractive for space vs terrestrial application because of the availability (it is said) of He3 on the Moon thanks to the solar wind. There remains the difficulty of obtaining optimum values of exhaust speed, which are typically much lower than the speeds of fusion reaction products. This means that not only must we achieve fusion with much smaller fusion reaction cross-sections (and/or higher temperatures) than for D-T, but the energy obtained needs to be shared efficiently with a larger mass flow to allow use with the open-cycle of a rocket. Neither of these concerns needs to be addressed by the present terrestrial fusion efforts based on D-T. By adding in the requirements for system and mission considerations to a problem involving electromagnetism and plasma flow, all of which are traditional EP concerns, we may define an appropriate challenge for electric propulsion folk. Here, the energy of the flow is provided initially at the highest temperature by nuclear processes, and electromagnetic forces are needed to insulate and channel this flow to create a directed exhaust at optimum specific impulse. The new S-curve involves new technology, but applies the same consideration of mission needs, component life times, and system trade-offs in a complex electromagnetic and plasma environment with which the electric propulsion community is very familiar.

Such an adventure will not be to everyone’s taste or abilities. It is perfectly reasonable to follow the present S-curve toward further improvements and applications of known electric propulsion techniques. There may even be new conceptual arrangements. For some, however, the lure of advancing technology to enable crewed space-faring around the Solar System will be compelling. We should not view their explorations at the beginning of the new S-curve with the same level of rigor as applied to the mature technologies of our present S-curve, but recognize how far we have come and how far they must go.

Coda

A memoir that extends back over 60 years can encourage recollections from others, especially beyond the largely US efforts discussed here. The watershed in the West of the introduction of Hall thrusters from the former Soviet Union needs proper exposition while principal players are still available. Equally, it would be useful to consider the evolution of efforts in Japan, Europe, and China. Other memoirs can focus on the last 30 years during which there has been so much progress in electric propulsion applications. Indeed, recent work [51] extends the introduction of S-curve considerations of EP [46, 47] quantitatively to a broad range of thrusters, (and provides a very substantial, albeit incomplete, list of references). Also, the classic texts of Stuhlinger [2] and Jahn [9] have been properly complemented [52] for Hall thruster and modern ion engine design. The applications of EP to satellite control now number in the thousands (e.g., Starlink), and we may soon continue the 1960’s goal of primary propulsion toward the outer planets with the Psyche mission [53] using Hall thrusters, enabled by continued progress with solar-electric systems. Beyond this, crewed exploration of the diverse worlds of our nearest “solar systems”, Jupiter and Saturn, still beckons.