share interactive transcript request transcript/captions live captions
download
|
MyPlaylist
SPEAKER: Hello, everyone. Good evening, and on behalf of the Cornell Physics Department, it is my pleasure to introduce you to this year's Bethe lecture. And I have two privileges tonight.
The first is I get to do-- I get to introduce our speaker, Professor John Foster, but I also get to tell you a little bit about the life and achievements of Hans Bethe. So Hans had several careers. Every one of his career was outstanding.
It's safe to say that no one will ever match the excellence he sustained over almost 80 years as an active scientist. So the time that we have this evening does not permit a full accounting of the breadth of Hans' achievements, but a short summary-- and this short summary was graciously provided to me by Ira Wasserman-- will give you the idea.
Hans began his career in the 1920s when he was among the first young physicists to explore applications of quantum theory. Upon leaving Germany to escape persecution for his Jewish heritage, Hans moved first to Great Britain, where he began his lifelong work in nuclear physics. And almost immediately, he became the world expert in the field.
Hans moved to Cornell in 1935. About four years later, he explained how stars burn hydrogen to helium. And for this, he received the Nobel Prize in physics for 1960-- in 1967. At the core of Hans' career was a fundamental question-- how does matter behave at extremely high density he was a theoretical physicist first and foremost, but fascination with this particular question brought him repeatedly to the realm of astrophysics.
Hans retired from Cornell in the mid 1970s after 40 years on the faculty. He helped build the physics department to the world class status it continues to enjoy to this day. And he fostered an informal collegial environment that made one of the defining characteristics of our department.
After retirement, there's still much more to say. Hans continued to do research almost actively almost right up to his death in 2005. He didn't merely putter around in this time. He led the worldwide effort of astrophysicists to understand supernova explosions.
And almost, as a lark, in 1986, he solved the stellar neutrino problem that had bedeviled stellar astronomers for 20 years. At that time, Hans was 80 years old. How did he sustain such a high level of achievement for so long?
Well, he loved solving problems. He was brilliant, of course, but physics was fun for him. For his entire life, he was as enthusiastic and forward looking as a new graduate student.
Throughout his career, Hans exemplified personal integrity and courage. For example, during the anti-communist hysteria of the McCarthy era, he was an early opponent of the development of the h-bomb. He helped protect Cornell physics colleague Philip Morrison from being dismissed as a result of his vocal opposition to the Korean War and his purported communist sympathies.
Famously, he defended J. Robert Oppenheimer, the former head of Los Alamos in his notorious security clearance hearing. He was a forceful and effective advocate for the limited test ban treaty, which forbade testing nuclear weapons underwater, in the atmosphere, and in space. He was a formidable opponent of the Star Wars program, and he was a relentless proponent of peaceful applications of nuclear energy.
Hans may yet become the prophet of energy, as Jeremy Bernstein dubbed him. Hans Bethe was a great man, and we are lucky that he gave so much to Cornell and left us with such a wonderful legacy. So now, I get to my second introduction. I get to introduce our speaker tonight, Professor John Foster.
Professor Foster received his BS in physics from Jackson State University in 1991 and his PhD in Applied Physics from the University of Michigan Ann Arbor. Throughout his academic and professional career, he's been involved in electric propulsion research, stunning thrusters from each of the categories of plasma-based propulsion-- electrothermal, electrostatic, and electromagnetic. And some of these actually appeared in-- if you're a movie buff-- 2001-- A Space Odyssey.
After graduation from Michigan, he served as a plasma diagnostics postdoc at the University of Wisconsin's center for plasma aided manufacturing, which actually deals and is related intimately with how the semiconductors in your cell phones work. He then later joined the onboard propulsion group at NASA Glenn to study advanced plasma-based propulsion. And his work focused on the design of ion thrusters, including the next ion thruster which flew on the DART mission, which I learned was a mission designed to test how well spacecraft could deflect life threatening asteroids away from Earth.
He left NASA in 2006 to join the nuclear engineering department at Michigan. His research at Michigan focuses on space propulsion, including the development of novel ion thrusters and nuclear thermal rocket engines, which we will hear about tonight. But also a very distinct branch of research, which we heard about in our Monday lecture, John studies environmental plasma processing, and in particular, Monday's riveting lecture, which you can find online, was about purifying water with plasma.
The environmental plasma work that John does includes terrestrial plasma-based water purification, as I just said, and plasma driven depolymerization to address the problem of plastic waste. So with that, it is my absolute pleasure to give you Professor John Foster.
[APPLAUSE]
JOHN FOSTER: Thank you. So in this lecture, I'm going to focus on nuclear electric propulsion in contrast to nuclear thermal rockets, which we talked about on Tuesday. So the goal is to give you an overview of nuclear electric propulsion and then look at some of the issues in regards to testing these systems.
So the talk will follow the flow of this outline. We'll first introduce what electric propulsion is, and then go into some opportunities afforded by nuclear electric propulsion. You'll see NEP throughout the talk as an abbreviation.
Then, we'll look at applications of it. For mission to Mars, it's competing with nuclear thermal rocket. And then the latter half is the-- we're looking to a little bit of a deep dive into how you test these engines on Earth, because they're going to operate at nearly an order of magnitude power levels than what they currently validated-- their state-- the state of the art is a tenth lower than what we need them to be. And then I'll do some concluding remarks.
So to kick it all off-- so the people from last time, you've probably seen this rocket equation. But to put the electric propulsion in context, it's worth taking a look at the rocket equation once more. So the rocket equation relates the change in velocity to the propellant required to do mission.
It's derived from the conservation of momentum. So most missions-- well, all missions are defined by delta v-- that is the change in velocity of rocket. We're in this gravity well in the solar system, and to move from one orbit to the next, you have to impart energy to your vehicle.
We go to outer planets-- the energy of those orbits are higher, so you have to increase your total energy to get onto those orbits, so you have to use delta v. And there's a cost to using the delta v, and that is the propellant expended. And so this is the relation.
We can rearrange it and see that the mass ratio-- the mass of the rocket before the fuel burned over the mass after fuel burn grows exponentially with delta v requirement. And in the denominator is the exhaust velocity. So for those missions that have large delta v's relative to the exhaust velocity, this mass ratio is huge.
And in general, the difference between the mass before and the mass after is the fuel that's expended, so it is also growing exponentially. So how do we define efficiency of a rocket? Well, before doing that, the thrust is just m dot times the mass flow rate times exhaust velocity.
If we divide the thrust by the mass-- the fuel consumption rate m dot, we get the parameter known as a specific impulse. And that's what, in rocket science, people refer to to determine-- to basically express the efficiency of a rocket. They call it the specific impulse, or ISP, and its units are in seconds.
And so it is the ratio of the thrust to mass flow times g. That's what makes it specific, because it's relative to Earth's gravity. And so simple math-- you see that this is just the ratio-- you do the math-- it's just the exhaust velocity over g.
And so the faster your exhaust velocity, the more efficient your vehicle is at giving rise to a certain amount of thrust, and thus giving rise to the delta v that you request. And so specific impulse is very important. And we'll put that in context when we talk about the electric propulsion systems, because we want to compare that to rockets. And I may allude to thermal-- nuclear thermal rockets as well.
And so I also wanted to just introduce you to how do we get around in the solar system? We're interested in Mars. NASA is keen on going to Mars in the next decade. And so the way we travel from Earth to Mars right now, because of the limitations of our technology, is the Hohmann orbit transfer.
And Hohmann transfer basically is illustrated here. So the way it works is we basically transfer to a transfer ellipse. Somehow, my arrow disappeared on the screen. But so here's-- so we're orbiting around Earth.
The way we get to Mars is we have to enter into a transfer ellipse that connects Earth to Mars. And so this transfer ellipse is a legitimate orbit that puts you in orbit around the sun. But the nice thing about it is that it accepts us as well as Mars.
So we do a fuel burn to get on this transfer ellipse-- just one burn and-- to get on-- to get to the total energy of this orbit, and we coast to Mars. So it's a boost and coast. When we get into the vicinity of Mars, we fire retros to circularize and enter into orbit around Mars.
And so this particular way, it's sort of the slow boat to Mars because you're boosting and you're coasting for a long distance. Kepler's law will tell you how long it takes because it's half the period of this elliptical orbit. A better way would be to thrust along the orbit, and that puts you on a transfer ellipse that allows you to intercept Mars at a closer distance.
And this is what electric propulsion affords you. And so what I have over here is the delta v's that's associated with that rocket equation. We can tell the cost of each one of these missions because we know the delta v to go from low Earth orbit to many destinations here.
This 3.6 kilometers a second, that gets you to the vicinity of Mars. You still have to do a burn to be captured in the circularized orbit. The cartoon shows you a simple orbit transfer from a lower orbit to an upper orbit. Notice that it enters into that transfer ellipse, and then does a burn to circularize its orbit.
So now, let's move to chemical rockets. So for the most part, everybody is familiar with these chemical rockets and the beautiful launches that you see, SpaceX in particular. And so this technology is good for getting us off the Earth and for some onboard propulsion once you're in space, but it has a limitation.
And to contrast with these electric propulsion systems, we're going to look at some of the limitations of chemical rockets and how we can do better. So the primary limitation of chemical rocket is that your heat source and your propellant are the same. So the energy that's released is in the chemical reactions and the combustion chamber.
And so there's a limit to how much energy that can be released from chemical reaction. And it turns out that if you look at the theory, it's around 5,000 meters per second or 5 kilometers per second. This is about the highest exhaust velocity that you can get out of a chemical rocket system.
And then chemical rockets in general are very good at transferring the energy released to direct the kinetic energy-- 60% to 90% efficient. But you have this chemical bond limitation here. And recall the rocket equation-- if your delta v is more than this, then you're going to have some problems.
That's why these systems are mostly propellant. And you have little small capsule on top that's your payload. So chemical rockets have limited exhaust velocity-- again, is due to the reactivity-- the energy release in those chemical bonds.
So this is a nice chart. I like this because it shows how chemical rockets as well as nuclear thermal rockets work really well. So you have your combustion chamber here. That subsonic flow, that is-- that flow is channeled through a converging/diverging nozzle where the flow becomes sonic at the throat, and then it expands and becomes supersonic in the bell.
And this is a plot that shows that the temperature drops because you're converting that thermal energy into directed kinetic energy, and pressure also drops. You end up with these supersonic speeds, but still you're limited to about 5 kilometers a second.
And you can see this specific impulses for various propulsion combinations. So LOX methane is what SpaceX uses. It's about 365 seconds multiplied by 10-- you're going to get the exhaust velocity.
And space shuttle main engines, 457 seconds specific impulse, which is pretty good. And so that's the state of affairs for conventional propulsion technology. Electric propulsion, which is a facet of rocket science that-- whose engines generate thrust by electrical means, is quite different.
The propellant and the energy source are separate. The energy source is electrical, and electricity can be derived from many sources-- nuclear, or electric, or radioactive decay. And so we can add arbitrary amounts of electrical energy into propellant to accelerate it to arbitrarily high velocities.
And so it is possible to have exhaust velocities 10 to 100 times that of a chemical rocket system. And that is the appeal of electric propulsion. The downside is that coupling electrical energy into the propellant is most efficient at low pressure.
So if you want to, say, accelerate ions, you've got to make a plasma, and that happens at low pressure. So inherently, your thrust density is low. And so this puts you in the regime of millinewtons to Newtons as far as thrust, but your exhaust velocities are high.
But that's OK, because if you want to get the delta v, you can either do it over a long period of time as electrical propulsion or a short period of time, as in chemical rocket systems. But because the electric propulsion systems are so efficient, they can also operate for long periods of time and still beat a rocket to its target. So that is the electric propulsion system.
Because its thrust is low, they can only be used in space. And right now, every satellite in the Starlink constellation has electric propulsion on it for attitude control and station keeping. And we're talking about thousands of engines, so this technology has proliferated in space. But the power levels are low.
Electric propulsion can be broken up into three categories depending on how the thrust is produced. The first category is electrothermal. And these systems use electrical energy to heat the propellant, and can do that two ways.
A physical heater to heat the gas up and [INAUDIBLE] resistojets. Or you can use an arc, and arc can reach very high temperatures, so you can get better specific impulse. And these use a converging/diverging nozzle to expand that hot gas into thrust. The first communication satellite to use electric propulsion used an arcjet.
The second category is electrostatic accelerators. And we'll talk about how these work, but these are mature technologies capable of very high specific impulse-- ion engine in particular. Power levels demonstrated, 30 kilowatts-- 50 here, although there have been some experiments that they run well above that.
But if you're talking about closer to state of the art engineering models, this is a good number. And so they're mature. And ion thrusters and Hall thrusters have reasonable efficiency. Gridded ion has the highest.
The final category is the electromagnetic accelerators, and these are a bit less mature, but they have a lot of potential. In fact, a single engine-- a magnetoplasmadynamic thruster can process a megawatt of power with just one engine that's about 30 centimeters in diameter, which is pretty exciting for missions that require thrust and nuclear power.
One thing I want to say as an aside-- electric propulsion systems have really been game changing. A lot of companies are moving to electric propulsion, making their satellites all electric. This is an example of the power of electric propulsion.
This is a plot of delta v's after these satellites have left Earth. Red is delta v imparted by the launch vehicle, the green is by the onboard propulsion. Dawn was an electric propulsion mission that utilized gridded ion thrusters to visit best in series asteroids. And look at the large delta v that that ion propulsion system imparted to that mission.
This is the largest delta v that any onboard propulsion system has imparted to a vehicle. Most of the chemical systems, they have two propulsion onboard. The launch vehicle gives them some delta v, but then you have onboard thrusters, like hydralazine or something. But that's small compared to the total delta v that's important to the vehicle.
This is Deep Space 1, which is net ion thruster mission-- very large delta V by the electric propulsion system. So they can operate for long periods of time and give you high performance. So let's take a look at the engines, but first, I just want to bring your attention to some perspective.
So for a Jupiter grand tour, where you're visiting multiple moons of Jupiter, the total delta v for something like that is going to be like 20 kilometers a second. A chemical rocket system, the mass ratio final to initial is 1 over 150,000. Impractical-- you can't do it [INAUDIBLE]-- doesn't make sense. But for electric propulsion system, 1 over 5.
So now, let's look at the thrusters. So all the thrusters that I'm going to show right now are being considered for high power missions to Mars-- humans to Mars. So the first is gridded ion thruster. Gridded ion thruster shown here-- basically, four components.
And so we'll just walk through how it works. So basically, have a cathode here and an anode. And on the surface, anode are magnets to help confine electrons. You have two closely spaced grids-- that's the third component-- and finally, the neutralizer.
So the way it works is you inject gas into this discharge chamber. This is cylindrically symmetric. You can think of it as a biconic shaped chamber. The gases are typically inert, like xenon.
Xenon is popular because it has high mass and it's easy to ionize. And then you heat up the cathode to temperatures where it will emit electrons. It's called thermionic emission. So those electrons are emitted out of the cathode, because you have a power supply connected between it and anode.
Electrons ionize the gas, producing a dense plasma. They don't go directly to the anode as a direct short, because the magnetic field increases impedance. And so you get nice voltage drop and good ionization efficiency.
The plasma is then extracted between two closely spaced grids. And this is a close up of those two closely spaced grids to see what's going on. A plasma and meniscus forms on the surface at the apertures of the grids. There are two.
One called the screen. The other accelerator grid-- this is negatively biased. This one is positive. Ions leave that plasma meniscus normal to the surface to form beamlets.
And a typical ion engine, like a 30 centimeter ion engine, has thousands and thousands of lined up apertures. So for 30 centimeters, you have about 15,000 apertures, and these are millimeter sized. The grid gap is submillimeter.
You end up with thousands of little beamlets that come out of the engine, and they merge to form the beam. Now, if you just pushing out ions out of the engine, you're going to charge up negatively. And so you have to pump the excess negative charge out, and that's the purpose of the neutralizer-- to neutralize the ion beam.
So it does both current and charge neutralization. And so this technology, as I mentioned, is being considered for these high power missions that we want to couple to nuclear reactors. So this is sort of a throwback engine from the late '60s.
So we actually ran a 200 kilowatt class mercury propellant ion engine, like '67. Can you imagine that-- because at '67, they thought we were going to Mars using nuclear power. And so things have changed a little bit.
We now have an annular ion engine. And this is a cool engine here to me that's almost like Iron Man's light in the middle. But yeah, this is a real picture of ion engine-- annular ion engine operating, capable of processing several hundred kilowatts of power.
If you don't use a single engine, you can use a raise of engines to also achieve that. So this is the array that we tested back in the 2000s. So the next technology is Hall Thruster, and space is being proliferated with Hall thrusters.
Hall thrusters are very popular technology because it has good thrust to power. And so we're going to take a look at how Hall Thruster works. This is a nice little graphic. So let me point out some things.
So you got the same hot cathode here. And notice here-- and these are magnet coils. So you have inner and outer winding magnets. And basically, some cross-section-- it looks like this. So this is your yoke or stator, if you will.
But you have magnetic wires wrapped around it to produce a radial magnetic field, as shown here. And so when you apply voltage, the electrons want to go to the anode, but that magnetic field is there. And the problem with that is that they have to cross the field lines to get to the anode.
This gives rise to the voltage drop at the exit plane of the Hall thruster. And also there's another effect that happens as well. Not only there is the impedance drop across the exit plane because the electrons have to occasionally walk across the field lines, but in a crossed electric and magnetic field-- you probably have done this in your physics class, but the electrons will execute a close drift-- a cycloidal drift called ExB drift.
So the electrons will tend to circulate around in the channel in close orbits. This ExB drift is a basic electron cloud, so that your neutral gas has to pass through that cloud and are quickly ionized. And those ions are not magnetized. Their Larmor radius is too large, and so they're just accelerated by the electric field ballistically.
And they are-- that's how you produce your thrust. And this is a plot of the magnetic field profile at the edge. And also the heat map shows the change in the voltage. So you see the voltage is very positive inside the channel, and it drops off. That's going to accelerate ions rearward.
You get these beautiful plumes here and the cathode jet coming out there. Hall thrusters also are scalable. One way to scale them up is to use multi-channel thrusters. So this is a Michigan multichannel Hall engine.
So you can divide the power in multiple channels, and-- or you can have arrays. So there's going to be a space station put in orbit around the moon. The hardware is already built.
The news release about the engines have been qualified, and these are 12 kilowatt engines in an array that'll drive this space station to its orbit around the moon. And so that should be happening shortly. The final propulsion scheme is one of my favorite, although I like gridded ion thrusters as well. It's the magnetoplasmadynamic thruster. It's too long to write that completely out.
So magnetoplasmadynamic Thruster looks very simple, but the acceleration processes get really complex. And it is still immature technology, but it looks like an arcjet. Basically, you have a cathode and anode.
You apply voltage between the two, and you get an arc. But this is not any arc. This is thousands of amps arc. So you're going to have an azimuthal magnetic field that's formed, and it is the cross field interaction between the current and the magnetic field that blows the plasma out to produce thrust.
This system can process kilowatts-- hundreds of kilowatts to megawatts of power with one engine. And so you don't need a large array of these things to be used for nuclear applications. This is an old school picture of an MPD thruster running on hydrogen. This is a water-cooled magnet around the MPD thruster.
Nowadays, people are looking at superconducting magnets to simplify the magnetic design, because it turns out if you use applied magnetic field, you can get more thrust production mechanisms, like rotational energy converted into directed kinetic energy. And this also allows you to reduce the amount of current that you have to run through the system, because you no longer need that self-magnetic field.
And so that increases the lifetime of things like the cathode. So this is the evolution of the engine. As you increase the magnetic field, you get a more developed plume. You can also see some structure here associated with the magnetic field. And so efficiency of these things, it used to be that these things were like 20% to 30% efficient, but these new magnetic field geometries, there have been literature publications that suggest you can get up to 70% efficiencies.
That has to be repeated in other laboratories, but this is very promising for this technology. And so the idea is to use engines like the MPD, ion engines, and hull thrusters for missions to Mars. Because it turns out that right now, these systems produce low thrust because they've been operated at low powers.
But it turns out if you operate them in a megawatt regime, you can get thrust levels that can push a vehicle to Mars in reasonable trip times. So much faster than rockets, it turns out. And so and that's what we're going to talk about.
So here's a review. These electric propulsion systems coupled with nuclear can allow you to make fast transits. And we need to do fast transits because of the health effects of space. Space is hazardous to your health.
So here, we see the muscle atrophy and the bone loss problem, and these are very real effects. And you're going to be in space for up to three years. You're going to get a good radiation dose.
In fact, it is this radiation dose that is the most problematic. I'll just read this for the sake of time-- the last part of this NASA document. It says a mission to-- a Mars mission will not be feasible unless improved shielding is developed or transit time decreased.
And this is due to the fact that the accumulated dose is just so large. You get the galactic cosmic ray passing through every cell of your body in a few days, and that accumulated dose is too much. You have to get there fast to limit your exposure.
And how do we get there? So there are two mission scenarios for nuclear thermal rockets as well as for chemical-- all chemical as well as for electric propulsion, and these are called conjunction and opposition class missions. And so for astronomy people, it would become obvious why they're called conjunction and opposition.
But the conjunction class missions, you start at Earth, and you do a Hohmann transfer boost and coast until you arrive at Mars. It takes about 200 days to do that. And then you stay on that surface until Mars comes back around again so you can take off.
But the stay on Mars is long. It's like 500 days, but you can do a lot of science there. And you take off from Mars and arrive at Earth 210 days. So you minimize the time that you're actually in space, which is good about the conjunction class.
The downside is you're on Mars for a long time. And it's not clear if we know how to live on another world for over a year like that. So NASA is kind of leaning toward the opposition class mission. And opposition class mission, you start with Earth departure.
It's about 200 days again. It's a Hohmann transfer, but you stay there for anywhere from 30 to 50 days, which is still quite long. But you could probably carry everything you need for a 30-day mission. And then you leave, but Mars is in the wrong spot. And so you have to catch Mars, and one way to do that is do a gravity assist about Venus.
So you actually cross Earth's orbit and do the gravity assist about Venus, and then you come around and land on Mars. This also [INAUDIBLE] exposure-- it exposes you to solar particle events from the sun as well, so it's a little bit dangerous from that standpoint. But it does reduce the total trip time at 650 days as opposed to 916 days, but you're in space for 403 days where you're getting galactic cosmic rays 90% of the time, you're in space.
And so how do you weigh that? But so these are the types of missions that you would use, whether you use a nuclear thermal rocket or a nuclear electric propulsion system. A nuclear electric propulsion system, in order to appreciate it, you have to look at the whole system-- the systems engineering aspect.
It consists of a reactor and a power conversion system. And most power conversion systems are not that great at converting that thermal energy-- that nuclear heat into electrical power. You're talking about 30% to 40% conversion.
And if your reactor is in the 10 megawatt range, that's a lot of power that you got to reject. And so you're rejecting a lot of heat using radiators. The power that you do convert goes to the crew cabin and the rocket engine.
That's why when you look at images of nuclear electric propulsion systems, you see they have these wings. And these aren't solar panels or the aerodynamic structures. They are radiators to radiate all that waste heat, because we still haven't kind of solved that power conversion problem.
Electric propulsion systems, to get the right thrust levels, will have to operate in the megawatt range. So you're kind of stuck with these sort of large scale structures. NASA has plans for nuclear electric propulsion. And right now, there's a bit of a competition between nuclear electric propulsion and nuclear thermal rockets. So there are roadmaps.
There's also experimental hardware being tested at the subscale for both technologies, more so on nuclear thermal rocket. But eventually, there's going to be a downselect, where NASA will pick either nuclear electric propulsion or a nuclear thermal rocket system, which we've talked about in yesterday's lecture. And then from there, we start building the system.
Right now, as I mentioned, investment in the nuclear electric propulsion is a little bit lower than the nuclear thermal rocket, because nuclear thermal rocket is one system, but nuclear electric propulsion, you got your power conversion, you have your reactor, and then there's the thruster itself. And also even competing with nuclear electric propulsion is solar electric propulsion, because it may be possible to use solar arrays, and that only works around Mars.
But the size of a solar panels to do 600 kilowatts is large, so it's difficult to pack it in the shroud of the vehicle. And so it is likely that nuclear electric propulsion is going to win out on that one. National Academies has essentially said that investment in nuclear electric propulsion is important in the modeling and subscale tests. So they have a lot of belief in nuclear electric propulsion.
As a recent development as far as this stuff happening for real, we have the Air Force-- they announced within the past month a new project. It's a nuclear electric propulsion system. Basically, a 20 kilowatt class nuclear electric propulsion system. They want to fly one of these around-- particularly between Earth and the moon to support DOD missions.
And so this is 20 kilowatts electric and your nuclear-- your reactor would probably be around a megawatt thermal. The project is called Jetson. Let me put that up there because I have to.
[CHUCKLES]
It's a joint-- and they worked hard on this one-- Joint Emerging Technology Supplying On-Orbit Nuclear Power. You have to give them credit. But that's-- so they've selected Lockheed Martin to do their paper studies.
But they will-- the plan is to work on component technologies and pick the engines. NASA, on the other hand, will leverage technologies like that to accelerate the development of nuclear electric propulsion systems. But in parallel with that, NASA has more of a evolutionary spiral.
And that evolutionary spiral-- basically, since we're going to the moon-- and that's soon. So by the end of this decade, we should be on the surface of the moon. Is to-- and to power those settlements on the moon is going to be fission surface power-- about 40 kilowatts is the design point.
So you're going to have a lot of practice running nuclear reactors in space. And so the idea is to let this system evolve to moderate power, and then you fly to moderate power systems, 100 kilowatts to a megawatt, and finally, get to the multi-megawatt systems, where you have piloted missions using nuclear electric propulsion. And so where are we with thrusters?
So right now, as far as mature state of the art, we're kind of stuck in the medium power-- 10 kilowatt range, but we need 100 to 200 kilowatts. There have been some technology demonstrations, so let's walk through these thrusters. So go back to-- this is the '90s and this is the '60s.
This is that-- this is one of the tests of that 200 kilowatt engine. It's huge. Look how big it is compared to the opening of the vacuum chamber. That's a problem with testing.
This is a 80 kilowatt MPD thruster. And limitations for it-- you could run that to several hundred kilowatts of power, but the vacuum chamber can't keep up. So that's why there's some limitations.
Here are some additional engines. These are two engines that were developed on the Jupiter Icy Moon orbiter mission-- a 25 kilowatt HiPEP engine. An opportunity to work on this one was pretty cool. This is in parallel with that-- NASA Glenn was developing a 20 kilowatt high power Hall thruster, the 300M.
And then you have the HERMeS Advanced Electric Propulsion system. This is the thruster that would be used on that space station gateway. And so it's very mature, but it's only 13 kilowatts.
This is the annular ion engine. Subscale 6 kilowatts, but that thing is scalable to higher power. Here is the 100 kilowatt X3, also tested. It got to 80 kilowatts, but it is capable of testing-- going more. But it's hard to test these engines to those power levels.
And finally, there's Advanced NEXT engine that basically doubles the power level of the NEXT engine that was on the DART mission that was just mentioned earlier. It's this system-- is probably capable of developing a 10 amp beam, which is pretty impressive for a gridded ion engine.
So that's where we are with thrusters. You can see we have a ways to go-- an order of magnitude, because this is the state of the art. So we're at basically 13 and 14, so we have to get further than that. And in parallel with the engine, we have to also develop the power systems, and that's compact nuclear reactors.
And so while we won't go into detail on nuclear reactors, we're talking about a little bit of history of the nuclear reactors. So it turns out the United States flew a nuclear electric propulsion system back in 1965, if you can believe that. That's SNAP-10A. That was a half a kilowatt rocket-- that's a half a kilowatt satellite.
It used uranium zirconium hydride fuel, and the mission was successful. There was an electrical fault that ultimately ended the mission, but it ran for a while-- 43 days in space. And so it had an ion engine on it, but the ion engine failed during the mission.
So it didn't-- it did turn on, but it didn't provide any delta v or anything. But I think we can call it a nuclear propulsion because it was onboard and they did turn it on. So you would think after '65, that would open the floodgates, but no, it didn't.
It wouldn't be until the '80s that the United States began looking at space nuclear reactors in any serious way. And this was supporting the Star Wars program for missile defense, and that was the SP-100 uranium nitride fuel. And it was a 100 kilowatt class reactor.
And NASA became interested in this reactor as well to support mission to Mars, but it was canceled. And then we get to 2000, project Prometheus. We're going to fly to Jupiter Icy Moon Orbiter mission.
That mission was also canceled, and it was replaced by-- as a consolation prize, Constellation was put forward. And the idea there was to land a nuclear reactor on the surface of the moon at 40 kilowatts of power using highly enriched uranium. And so that was, again, 40 kilowatt system, but that didn't last long either-- cancelled.
And 2010's, KRUSTY-- small compact power. We thought for sure this was going to happen. 1 kilowatt reactor-- could also do 10 kilowatts. Very simple designs-- also cancelled.
I should point out that Russia has flown almost 40 nuclear reactors in space, and this was for their radar sats. Deep penetrating radar to detect submarines. And so some of these are in orbit leaking coolant.
So here's a picture of the SNAP-10A reactor. And this is a picture of a core for SNAP-10A that's-- it's from SNAP-2. But it is-- the fuel pins are similar. Basically, you have stainless steel tubes with your zirconium hydride uranium fuel and these pins.
It was cooled by liquid metal using an electromagnetic pump. So it was circulated in the interstitial spaces between these fuel elements. This is the way SP-100 would look, and this is the way it would look if we went to other worlds with plasma thrusters.
So it was a lithium cooled reactor, so higher temperature. And the lithium was circulated using an electromagnetic pump as well. And its power conversion was thermoelectrics just like SNAP-10A. And you had two loops-- one for the hot side of the thermoelectric converters. And for the cold side of the thermoelectric converters, which are basically like thermocouples, it's a-- this coolant loop passed through a radiator.
These are radiator structures for that one. And so Jupiter Icy Moon Orbiter also had some development work, but SB-100 was significant amount of development work. What are we doing right now as reactor development for nuclear electric propulsion? So there are three reactor designs that are being traded right now.
These are shields, but the reactors are up here. There's a heat pipe cooled reactor, there's a liquid metal cooled reactor, and then there's gas cooled reactors. And what's different is that in the past, we've always used highly enriched uranium. That means that you could-- the size of these things could be smaller and more compact.
But because we like commercial entities to be able to help the development of these technologies, we've moved to High Assay Low Enriched Uranium-- HA-LEU-- which is less than 20% enrichment. So they're going to be bigger, but now, you can have companies and startups develop in these reactors for these types of missions. Just for some scale, how compact nuclear electric propulsion is compared to solar electric propulsion.
These are vehicles for near-Earth orbit rendezvous, and there's no comparison. You had to fold this up into your shroud, and that's very difficult. You can't do it with solar.
And if we're going to Mars, though, you got to use nuclear. Here is some artist's conception of-- actually, these are based on mission studies-- piloted mission going to Mars. This is a 2.5 megawatts electric system.
It has many ion engines on it. In this particular case, you have 10 280 kilowatt graded annular ion engines as your propulsion, and those operate at 5,000 seconds specific impulse. So this is a baseline mission. And just for perspective, for going to Mars, the amount of xenon required to do that mission is equal to the annual production of xenon in the world, not just in the United States.
So that's a lot of xenon. So let's take a look at going to Mars using electric propulsion. So these may look a little fuzzy, but these are straight from the mission analysis people. So this is as crisp as we could get those.
So this is a full-up opposition class mission to Mars using all nuclear electric propulsion. And so we start with-- let's see. We're departing Earth here, and we're basically thrusting the full duration.
Thrusting the full duration. You talk about tens of thousands of hours of continuous operation until you arrive at Mars. And then this is a 45 day surface stay on Mars. And then you leave-- turn your thrusters on again. And this one includes a swing by of Venus for a gravity assist.
And finally, you arrive at Earth. The total trip time is only 700 days, so this is not bad, and this is using total electric propulsion. And the total delta v for that mission is 18.7 kilometers a second.
On time is probably on the order of 30,000 hours, so you need that kind of life. The problem with this particular system is that it requires a high powered nuclear power system to run it. And so that's problematic, because from a technology readiness level, we're just not there at developing such high powered nuclear reactors.
And so NASA went to the drawing board and developed an alternative to all nuclear electric propulsion. It's a hybrid that they came up with, and that is a system that uses electric thrusters and chemical. Chemical you the boost and NEP gives you thrust throughout.
And that can save you a lot, because mass ratio of all chemical, 16-- you can slash that by a third if you use the electric propulsion system along with it. And this plot is telling-- so here is showing the mass required to do the mission. And this gives you a feel for how much propellant is required.
And so the amount of propellant required to do the mission goes down as you run the reactor at higher and higher temperatures, because you're putting more power to the thrusters. You can see that in all electric propulsion system, it's just too heavy. But if you increase the operating temperature of those engines, you can operate at higher power levels and actually be lighter.
But the technology readiness level of something like this is just it's just too low. And so the selection is to just do NEP chemical. So the system, from NASA's standpoint, looks like this. This is their vehicle.
You have radiators, chemical tank-- LOX, so liquid oxygen and liquid hydrogen. And two wings of electric propulsion thrusters. In this case, it's 1,600 kilowatt class Hall thrusters.
[CLEARING THROAT]
Sorry about that. This is a concept of operations of going. So for all these missions, there's always a concept of operations. That is, how do you execute the mission, how do you deploy the assets into orbit before you take off and go, ultimately, to Mars?
And probably, you can recognize these are Starship launchers to assemble the vehicle in orbit. And all of this stuff happens, and 170 days later, you're ready to go. This is-- Aerojet Rocketdyne also posed a mission to Mars using a mixture of chemical and electric propulsion. And--
[CLEARING THROAT]
--sorry. So what I wanted to show you-- bring your attention to is that for this mission, the chemical rocket system's total burn time is only 36 minutes. The electric propulsion system [INAUDIBLE] have to run for 17,000 hours.
And so if we look at this-- so we're departing Earth here. The green is electric propulsion thrusting. And then you coast for a while, and when you arrive at Mars, you do some hyperbolic slow down, and then chemical actually arrive at Mars.
And you stay on Mars, and this is a short stay. This is about a 50 day stay. Then, you use chemical to leave Mars, and electric propulsion to shave off a lot by adding a lot of delta v.
And in this particular scenario, you don't need a swing by of Venus. You just go directly to Earth, as you can see here. And so we almost did a nuclear electric propulsion system. The Jupiter Icy Moon Orbiter mission would have been the first. That was in the 2000's.
That mission was a 1 to 200 kilowatt system-- electric-- that would go to Jupiter and explore Callisto, Ganymede, and Europa. It would have used gridded ion engines-- 7,000 seconds specific impulse. Those each engine would develop a little over half a Newton of thrust, and you'd use four of them.
But unfortunately, that program was canceled. What's happening right now is we have a situation where DOD-- just like what I mentioned with Jetson and nuclear electric propulsion system, DOD has developed a compact nuclear reactor. So there was-- this is a development activity that DOD is interested in, because in the battle theaters, often, the diesel trucks used to resupply the soldiers are targets.
And this is actual footage of a picture of a convoy that had been attacked. And so so to avoid, this it would be better to have a nuclear reactor that provides you with power. And that's-- this compact reactor system here is up to 10 megawatts electric.
You can drive it around in a shipping container. In fact, that's how you would deliver it-- in a shipping container. This is a computer generated image. So it would run for greater than three years without refueling.
And this thing, 2025 is when it's supposed to be delivered and tested. So it is-- and this geometry is consistent with what you need for space as well, so NASA may be able to leverage project Pele, which is kind of exciting. And so that's the intro to nuclear electric propulsion.
The problem right now with nuclear electric propulsion, as I mentioned, is that we just-- we're like an order of magnitude too low in power. So we have to get the power up. And so-- but that's a problem, though, because we have to be able to test and qualify the engine in ground facilities.
So in fact, it's listed here. We have to assess the performance, assess the lifetime, because these things have to operate for, say, 20,000 hours at least. And you have to put margin to that. Sometimes it's 1.5, sometimes it's 2.
So maybe 50,000 hours is the right number. And the thermal margin because-- and so in a space like environment. And so right now, we have a challenge there because our ground test facilities are incapable of handling those flow rates.
The high power means high plasma currents, high plasma density unlike what you see in space. And so an engine will behave differently. And the question is, how do you unpack that?
How do you tease out how it would actually operate in space? And also practical considerations. You sputter and damage the chamber, and also the thermal loading on the cryopumps can be problematic for the vacuum facility as well. So it makes it hard to keep up.
And so these basic issues have to be solved and addressed. This chart shows you where we need to be and where we are. So low-Earth orbit, 10 to the minus 9 Torr in low-Earth. Ground test-- this is just base load-- no engine running at all. 10 to the minus 7.
When you're running, you're much higher. Plasma density, 10 to the 6 for CC and low-Earth orbit. And on the ground, you can get at the 10 to 9 per cubic centimeter. And so what the long-short story is is that these facilities are almost incapable-- the current facilities are incapable of giving us-- if we ran these engines at high power-- giving us the pressure levels so we can credibly say that this is how the engine will operate in space.
And so what are our assets? So some of our biggest assets as far as testing these facilities are shown here. A study was done a while ago that says that for these megawatt class engines, we need 10 to 50 megaliters per second pumping speed.
Right now, if you look at L3 Technologies, this is the xenon ion propulsion system live test chamber. Basically, zips are gridded ion thrusters aboard these galaxy class communication satellites. If you use DirecTV, then they're on your satellite. So you talk about 800,000 liters per second. How do you even-- 1 mega liter per second.
Vacuum Facility 5 at NASA Glenn, 800 as well. And at Michigan, we have a [INAUDIBLE]. It's only 520. And then there's VF 6, which is a huge vacuum facility, but it also has limited pumping speed. It's got large size, but limited pumping speed.
And most recently, we have some exciting news from Aerospace Corporation. They just built a large tank that has a megalitre per second pumping speed. And that thing-- this tank is now operational.
But even that is just not enough. If you look at the pumping speed requirements actually necessary to do things assess the lifetime, assess spacecraft plume interactions, and assess performance, these plots are for 0 to 60 kilowatt power systems. And if you look at the pumping speeds required to credibly do that-- just pick 50 kilowatts for example-- your pumping speed to assess life has to be 3.5 million liters per second. We're at 800 kiloliters per second.
So that's not going to work. So one thing's for certain-- adding more pumps, we're going to run out of space to put the pumps. We can't get the pumping speed. We can test it in orbit.
That would be cool, but we don't have a platform to do that. This is the solution. You can use liquid metal propellants, which are easy to pump. But the technology readiness level for running these engines on lithium and bismuth is very low, and we can't run mercury anymore.
So the only real solution is-- and I guess I'm about to wrap up here. The only real solution is to combine subscale testing with analysis. So using computational models to interpret the test results in the chamber.
So if you make accurate measurements during the tests that exquisitely characterize the test environment, then it is possible to use a computational model to subtract out the effects associated with the facility. And what are some of those effects? Those effects include that when the plume hits the walls, it's going to knock material back and deposit on the engine.
And in this case, if it's carbon, it can make the engine immortal because the sputtering rate of carbon is very low. The plasma environment affects the coupling of electrons from, say, your neutralizer to your ion beam. And the ingestion of gas from this elevated pressure goes into the engine and makes it operate like it's getting additional propellant-- operating rich.
So your performance looks exquisite, but in actuality, it's not. When you go into space, it will operate in different modes. So that is-- so we have this institute that multiple universities, including Michigan, are a part of to combine experiment and modeling to understand how these engines operate in space from test data, where we are not running at full power, but at subscale so we can at least get these effects and subtract them out.
And so I do have-- so as far as this center is concerned, my focus is on the gridded ion engine. And the goal is to look at those facility effects associated with it, because not only this engine has a lot of potential for nuclear propulsion. It has done a lot of work in the past on a lot of missions. And I'll just leave it there with these missions you see. I'm at time.
[APPLAUSE]
SPEAKER: So we will definitely take questions, but before we do, I will invite you all to a reception after this right outside the physics department office. And if you don't know where that is, it's going to be pretty easy to find. Follow the big group of people going towards the beverage and food. So let's have some questions for John. Yep, Abby?
AUDIENCE: My question is about how you actually do the pumping in the large chambers. You mentioned cryopumping, but are there other [INAUDIBLE] primary [INAUDIBLE] that you use for that?
JOHN FOSTER: Yeah, so we've-- and the whole field has moved almost exclusively to cryopumps. There's always backup. There's always turbopumps to pump things like hydrogen and helium. But for the most part, yeah, we pump down with dry pumps initially, and then turbo to take us down to very low levels, and then use cryo tubs.
The facilities are peppered with cryo panels and cryo tubs. And those things can't take much heat load, and so that's a problem as well. But it's mostly-- yeah.
SPEAKER: More questions for John?
JOHN FOSTER: There are no questions?
[CHUCKLES]
SPEAKER: Right there.
AUDIENCE: Silly question, but like [INAUDIBLE] you're saying [INAUDIBLE]?
JOHN FOSTER: Oh, you mean just running the engine?
AUDIENCE: Yeah.
JOHN FOSTER: Yeah-- in a poor vacuum?
AUDIENCE: Yeah.
JOHN FOSTER: Yeah, that is exactly what we're doing. So I have a lot of backup charts, and let me go to one. This is embarrassing, but I have a lot of backup charts. But I will go to-- I made a lot. So--
[LAUGHTER]
--let me go to-- yeah, this one. So this is a test that we did recently. So you see the engine, and it's operating. And that's a graphite target. That different color is actually carbon plasma, and that stuff can come back onto the engine and coat it.
And in fact, for life tests, the coatings can be so thick, you can see it's flaking off. So this is two problems. One is that that coating is-- it makes the engine immortal, because as I said, it's like, it's a carbon coating. It's hard to sputter it off.
But when it does fall off, because it's so thick, it can short the grids out, at least for gridded ion engines. And so when we had this very long test-- this 50,000 hour life test of an engine, it didn't have a whole lot of wear, but it had a whole lot of carbon coatings on it. And so the modelers, though, can take that data and try to subtract out those effects associated with that. So the question is a good one, and that's really our only hope is to run it because that's the best we can do and try to use models to figure it out.
SPEAKER: Right there.
AUDIENCE: One of the previous slides, you had a graph that showed pure nuclear electric propulsion system would actually have more-- higher mass to orbit. Could you go over--
JOHN FOSTER: Say that again?
AUDIENCE: So if you had pure nuclear electric propulsion [INAUDIBLE] would actually have a higher mass than chemical? Could you go over that?
JOHN FOSTER: Yeah. Oh, that's really far away. Let me get out of this and go there. Yeah, that's because the-- for that high-- so that's all nuclear electrical propulsion. So it's like, it's high power. Let me see where it is.
SPEAKER: There it is.
JOHN FOSTER: I think it's this one.
SPEAKER: It's that one.
JOHN FOSTER: Yeah. Let me-- yeah, this one. That's the 1,200 kilowatt system. So you're going to need more propellant, because you don't-- at 1,200 kilowatts, you're not producing as much electrical power.
And so you have basically lower power going to your thrusters. So they have to-- yeah, they have to-- you're going to need more propellant in order to get the delta v, because your-- yeah, I guess that's it. I mean, the system is heavy and it is lower powered.
And so you just-- yeah, it has to do with the power level of the system. You need more xenon to do the mission. Does that make sense?
AUDIENCE: [INAUDIBLE]
JOHN FOSTER: So you're-- it's still using the same ISP. It's just that-- so as you go to higher power, you can either add more engines to get more performance, because you have more power available at the higher temperatures. So you have better conversion efficiency.
And also if you run at the higher powers, your-- at higher temperatures, your radiator mass also goes down because it's-- a lot of the mass of the system is in the radiator. And this is at 1,200 as opposed to 1,800, so your radiator mass goes way down.
It's just that our technology readiness level to run-- to build engines at these high temperatures is just not there. And so so you're stuck with-- essentially, for these systems-- yeah, I guess if you're forcing it-- if you're forcing the system to run in this power level range, you got a lot of waste heat because you're only at 1,200. And so you radiate-- you have a lot more radiator mass.
I guess that's really the answer-- it's the radiator mass, because you got to get rid of the heat. And at 1,200, you have to have larger arrays to get rid of it then than if you were at higher temperatures, because it goes-- it's T to the 4. Yeah, that's a good question.
SPEAKER: I think there was a question over here. Carl?
AUDIENCE: I remember hearing that people worked really hard on thermoelectric generators, [INAUDIBLE]. But he mentioned that in the past including space as power sources for electric power. Can you comment about where they're useful right now [INAUDIBLE]?
JOHN FOSTER: You mean thermoelectric converters.
AUDIENCE: Yeah.
JOHN FOSTER: So right now, the state of the art is under 10%. And so the power conversion efficiency is still fairly low. And with these megawatt systems, it's-- the rate of your mass is going to be pretty large.
AUDIENCE: There were applications where that was the best thing to do?
JOHN FOSTER: So the first reactor that we ever flew was-- used thermoelectrics for power conversion. SB-100 was also baseline for thermoelectrics. And SB-100 had a lot of radiator surfaces. This could actually deploy out like a fan. These are radiators.
So what's nice about thermoelectrics is it's simple and it's reliable. All of the RTGs use thermoelectrics for power conversion. So it is effective at converting the heat to power, but it is-- its efficiency is just low.
Even if you projecting, it's-- I don't think it's much above 12%. But yeah. I mean, it's actually on the list. I mean, this is from NASA. They put it there, but radiator mass would be substantial.
SPEAKER: We have time for one more question in the very back. Yep, go ahead.
AUDIENCE: How was the ISP or ion engines affected by increases in [INAUDIBLE]?
JOHN FOSTER: No, it's-- for gridded ion engine, it's almost exclusively due to the voltage between those grids. Hall thrusters, on the other hand, you can control the-- if you run in constant current mode, you can control the accelerating potential by varying the radial magnetic field and get more acceleration. But gridded ion thrusters, it's all the voltage to the grids.
Magnetic fields-- all the magnetic field geometry does for you is it increases the plasma density. And so I guess it would have a second order effect, because if you can improve your propellant utilization, your effective specific impulse is going to go up because you're not wasting as much propellant. But yeah, it's second order. For the most part, it's like those things run at 90% propellant utilization. So they're pretty sporty.
SPEAKER: I think that's a great place to end. Let's thank our speaker again.
[APPLAUSE]
This public lecture was held on Wednesday, November 15, 2023 in Schwartz Auditorium, Rockefeller Hall by Prof. John Foster, University of Michigan.
High power electric propulsion and nuclear propulsion systems offer the promise of liberating humans from low Earth orbit and enabling quick trip times throughout the solar system. But how do we practically test these engines on the ground? The energetic plumes these rockets produce interact with the walls of vacuum chambers which are supposed to simulate space-like conditions, leading to high backsputter rates and elevated chamber pressure. If these problems are unavoidable, how do we account for these processes when interpreting the test data and then extrapolating to space operation? In this talk, we survey the challenge to testing high power electric propulsion on the ground using the gridded ion thruster as an example, unpacking the myriad of physical processes that can impact interpretation of on-orbit performance. We also survey past ground test methods of characterizing nuclear thermal rockets—which are prohibited today—and current approaches being posed to test, understand, and characterize performance of these engines on the ground.
