WEBVTT 1 00:00:04.940 --> 00:00:08.380 David J McLean: Thanks. 2 00:00:08.710 --> 00:00:17.610 David J McLean: Good afternoon, everybody. Good afternoon. My name is Tony Boss. I'm the chair of aerospace engine here at Michigan. 3 00:00:17.790 --> 00:00:20.480 David J McLean: and it's a pleasure for me to introduce 4 00:00:20.660 --> 00:00:34.390 David J McLean: today's seminar speaker, who is our own Ben Jones, who is an assistant who is an associate professor, Sorry, genuine associate professor in the department of Aerospace engineering at Michigan. 5 00:00:34.490 --> 00:00:36.410 David J McLean: So, by way of background. 6 00:00:37.730 --> 00:00:45.110 David J McLean: Ben Co. Direct the plus my dynamics and electric propulsion lab here in Airlines, Michigan. 7 00:00:45.440 --> 00:00:50.100 David J McLean: prior to joining us. Ben was a member of the Electric Propulsion group 8 00:00:50.280 --> 00:01:04.800 David J McLean: at NASA's Propulsion Lab in Pasadena, California, where he's work from buying experimental and analytical techniques to investigate proportion systems for the next generation of NASA robotic missions. 9 00:01:05.170 --> 00:01:07.920 David J McLean: His primary research interests include. 10 00:01:07.930 --> 00:01:22.720 David J McLean: We are mechanisms and stability in electric propulsion systems, turbulence and nonlinear processes in low temperature, plasma developing new plasma diagnostics and investigating breakthrough forms in space propulsion. 11 00:01:23.260 --> 00:01:36.520 David J McLean: He's published over 150 journal papers and conference proceedings in the field. He's an associate fellow of AI and a member of I Triple Eps and the Electric Rocket Propulsion Society. 12 00:01:37.090 --> 00:01:46.790 David J McLean: His work has been recognized with 5 a. Iw. Based paper awards The Air Force Young Investigator Program award The DOE Early Career Award. 13 00:01:46.870 --> 00:01:51.620 David J McLean: The Iw. Is very award, and a number of NASA technical 14 00:01:51.870 --> 00:01:58.380 David J McLean: Dr. Johns is currently serving as co-director of the joint advanced propulsion institute 15 00:01:58.480 --> 00:02:13.160 David J McLean: which is a 5 year strategy technology Research Institute supported by NASA to Investigate testing of high power electric propulsion systems. I'm sure some of you have seen him here. So let's welcome Ben to give this lecture. 16 00:02:21.220 --> 00:02:27.560 David J McLean: Okay, thank you, Tony. It's always great to be back in Ann Arbor. So thank you for that. Okay. 17 00:02:28.400 --> 00:02:32.820 David J McLean: I I am grateful for this opportunity to talk about 18 00:02:33.010 --> 00:02:40.900 David J McLean: this research we've been doing in my group, which is a broadly, a technology development program that we undertook internally 2 years ago. 19 00:02:41.550 --> 00:02:51.240 David J McLean: and the overarching goal was to, as the title says, Push the limits of hall effect, trust, or technology to explore its capabilities for next generation 20 00:02:51.750 --> 00:03:01.370 David J McLean: space exploration. And by that I mean, there are rising challenges, both for deep space exploration as well as in this lunar that are calling for new technological advances. 21 00:03:01.660 --> 00:03:04.820 David J McLean: and to meet those challenges we need to innovate 22 00:03:04.980 --> 00:03:12.950 David J McLean: on multiple forms of subsystems and technologies for space transportation. primarily for both power as well as propulsion. 23 00:03:13.740 --> 00:03:22.040 David J McLean: So with that in mind, hall effect thrusters are by far and large, the most widely flown types of elect propulsion in space to date. And we'll talk about Why. 24 00:03:22.200 --> 00:03:24.180 David J McLean: that is the case here shortly. 25 00:03:24.260 --> 00:03:42.160 David J McLean: but they also may potentially have a niche for meeting these growing challenges for deep space exploration. That being said, there are some fundamental limitations on how they operate that historically have thought to limit their extensibility to these deep space applications. 26 00:03:42.410 --> 00:03:46.740 David J McLean: and her overarching purpose was to explore how valid those ultimately are. 27 00:03:47.180 --> 00:03:53.670 David J McLean: So the purpose of today's talk is to first outline why hall thrusters are so great in particular exp propulsion. 28 00:03:54.100 --> 00:04:04.470 David J McLean: The next turn of the question of why we think they are currently limited. What is the state of the art understanding? Why, in turn, we thought that we might be able to push the envelope beyond those current limitations and capability. 29 00:04:04.660 --> 00:04:08.230 David J McLean: how we went about doing that. And finally, what are the implications of those results? 30 00:04:08.900 --> 00:04:28.510 David J McLean: So to give some mission context to this, let's start with the big one, which is Mars, and how ultimately do we get there? So most modern mission projections suggest. We need to send about 100 metric tons of dry mass to Mars in order to support a crude mission. Somebody would last name from 60 days to several 100 days on the surface of Mars. 31 00:04:28.510 --> 00:04:32.490 David J McLean: and the key question is, in order to enable that type of 32 00:04:32.650 --> 00:04:34.040 David J McLean: architecture. 33 00:04:36.840 --> 00:04:49.840 David J McLean: how much mass do we ultimately need to send from a low Earth orbit to get out there, and of course we need some reactory mass in the form of propellant to enable this mission, and the fundamental systems level question is, how much is associated with that. 34 00:04:50.290 --> 00:05:04.420 David J McLean: Well, the lowest thing you approve is to start with established technology, which should be chemical propulsion engines. We use locks. L. H. 2 we use locks, Rp. One their near-term projections for using locks methane to propel spacecraft from a low earth orbit to a Mars trajectory. 35 00:05:05.270 --> 00:05:21.690 David J McLean: And indeed, if we baseline that architecture and others have done this study. Here's an example of a locks on H, 2 engine. We find that the actual mass. This includes both the dry mass, 100 metric tons distributed across cargo and crew vehicles, plus the amount of reactionary mass or propellant on board scales. 36 00:05:21.690 --> 00:05:31.720 David J McLean: quite substantially, almost 10 times in the case of Lox. Lh. 2 and energetically favorable propellant compared to the dry mass, and 20 times can, when we use locks. 37 00:05:31.740 --> 00:05:33.620 David J McLean: methane combination. 38 00:05:34.820 --> 00:05:50.420 David J McLean: So the key logistical challenge when you recognize this kind of mass budget required for a Mars architecture leaving lowered orbit is, how do you actually get that mass initially into space and to date there's effectively no launch vehicles. It can send a 1 million kilograms like that in the space at once. 39 00:05:51.200 --> 00:05:59.440 David J McLean: So we also have to adopt a staged architecture, and this is kind of where the key challenge within a Mars approach comes to a. For 40 00:06:00.140 --> 00:06:17.650 David J McLean: so we could baseline, for example, the recently demonstrated Sls Mission, which has 130 metric tons to a low Earth orbit, and recognizing that if we use a locks L. H 2 combination. We'll need to get somewhere at the tune of 1,200 metric tons in the space. So one launch isn't going to cut it. In fact, we'd have to do 41 00:06:17.900 --> 00:06:26.070 David J McLean: 10 launches right. So this is becoming prohibitively difficult, and the probability of failure for these launches actually can approach somewhere on the order of 60. 42 00:06:27.640 --> 00:06:38.880 David J McLean: Now, if you do a locks methane architecture, which is a spacex approach, and you're going to leverage starship here, which hopefully will be going up here in the next couple of weeks they project 115 metric tons to Leo. 43 00:06:38.880 --> 00:06:51.790 Now, again, because locke's methane is a less energy genetically favorable, propellant. There needs to be more wet Mass and Leo, and that translates to more launches, despite the fact that starship has a little bit more margin in terms of launch mass to space. 44 00:06:52.250 --> 00:07:03.840 David J McLean: So if you go with the Spacex approach, which is where they're going to send the mission to Mars and then recycle, reuse some of the Martian atmosphere to make methane. Then you could go away with something like 8 launches 45 00:07:04.070 --> 00:07:08.930 David J McLean: alternatively, if you have to take all the propellant on board with you. Now we're talking about 16 launches. 46 00:07:09.050 --> 00:07:21.540 David J McLean: Now, not to say this is impossible, and of course sarship building on Spacex heritage is meant to be reusable, but I think at the they, just at the first blush. Look at this. We can imagine launching these 47 00:07:21.540 --> 00:07:31.180 David J McLean: 30 story high skyscrapers 16 times reliably, and assembling the subsequent payload in orbit can become logistically complex and extraordinarily difficult and expensive. 48 00:07:31.340 --> 00:07:41.020 David J McLean: And, indeed, I think, broadly speaking, this is one of the major limitations for a Mars architecture, Leaving aside the human element, which is a number of launches required, becomes expensive and risky. 49 00:07:41.640 --> 00:07:59.690 David J McLean: So with faced with that kind of technological limitation invites the question: Are there other transportation solutions, other propulsive technologies that could enable us to get to Mars, but with fewer launches and indeed that boils down to the key question, Can we start off with less wet mass and a low Earth orbit to ultimately deliver the same payload? 50 00:08:00.200 --> 00:08:11.620 David J McLean: The answer is that if we rely on next generation technologies most notably so-called electric repulsion technologies it does substantially buy down that risk translating to substantially fewer launches to a lower orbit. 51 00:08:11.880 --> 00:08:17.920 David J McLean: So let me say a quick word about why electric propulsion has that capability, and that will hopefully motivate our subsequent discussion. 52 00:08:17.970 --> 00:08:33.250 David J McLean: All right. So the best way to talk about electric propulsion is to draw an analogy, and just as we have our gas guzzling humvees on the on the ground. We have our lease. What? I drive here on the right hand side. Okay. The lease of space. And then we have a lecture repulsion here on the lower bottom 53 00:08:33.250 --> 00:08:45.770 David J McLean: effectively. These translate this analogy of holes. Gas Gasoline cars rely on combustion of hydrocarbon-based fuels, as do chemical rockets. Electric Propulsion system is fundamentally relying the conversion of electrical energy into propulsive energy. 54 00:08:46.330 --> 00:09:10.730 David J McLean: Okay. So to understand why Ep. Has these advantages, let's contrast it with these chemical rockets Specifically, we can consider kind of the canonical chemical rocket architecture, which is comprised of an oxidizer and fuel. In this case i'm. Showing liquid oxygen and liquid hydrogen, which then combusts in a combustion chamber, releasing that chemical energy converted to thermal. The thermal is converted to direct to kinetic through the action of a shaped duct right 55 00:09:11.100 --> 00:09:32.190 David J McLean: at the end of the day. The degree to which you can accelerate that propellant is going to be intrinsically linked to how much energy is stored up in the chemical bonds of that oxidizer and fuel. And indeed, it can be shown through some energetic scaling arguments. The max achievable exhaust velocity of that propellant is going to be dictated by the energy per unit mass stored up in those oxidizer and hydrogen bonds. 56 00:09:33.120 --> 00:09:44.720 David J McLean: As a direct consequence, we are fundamentally limited by the periodic table in terms of how fast we can send propellant up the back when we use a chemical propulsion architecture, and indeed, for locks at H 2, which is the most energetically favorable 57 00:09:44.730 --> 00:10:04.670 David J McLean: chemical propulsion system that we use to date and, Max, we can achieve exhaust velocities on the order of 4.5 kilometers a second. The let's repulsion breaks free this paradigm, and so much as it it no longer relies on the release of stored up chemical energy to energize repellent instead, it can rely on external power supplies. I've shown symbolically here 58 00:10:04.850 --> 00:10:22.060 David J McLean: in this case the external power supply, as I'm. Shown in, is solar arrays, or it can be a nuclear propulsion system which converts, or it takes in electrical power, uses that electrical power to ionize and then energize the repellent and then accelerate it out, the back of the vehicle generating reactionary force. 59 00:10:22.390 --> 00:10:38.910 David J McLean: Now the key enabling feature of this is that it can be shown again from an energetics argument that the speed at which we can accelerate for pound off the back does not scale like the energetic content of the propellant, but actually scales monotonically with the amount of power that we deliver per unit flow rate up propellant. 60 00:10:39.130 --> 00:10:44.570 David J McLean: This is a really enabling feature, because in principle, if we scale up the amount of power delivered to the propellant 61 00:10:44.720 --> 00:10:53.480 David J McLean: Steve, Arbitrarily high exhaust velocities, and indeed, state of the art systems can achieve orders of magnitude higher than chemical rockets somewhere to the tune of 30 kilometers per second. 62 00:10:54.140 --> 00:11:08.840 David J McLean: All right. So practically, why is this ability to accelerate propellant to high speeds? Translate to major savings and fuel economy. Well, that all comes down to Isaac Newton all right. Who effectively tells us that the only way we can travel in space 63 00:11:09.070 --> 00:11:13.180 David J McLean: is the reactionary force. There's nothing for us to push off against except for ourselves. 64 00:11:13.640 --> 00:11:33.280 David J McLean: So at the end of the day, if we want to move in the opposite direction. Generating some sort of impulse, we have to repel both propellant propellant and momentum out the back of our vehicle. Now the ultimate amount of momentum that we can transfer in the opposite direction is denoted by this trust factor is a product of the amount of propellant that we found out. The back as well as the speed at which we send it. 65 00:11:33.370 --> 00:11:41.610 David J McLean: Momentum goes like mass times velocity. This is kind of the traditional chemical architecture When we send a lot of propellant out the back at relatively slow speed. 66 00:11:42.220 --> 00:12:03.580 David J McLean: With electric propulsion systems. However, we can get away with sending just a little bit of propellant out the back the smaller cloud at very high speed. and the implication is that at the end of the day we can still impart the same total momentum to the vehicle, but with substantially less reactionary mass, and then all stems directly. From this momentum goes like mass times velocity savings. 67 00:12:03.860 --> 00:12:19.510 So at the end of the day. The ability to send for pound out the back of your vehicle at higher speeds translates to overall higher fuel economy. As a direct consequence. This is an enabling feature for Mars architectures, allowing us to get away with less reactionary mass and orbit for our transit to Mars. 68 00:12:19.600 --> 00:12:29.870 So let's reconsider this mission architecture, where, instead of using a chemical propulsion system, now we've baseline in a lecture propulsion system, which in this case is powered by these massive nuclear reactors. 69 00:12:30.960 --> 00:12:36.820 David J McLean: and we're going to zoom in on the propulsion system. And in this case I've shown our favorite here, which is a hall effect thruster. 70 00:12:37.240 --> 00:12:47.210 David J McLean: So if you work through the math. Here, again for comparison are 2 chemical propulsion systems, about 1,200 metric tons, 2,400 metric tons, depending on the propellant oxidizer that you choose. 71 00:12:47.310 --> 00:13:01.320 David J McLean: If you let the state of the yard election portion system in terms of the achievable exhaust velocities, you can get a reduction of overall mass of 500 metric tons. Okay, all stems from the fact that we can accelerate them, filling out the back in much higher speeds. Right. 72 00:13:01.440 --> 00:13:14.940 David J McLean: So if we revisit this mission architecture for locks, methane for sending all that dry mass to Mars, using a chemical portion system, if we use an electric propulsion system, we can get away with just 4 launches, right? 73 00:13:15.010 --> 00:13:30.480 David J McLean: And so, right off the back, contrasting those 2 we see strictly from my risk and costs perspective, Elect propulsion has a major advantage, and for this reason that we say Ep. Could ultimately be a high in the highly enabling technology for enabling future. Mars architectures 74 00:13:30.750 --> 00:13:46.970 David J McLean: with that said, Although electric repulsion systems are widely flowing and considered a mature technology for geocentric applications, those that exist between the Earth and the moon, and to a limited extent for robotic exploration to deep space. These types of technologies have never really be considered seriously 75 00:13:46.970 --> 00:13:56.220 David J McLean: in in kind of the state of the yard, or today's technologies for Mars based architectures, and that stems from the fact that there are several key technological challenges. 76 00:13:56.510 --> 00:14:02.980 David J McLean: So to capture those at a high level. They're represented again by the mission and architecture here Oops. 77 00:14:04.610 --> 00:14:06.110 David J McLean: So the first is 78 00:14:06.440 --> 00:14:23.380 David J McLean: for us. Okay. So modern electric proportion systems, I said, are used for Cis lunar applications. The maximum amount of force that they can generate on those vehicles for robotic applications has been demonstrated about a half in Newton. Right? So imagine a handful of sheets of paper on your hand. That's about as much thrust as they can generate. 79 00:14:23.450 --> 00:14:24.120 Right. 80 00:14:24.510 --> 00:14:29.890 David J McLean: You contrast that with something like the Saturn 5, which is able to launch a skyscraper effectively. 81 00:14:30.750 --> 00:14:41.460 David J McLean: Now, in order to enable this mission architecture, we need orders of magnitude and increase in the amount of thrust generated by this vehicle to in turn translate to the accelerations required for a relatively rapid transit to Mars. 82 00:14:41.990 --> 00:14:57.850 David J McLean: Okay, at the same time hand in hand with that requirement. Is it the frost that can be generated by an electric propulsion system intuitively scales? With how much power that we're applying to the electric propulsion system to date. Electriculsion systems use solar power, and at most that's been available on orbits. Been about 4 and a half kilowatts. 83 00:14:57.850 --> 00:15:06.830 But we need orders of magnitude higher than that to enable a Mars architecture somewhere to the tune of 2 to 4 megawatts, which would be a little bit shy of what you need to power in our room. 84 00:15:06.900 --> 00:15:24.500 David J McLean: All right. So these kind of key technological questions, of course, invite the larger technological question. How do we develop the technologies, and what kind of technical roadmap is viable to achieve them. To accomplish a Mars architecture in the realizable near term, let's say 10 to 15 years 85 00:15:25.140 --> 00:15:39.090 David J McLean: with this in mind a NASA commission to study, just 2 years ago by the Nash, conducted by the National Academy of Science, which looked at this key question, and of course they identified that the 2 lemonade technologies 86 00:15:41.010 --> 00:15:56.450 David J McLean: for enabling this architecture are the power system, the propulsion system. And in here, in broad strokes is a recommendation which is, we need to start this technology development yesterday. All right. Not only we need to start yesterday. You need to start dumping billions of dollars into it. But of course I love to hear maybe Congress not so much 87 00:15:58.020 --> 00:16:15.280 David J McLean: so, for now i'm going to give you the power system for free. And indeed, NASA, as well as the Department of Defense, have already start to invest in next generation nuclear electric propulsion systems for space, right. There are lots of applications that go beyond propulsion for having reliable nuclear power and space. Right? 88 00:16:15.280 --> 00:16:25.050 David J McLean: So let's assume that that's a given the key question that we've been addressing as a community, and in my group as well, is what type of electric propulsion system is best suited for this architecture. 89 00:16:25.580 --> 00:16:26.680 David J McLean: Right? 90 00:16:27.610 --> 00:16:45.230 David J McLean: Well, electric repulsion systems fall under a broad, wide, ranging series of categories, and effectively it boils down to the point that there are many ways you can turn electricity into accelerated kinetic energy. You can heat up a plasma and expand it much like a dimensional nozzle. Those are thermal expansion propulsion. Systems. 91 00:16:45.230 --> 00:17:05.119 David J McLean: You can convert that electricity into static electric fields, ionize a plasma and accelerate the heavier species, the ions downstream that so called electrostatic acceleration, or it could leverage a combination of electric and magnetic fields which are generated by that initial electrical power to accelerate your ionized propellant downstream again at very high speeds. 92 00:17:05.119 --> 00:17:18.700 David J McLean: I would say to date by my last estimation, there at least 30 or 40 different propulsion concepts that are out there, that leverage electric propulsion. But of course that invites the key question which one is most applicable for the near term Mars architecture. 93 00:17:18.819 --> 00:17:40.070 David J McLean: And i'd say, broadly speaking, the community has landed on this consensus that do either to their existing maturity or potential major benefits. These are considered the leading contenders, so called, gritted ion thrusters all effect thrusters which we'll talk about in more detail today, as well as these things called magneto plasma dynamic thrusters, which is a mouthful to say so. Subsequently. I will call them Mpds. 94 00:17:40.070 --> 00:17:40.660 All right 95 00:17:43.210 --> 00:17:54.330 David J McLean: Now the key challenge with this is that all of these technologies largely have been demonstrated at 10 kilowatts, and remember that our goal is to be able to scale them up to Megawatt level power sizes. 96 00:17:58.880 --> 00:18:16.190 David J McLean: So to give you a sense of scale here, there's a person standing next to each of those engines that has been demonstrated at the 10 kilowatt power level, and again till it 10 kilowatts is about the amount of power that's available to us in orbit. Right now many of these have been successfully flown for commercial applications as well as deep space, robotic exploration. 97 00:18:16.580 --> 00:18:43.400 David J McLean: The challenge, however, becomes, when you want to scale these technologies up to higher power. And indeed, if we want to do that by an order of magnitude, we start to run into limitations in size and weight. So let's take these 10 kilowatt systems and pose the question using existing design laws, how big will they ultimately be? So now, here's our little person on the side, and we scale these technologies up to 250 kilowatts. So a factor of 25 times higher. And remember, we need 4 megawatts to the 4 megawatts in orbit. 98 00:18:46.890 --> 00:18:47.730 David J McLean: Okay. 99 00:18:49.430 --> 00:18:53.230 David J McLean: let me pause here because it is playing through for me. 100 00:18:55.980 --> 00:18:56.810 David J McLean: All right. 101 00:18:58.120 --> 00:19:07.220 David J McLean: So suddenly these engines start to become kind of prohibitively large. And indeed, if we need 10 of these engines on a spacecraft, soon it dwarfs the overall spacecraft size. 102 00:19:11.510 --> 00:19:16.530 David J McLean: It's for this reason that if you can contrast this all thruster with the gridded Ion thruster. 103 00:19:19.630 --> 00:19:22.230 David J McLean: All right. Sorry. Let me take a pause here 104 00:19:28.990 --> 00:19:30.120 David J McLean: hopefully. That will do it 105 00:19:34.060 --> 00:19:48.840 David J McLean: if you can contrast these different technologies, magnetoplas, dynamic registers, Npds seem to be far and away. The winner in so much is that you can get away with much higher power through a much smaller footprint. So this is an attractive from an overall vehicle development, integration, perspective. 106 00:19:49.360 --> 00:19:59.970 David J McLean: and it is in large part for that reason that I'd say the conventional wisdom in the broader electric propulsion community has been that magnetoplas dynamic thrusters are the de facto choice for enabling a Mars architecture. 107 00:20:00.250 --> 00:20:04.070 David J McLean: Okay, that all comes with one major caveat right? 108 00:20:04.140 --> 00:20:28.800 David J McLean: And to illustrate that, let's consider the flight heritage of each of these technologies. So let's start with hall thrusters and hall frosters. At this point there are 3,000 on Orbit credit. Where credits do this is primarily Spacex is doing this. Each one of their starling satellites has a hall thruster on board, but before they got into the business there was someone around 300 to 400 hall thrusters successfully demonstrated on orbit since the early 19 nineties. 109 00:20:29.230 --> 00:20:39.800 David J McLean: Okay, grid an iron thruster is not as impressive. Okay, they become a less popular and more recent decades. But there's have been 200, demonstrate, and some quite successfully, in recent robotic exploration missions. 110 00:20:40.370 --> 00:20:49.590 David J McLean: Now let's compare that to Mvd. Thrusters thanks. and I say approximately one. Why, approximately well as a demonstration mission that flew for about 10 s. 111 00:20:51.420 --> 00:21:20.450 David J McLean: So if you look at this graph, and you contrast that with the advantages in terms of footprints. It seems like there might be a trade space there. We seem to to know how hall thrusters work, at least reliably enough that we can fly them for operational applications. But Mpd. Thrusters, and despite all the major advantages that they have, at least theoretically do not have any flight heritage, and that invites a lot of key risk and key. Expensive questions about how you'd ultimately flight qualify them for these next generation. Mars architectures 112 00:21:21.270 --> 00:21:35.240 David J McLean: with that said, Despite the fact that Npds have low heritage. They still seem to be the de facto choice, and again it largely comes down to that mass consideration. You can get away with much higher thrust levels for a much smaller footprint. 113 00:21:36.450 --> 00:21:51.250 David J McLean: All right. So here we are again at the trade space, and this is where we started 2 years ago in this technology development program that we started at the University of Michigan. And the key question that led this, this, this investigation off. Was this okay. 114 00:21:51.250 --> 00:22:00.470 The reason why the hall thruster is so large is that we have been scaling it based off of conventional scaling laws that have been given to us by hall effect thruster developers 115 00:22:00.620 --> 00:22:03.670 David J McLean: and perfected over the past 20 to 25 years. 116 00:22:03.680 --> 00:22:20.470 David J McLean: Okay, leveraging these scaling laws, they built several 3,000 successful fight units to date. But the question that we wanted to ask is, could, in fact, be possible to circumvent those stealing laws? Is it possible to make a same level of thrust as an Mpd. Fruster for a comparable area. 117 00:22:21.220 --> 00:22:34.770 David J McLean: if that were in fact possible. And we can leverage all the technical know how that we've developed for lower power hall thruster systems soon. This might be an attractive and even more competitive candidate compared to the less mature and meneto plasm dynamic thrusters. 118 00:22:35.190 --> 00:22:36.060 David J McLean: Okay. 119 00:22:36.330 --> 00:22:41.660 David J McLean: So that's kind of the central question that we started with. And to take a deeper 120 00:22:41.660 --> 00:22:56.450 David J McLean: discussion of that. Here are the 3 kind of key questions that I want to talk about in the context of this problem. So first is, what are hall thrusters? Let's talk a little bit more about the physics of them number 2 is, why do we currently think that thrust them? And density is, listen limited for them? 121 00:22:56.450 --> 00:23:03.510 David J McLean: And third is, why do we think, from a first principals perspective, we might be able to overcome these current limitations. 122 00:23:03.570 --> 00:23:12.660 So i'm going to talk a little bit now about the principal operation of hall thrusters, and we're going to leverage that subsequently to inform our discussion of these other 2 T key questions. 123 00:23:13.090 --> 00:23:23.170 David J McLean: So Halifax frustrated. Let's start with this 3 quarters view here. So on the right hand side this is an actual hall effect, frustrating operation. It's characterized by this access symmetry 124 00:23:23.170 --> 00:23:42.910 David J McLean: with a kind of glowing blue doughnut in this case operate on the Xenon gas, which is propelling propellant downstream and relatively high speed, and for comparison here I have a 3 quarters view the overall Canonical Hall thruster. Architecture is characterized by a central core as well as an external core, that 125 00:23:42.910 --> 00:23:49.830 David J McLean: it serves to generate and direct a magnetic field which is directed radially across this confining discharge channel. 126 00:23:50.150 --> 00:24:01.130 and a strong electric field is applied between an upstream. Anode is shown here, and a downstream cathode that serves as the electron source as well as the termination for that applied electric field. 127 00:24:01.780 --> 00:24:15.330 David J McLean: So i'm going to walk through the multiple steps of this operation. I'm going to translate now from that image, keeping my 3 quarters view here and now we're going to do just a pure across section. So remember it has X symmetry around the central axis. 128 00:24:15.400 --> 00:24:33.750 David J McLean: and the first step of operation for all hall effects is that a thermionic cathode is heated up until it starts to emit electrons. The next step is to apply our power supply members and electric propulsion system, relying on external power to generate an electric field from an upstream ano to a downstream cathode. 129 00:24:33.910 --> 00:24:38.160 David J McLean: When this is generated it starts to pull electrons into the discharge channel. 130 00:24:38.200 --> 00:24:43.360 David J McLean: and at this point we Haven't made a very effective fault, but we've made it a very effective electrical circuit. 131 00:24:44.060 --> 00:24:55.480 David J McLean: Now the key enabling feature for Halifax thrusters is to to apply another fundamental field, a magnetic field which is transverse to that applied electric field in the radial direction. This canonical architecture. 132 00:24:55.740 --> 00:25:12.930 David J McLean: this has 2 effects. The first one is that it starts to magnetize the electrons. When the electrons are susceptible to relatively strong mining fields, they start to undergo what's called Larmer procession, where they experience high frequency, circular drifts around the confined amending fields, and you can see one right here kind of making its way 133 00:25:12.940 --> 00:25:14.180 David J McLean: through the thruster. 134 00:25:14.620 --> 00:25:28.780 David J McLean: At the same time the combination of electric field Cross, with that many field gives rise to a cross field drift in the E Cross B direction, which in turn is in the as mutual direction of the thruster. So you can see the electron making its way 135 00:25:28.780 --> 00:25:41.900 David J McLean: around the channel here. If we are to envision in this architecture to come out of and back into the page. This strong electron drift in the as mutual direction is called the Halifax Drift, and once the thruster drives its name 136 00:25:43.340 --> 00:26:04.050 David J McLean: all right. So these electrons are effectively trapped in the strong electron drift, going a 100 or 200 kilometers per second. We next flow neutral gas from an upstream manifold into the discharge channel. The initial gas in this case is the note of the Xenon. That neutral gas now has to traverse the gauntlet. A very strong electrons which are traversing in the as you. As the direction 137 00:26:05.230 --> 00:26:20.910 David J McLean: this got little electron effectively acts like a buzz solid ionization. So as those heavy particles traverse that trapped region they are run into by the electrons that kicks off additional electrons, which ionizes the plas, the gas into a plasma state characterized by eco parts, positive and negative particles. 138 00:26:21.470 --> 00:26:40.050 David J McLean: Now a central design principle of hall thrusters is that the magnetic field is tailored in such a way that while the electrons are magnetized, the ions are sufficiently heavy that they are not magnetized, and therefore only feel the electric of field which is applied to them. As a consequence, those ions are accelerate out the geometry at very high speed by the action, the electric field. 139 00:26:40.370 --> 00:26:51.390 David J McLean: This is an enabling feature of hall frustr, in fact, because the degree to acceleration depends on the electric potential that is applied. As you increase that electric potential, you can get the exhaust losses to become progressively higher 140 00:26:52.260 --> 00:27:07.240 David J McLean: at the end of the day. Once we've expelled these heavier species. If we took no action, then the fruster itself would start to charge negative. So this cathode or electron source serves the dual role of injecting additional electrons downstream to neutralize the overall plume. 141 00:27:07.290 --> 00:27:07.880 David J McLean: And 142 00:27:07.990 --> 00:27:20.520 David J McLean: so a key enabling feature, the acceleration process of hall effect thrusters fundamentally is that this acceleration process is so-called, causing neutral and therefore it can generate very strong electric fields and currents for a relatively small footprint. 143 00:27:20.900 --> 00:27:38.170 David J McLean: So to give a sense for kind of typical operating conditions, we most trust us on orbit run Xenon or Krypton. The exhaust velocities can be as high as 30 kilometers a second. Sometimes i'll use that interchangeably with specific impulse, which is just another way of saying, Normalize exhaust velocity. Take your exhaust loss and divide by 10 144 00:27:38.170 --> 00:27:40.930 cross levels anywhere up to half a Newton 145 00:27:40.940 --> 00:27:57.320 David J McLean: for us. Efficiency, which is an indication of the conversion of the power supplied into directed kinetic energy is somewhere around 67. And then finally, this is a key metric. We're going to talk about later which is they can generate about 10 Newtons per meter squared, which again is not much right home home about about a kilogram of fast 146 00:27:57.380 --> 00:27:58.170 David J McLean: right 147 00:27:58.180 --> 00:28:13.960 David J McLean: with all that being said, despite the fact that they have relatively low thrust. These kind of key performance. Metrics have translated to hol attack thrusters being widely leveraged for geocentric applications, and indeed they have been baseline for the upcoming deep space gateway, which is supposed to be the space station that goes around the moon. 148 00:28:16.030 --> 00:28:27.800 David J McLean: All right. So with this under our belt to kind of a fundamental understanding of how hall effect thrusters work. What's next access Question: Why, fundamentally has thrust density been thought to be limited in this this technology. 149 00:28:28.490 --> 00:28:42.730 David J McLean: So to do that, let's first pose the question. How would you actually increase the thrust density of a hall effect thruster? And to illustrate that i'm going to show here on the left hand side kind of a nominal operating hall effect thruster, which has moderate the low thrust density. 150 00:28:42.730 --> 00:28:50.290 In this case it's characterized by relatively low flow rates or number of particles and neutrals were injected and converted to ions and accelerated downstream. 151 00:28:50.880 --> 00:29:09.790 David J McLean: The central idea here is that one easy knob for increasing the thrust density. The amount of force generated per unit area of the thruster is simply to turn up the flow rate. So we push more particles through the system which subsequently get a good ionize. More particles are accelerated. The same speed that generates more reactionary force or in simple approach. 152 00:29:09.790 --> 00:29:32.550 David J McLean: And indeed, we can see that on the right hand side this technology, which has more flow rate to the thruster commensurately, we would anticipate, should have a higher thrust per unit area. Now, moving forward, I'm. Actually going to use mass flow rate, as shown here. Kilograms per second interchangeably with this metric, which is called discharge current. How much current is supply to the thruster to generate the ions which are accelerated downstream. 153 00:29:33.460 --> 00:29:57.110 David J McLean: Okay. So in principle, you could increase the thrust density of a hall effect thruster just by turning up the flow rate or amount of current that's delivered to the system. But there are some t technical challenges which have to date fundamentally limited our keep capacity to do that. So i'm going to talk about this at the high level. The first one, of course, is thermal. So if I just put more particles in here, and these particles are getting energized, and the electrons to be millions of Kelvin. 154 00:29:57.110 --> 00:30:12.100 David J McLean: those electrons can diffuse to the walls, and they can heat up the thruster. And if you're not carefully, you can so get so much power to the walls that can actually overheat the thruster and melted right. So there's a key thermal limitation right off the back, for those of you are familiar with chemical rockets. Thermal limitations are a big concern as well. 155 00:30:12.790 --> 00:30:35.440 David J McLean: Another key limitation is lifetime, so hall effect thrusters can, in fact, a road. If you're going to accelerate Ions at very, very high speed. Some of those ions, if you're not careful, can go on the wrong direction, and, in fact, they might be able to impact the sides of your thruster. If they hit the material with sufficient energy, they can kick some additional material off in a sputtering process, which can also eat away at the walls of your thrust or lead into overall failure. 156 00:30:35.440 --> 00:30:41.220 David J McLean: I mean, this problem becomes more pernicious as the current and the thrust density increases 157 00:30:41.770 --> 00:31:00.760 David J McLean: finally, and this is kind of the key one we're going to talk about is loss and plasma confinement now at a high level effectively. What is going on here, we think, is that you introduce more particles, and that compromises the electron drift which in turn leads to electrons being able to traverse across the confining field lines that also can lead to a performance drop. 158 00:31:01.010 --> 00:31:06.470 David J McLean: I'm going to talk about that in more detail subsequently. So let's reserve our discussion for that for now. 159 00:31:06.640 --> 00:31:24.240 David J McLean: And, indeed, what I want to highlight here is that these 2 2 key challenges are effectively engineering challenges, and we think there are potential solutions, and we have some ongoing projects to address them. We think we can mitigate the thermal issues, and there have been some recent developments in hall thruster design, which we think will mitigate the lifetime problem. 160 00:31:24.480 --> 00:31:41.510 David J McLean: But as I just alluded to this other problem in terms of plasma, confinement is one that is not well understood to a certain extent, and could also be the fundamental limiter, and has been the fundamental limiter, at least a phenomenological level. For why hall thrusters have been limited and thrust density. 161 00:31:42.100 --> 00:31:59.900 David J McLean: So now I want to talk about why we think of that. Loss of confinement occurs in higher thrust density, and why it adversely Impacts performance. So to illustrate that again. Here's just a cross-sectional view of my channel as shown here and we're going to introduce a little math to talk about the fundamental operation. 162 00:32:00.160 --> 00:32:10.360 David J McLean: So, broadly speaking, this is this parameter. Again, electrical efficiency which is characterized the amount of thrust power generated by the system compared to the amount of electrical power put into the system. 163 00:32:10.360 --> 00:32:29.810 David J McLean: the higher the electrical efficiency the better. And if I have a one megawatt reactor power in my system and my electric efficiency is only 10. That means only a 100 kilowatts of my power is going to thrust generation, and I have to somehow deal with that other 900 kilowatts all right through thermal rejection. So electrical efficiency is actually a key parameter for characterizing the operation of a system 164 00:32:30.710 --> 00:32:41.960 David J McLean: in turn the electrical efficiency for a thruster. It can be broken down into many different modes. But i'm going to highlight the 2 key ones which were related to plasma confinement one is which is called the so called mass utilization, efficiency. 165 00:32:42.540 --> 00:32:50.490 David J McLean: which in words represents the fraction of inflowing gas which is converted to a plasma state. The more conversion you have, the better. 166 00:32:50.690 --> 00:33:02.870 David J McLean: and the second one, which is called the beam utilization efficiency, which is an indication of however much current you're putting into the thruster. How much of that current is actually being converted into accelerated current, or useful concern for thrust production? 167 00:33:03.860 --> 00:33:13.830 David J McLean: All right. So a few key parameters to introduce one is that discharge current. This is furnished by the power supply. One is the characteristic length of the thruster owl. 168 00:33:14.860 --> 00:33:32.260 David J McLean: One is the characteristic electron confinement time. So how long electrons actually exist in their hall effect drifts before they are driven out of the geometry. And then there's that cyclotron frequency again, which is the rate in which those electrons actually go around those magnetic field lines. This the frequencies here can be on the order of 10 to the 9 169 00:33:32.390 --> 00:33:33.900 David J McLean: per second. 170 00:33:34.210 --> 00:33:41.800 David J McLean: and then a couple of other free parameters, which represents the ionization cross-section of the gas a geometric parameter, and the I and Larmer radius. 171 00:33:42.350 --> 00:33:53.410 David J McLean: all right. So, using some simple 0 dimensional scaling laws, we can come up a few analytical prescriptions for the mass utilization, efficiency. as well as the beam utilization efficiency. 172 00:33:53.680 --> 00:34:02.640 David J McLean: I don't want to belabor the underlying expressions too much, except to highlight this, which is that both of these fundamentally defend on what I've called the electron confinement. 173 00:34:02.640 --> 00:34:14.050 David J McLean: and we can see mathematically that is the electron confinement. Time increases the beam utilization will increase and If the electron confinement time increases the so-called Mass conversion. Efficiency will also increase 174 00:34:14.389 --> 00:34:34.710 David J McLean: All right. So i'm going to keep these expressions, in on the the slides, just to highlight the fact that we're evaluating the mathematically. But I want to talk more qualitatively about what they represent. All right. So let's call our code this, and we're going to talk about the overall efficiency. This shouldn't be a to end. This should be overall efficiency, which is a product of mass utilization as well as beam utilization. 175 00:34:34.770 --> 00:34:40.750 David J McLean: And the first thing we're going to plot parametrically is these 2 efficiency modes as a function of the confinement time. 176 00:34:40.850 --> 00:34:51.620 David J McLean: I say, normalize confinement time because it's a cyclotron frequency times the confinement time. So how long the electrons stay in the thruster for those of you with a background of plasma physics. This translates to the so called. Hall Parameter. 177 00:34:52.280 --> 00:35:05.080 David J McLean: All right. So mass utilization. If the confinement time increases for electrons as electrons are sticking around longer, they have a higher probability of ionizing gas, and therefore will translate to more efficient conversion of our neutral gas into a plasma state. 178 00:35:05.960 --> 00:35:17.950 David J McLean: Similarly, if these electrons stick around longer, few of those electrons are actually contributing to the total amount of current required by this of power supply, and therefore the so called beam Utilization increases monotonically as well. 179 00:35:18.370 --> 00:35:31.770 David J McLean: and in turn the overall frost efficiency. Your electrical efficiency will also increase monotonically as a function of confinement time. So the key take away from this intuitively, is that the hall effect thruster which relies on the confinement of electrons and drifts 180 00:35:31.770 --> 00:35:49.770 David J McLean: there. This efficiency in turn depends on the ability to confine those electrons for longer durations of time to translate the higher levels of performance. Okay, and the we can kind of show here a typical hall parameter for all of our thrusters, and we see this matches up to what I showed earlier in terms of 60 to 70 electrical efficiency. 181 00:35:50.810 --> 00:35:52.020 David J McLean: All right. 182 00:35:52.260 --> 00:36:04.520 David J McLean: Now, here's the rub, which is in this analysis. So far I've assumed that the confinement time is largely independent of the thrust density. How much current i'm pushing through the thruster. But in practice. This may not be the case. 183 00:36:05.180 --> 00:36:14.830 David J McLean: and to highlight that in a kind of high level or a qualitative level, let's again look parametrically. If the operation of the thruster as a function of the amount of neutral gas, and I putting into it 184 00:36:14.830 --> 00:36:37.790 David J McLean: so on the left hand side, I kind of have my so called, low. Thrust density, case, or load discharge current case, where I just have a few neutrals which are being transited and converted into a plasma state. Once those electrons are formed, they have a very low probability of running into another particle. There, there, just aren't enough there in that magnetic field region. As a consequence, they're going to stick on their qualified drives for very, very long duration times. 185 00:36:38.610 --> 00:36:50.780 David J McLean: Now, if I start to increase the number of particles in the channel. I can start to create more plasma in that strong mining field. And now a few of those electrons can actually run into those heavier particles, and they'll start down to go a random walk 186 00:36:50.780 --> 00:37:09.550 David J McLean: where they are. Random walk through through a collisional process across the magnetic field, traversing them at any field. And now those electrons contribute to the overall current. This current is parasitic to the operation of the thruster. We actually don't care about electrons traversing many field lines. We don't want any to do that. We want all the current we're supplying to the power supply to go into directed kinetic energy. 187 00:37:10.290 --> 00:37:36.450 David J McLean: Right now in the extreme case. If we have very, very high thrust density, a lot of mass flow rate going through the system. Now, we just have this: a huge number of particles, density, particles in the many field region, and the electrons all of them start to collide, and they kind of plank go their way here across to the v anode, and in this case we have virtually no electron confinement. All the power for the thruster is converted to electron current, and as a consequence of the beam Utilization goes down 188 00:37:36.830 --> 00:37:45.320 David J McLean: all right. So this is kind of a qualitative description; for why we'd expect that the performance of the thruster should decrease monotonically as the current density increases. 189 00:37:46.360 --> 00:37:58.390 David J McLean: So to kind of illustrate that mathematically we would expect, based off this classical collisional process, that the confinement time should inversely depend on the mass flow rate, or by extension, the amount of current that we're sending through the Thruster 190 00:37:59.020 --> 00:38:16.420 David J McLean: and the CD. Implications of that. Now let's revisit this parametric plot, where i'm showing efficiency modes on the Y access and discharge current on the x-axis, and for comparison a typical hall thrust or discharge current. So many of the technologies that have been developed and qualified today operated about 15 to 10 discharge currents in terms of amps. 191 00:38:17.010 --> 00:38:27.280 David J McLean: So first we'll denote symbolically that the confinement time now depends on discharge currents. and we're going to plot parametrically how these efficiency modes vary 192 00:38:27.480 --> 00:38:45.150 David J McLean: well. The mass utilization in terms of its trends is not change, and this is intuitive, because if I put more charged particles in my thruster, those neutral particles which traverse from the upstream region are just going to have to hit them eventually. As a consequence, I'm just going to have a higher likelihood of ionization, no matter what the more particles I put in the system. 193 00:38:45.660 --> 00:38:50.500 David J McLean: On the other hand, as we had intuited earlier from that phenomenological description. 194 00:38:50.820 --> 00:38:55.570 David J McLean: as the discharge current increases, we'd expect that classically the beam utilization should go down. 195 00:38:56.680 --> 00:39:14.000 David J McLean: which is a which is to say that the confinement of the plasma should be adversely impacted, which should translate to a loss in that efficiency mode. and if we ultimately take the product of these 2, we into it that there should be a maximum in terms of discharge current as an extension in terms of current density which a hall effect thruster should operate. 196 00:39:14.510 --> 00:39:30.710 David J McLean: It's in large part for this region, though there's some debate as to why are also many scaling laws came from that hall effect. Thrusters have been relegated to this relatively moderate to low thrust density, operating regime. The conventional wisdom is that you can't go much above 100 millionper centimeters squared before performance starts to fall off. 197 00:39:31.630 --> 00:39:47.150 David J McLean: So to look at that in parametric space I can show this plot, which is a function of demonstrated thruster power going all the way up to 100 kilowatts as well as thrust density for different propulsion systems, and we see here that electro sprays exist in very low thrust densities. 198 00:39:47.150 --> 00:40:06.860 David J McLean: I am thrusters which are a different type of acceleration. Mechanism existed moderate the low thrust densities. Mpds, which are kind of the kings so far exist at these via high thrustencies, which again translates the ability to generate more thrust for small footprint and hall effect for users have been relegated to this space, which is, it can generate about 10 Newtons per Meter square. 199 00:40:07.110 --> 00:40:20.270 David J McLean: Okay, so the fundamental reason why this limitation exists, and most hall thrusters are built. Following the scaling law is that it's been thought that if you go above that current density or thrust density, the overall performance of the thruster will suffer. 200 00:40:22.580 --> 00:40:31.300 David J McLean: and we can now revisit this plot and understand where it comes from, which is to say that if we try to scale a thruster up using the conventional scaling laws, it gets prohibitively large. 201 00:40:32.230 --> 00:40:44.650 David J McLean: all right. So with that in mind and kind of accepting that conventional wisdom which has been handed down to us for the past couple of decades. Why did we naively, potentially think that we'd be able to overcome it. 202 00:40:44.680 --> 00:40:57.940 David J McLean: What? Why do we think this limitation might not be absolute? So let me revisit this fundamental description here, and everyone's on the same page here in terms of more neutral gas density flowing in the system translates to lower 203 00:40:58.140 --> 00:40:59.720 David J McLean: confinement. Right? 204 00:41:00.020 --> 00:41:04.690 David J McLean: Everyone's on their heads. Okay? Okay? Well, this none of this is a lot of bull. 205 00:41:04.960 --> 00:41:10.000 David J McLean: all right. This description gets the electron dynamics all wrong, right? And in fact. 206 00:41:10.450 --> 00:41:16.580 David J McLean: this description is a so-called classical approach in which the electron confinement is dictated by collisional effects. 207 00:41:16.630 --> 00:41:23.790 David J McLean: but in reality. And this is a common truth. Across all disciplines of plasma physics the transport is much more non-classical 208 00:41:25.150 --> 00:41:26.090 David J McLean: right. 209 00:41:27.240 --> 00:41:35.450 David J McLean: and fundamentally in many ways, for healthcare. We don't understand what causes the electrons to move across field lines. This is an active area of research. 210 00:41:35.550 --> 00:41:45.320 David J McLean: and that to very research that's gone on. For 40 plus years now there' been a lot of breakthroughs recently. People tend to think, at least in hall frustrations that this non classical transport is attributed 211 00:41:45.330 --> 00:41:59.460 David J McLean: to the existence of turbulence, and we have beat our heads against it here in Michigan, as well as other centers for many, many years. So here's this proof of that. And in the many ways we keep trying to solve and unsuccessfully solve the problem all right. 212 00:41:59.460 --> 00:42:11.400 David J McLean: But fundamentally the key takeaway from this is that that electron transport is not classical, and if it's not classical that invites an interesting question, maybe this kind of fall off in performance is not actually absolute as we go to stronger currents. 213 00:42:12.010 --> 00:42:22.310 David J McLean: Okay, so to kind of illustrate this one common shortcut, and i'll say this is definitely not right to do, but a common shortcut that's applied in all thresholds is to assume that the transport is so called bone like. 214 00:42:22.410 --> 00:42:31.670 David J McLean: Now I won't go into the details of that. But effectively. What this translates to is the assumption that the confinement time does not depend on discharge, kind of flow rate, but is actually independent of that. 215 00:42:31.950 --> 00:42:44.000 David J McLean: So if we once again revisit our phenomenological breakdown in terms of overall thruster performance Again, the mass utilization efficiency increases monotonically in asymptotes, and we're converting all of the neutral gas into plasma. 216 00:42:44.090 --> 00:42:50.200 David J McLean: But the beam utilization in principle should remain constant right. If the confinement is independent of discharge current. 217 00:42:50.310 --> 00:42:56.670 and the implication would be that we could actually operate a much higher current densities within no drop off and thrust efficiency. 218 00:42:56.770 --> 00:43:12.620 David J McLean: So this is kind of an enticing idea. And indeed, if you take that to its logical inclusion, as these authors did in this study, which was published a couple of months ago, you come to the conclusion that hall effect Thrusters are for 2 to 3 orders of manage to lower in terms of thrust density compared to where they could be 219 00:43:14.550 --> 00:43:28.230 David J McLean: now in practice. If we actually look at those 2 limits, the bone like, and the classical diffusion in terms of efficiency. The truth, we think, probably lies somewhere in between those 2. And, in fact, that's kind of the hypotheses hypothesis and motivated this work. 220 00:43:28.960 --> 00:43:35.740 and the central idea was maybe maybe if the transport is non-classical somewhere between bomb like and classical. 221 00:43:36.090 --> 00:43:42.970 David J McLean: if you go to higher current densities the drop off, and performance might be acceptable. maybe the drop off and performance doesn't even occur right? 222 00:43:42.990 --> 00:43:47.880 So this is kind of the key question that we started with. If we push the higher current and see what actually happens. 223 00:43:48.190 --> 00:43:55.950 David J McLean: So that might mean that the press density limit actually lies between these 2 regions. But it's still a 2 order of magnitude boost, and that could be substantial in terms of application. 224 00:43:56.630 --> 00:44:09.040 David J McLean: So how do we go about addressing this question? Well, of course you could try to model it and try to come up and improve fundamental understanding of what we think from first principles. That limit should be, or you could turn it up to 11. 225 00:44:09.550 --> 00:44:23.110 David J McLean: Now we end. Sue said this in an article that was published on this I'm. Going to give her a credit for that. Okay, based on the reaction. No one seen spinal tap. All right. What's the Max? You can go to an amplifier for your guitar 226 00:44:25.610 --> 00:44:26.490 David J McLean: right? 227 00:44:26.650 --> 00:44:33.180 David J McLean: 1010, 10 notionally, unless your spinal tap, which is, you can turn it up to 11, right? So you can make it even louder. 228 00:44:34.000 --> 00:44:50.410 David J McLean: Alright, so instead of modeling, we just went for it, and effectively. We we baseline this technology which the 4 and a half kilowatt thruster that we co-developed with NASA and the Air Force Research Laboratory and we did a bunch of modifications to it, such that we were comfortable that it could operate at these higher power densities. 229 00:44:50.410 --> 00:45:08.790 David J McLean: Now I don't want to belittle this. It took 2 years of iteration to get to the point that we could comfortably test this, and I think it's in part, for that reason, that people have not at least openly published on the ability to do this earlier. But I won't. Go into the details in terms of the modifications where, suffice is to say that we were able to push the current density to these high power levels. 230 00:45:08.930 --> 00:45:14.900 and our overarching goal was to see ultimately how the performance responded as we go went to those higher current densities. 231 00:45:15.240 --> 00:45:30.500 David J McLean: Now all this testing was done in a large vacuum test facility here at the University of Michigan. We're about a mile off campus. Here's a chamber for reference dates back to the Apollo days, 1,961. Here's a sense for scale. This is the entire lab group inside. And there, right, there is a whole thruster thing. 232 00:45:30.500 --> 00:45:38.950 David J McLean: So you need this kind of scale to better approximate a vacuum like environment, but also to accommodate the flow rates and powers that the thruster can generate. 233 00:45:39.410 --> 00:45:46.440 David J McLean: And in turn we applied a series of diagnostics which allowed us to characterize experimentally these multiple efficiency modes. 234 00:45:46.960 --> 00:45:49.290 David J McLean: So here's a thruster 235 00:45:49.540 --> 00:46:08.080 David J McLean: firing and we get a video sweep through the plume as it's operating. And while it's doing that i'll just say a couple of details. So the discharge voltage we applied again. That's the electrical potential of starting the ions is 300 volts. Not only that translates to about 2,000 to second specific impulse for 20 kilometers per second. 236 00:46:08.080 --> 00:46:19.320 We did this test, both on Xenon and Krypton, and we turned the current all the way up from its nominal current at 4 and a half kilowatts, 15 amps all the way up to 150, am so trying to demonstrate a 10 X in terms of current density. 237 00:46:21.790 --> 00:46:31.480 David J McLean: All right, so let's get to the results here on the left hand side. I'm going to show my qualitative expectation based on those arguments I made earlier, and on the right hand side. I'm going to show the measurement. 238 00:46:31.700 --> 00:46:47.620 David J McLean: So this is a mass utilization which we expect to monotonically increase in asymptote as a function of discharge current. And indeed, we saw that reflected for the hall effect thruster. Now a couple of these mass utilization goes above 100. We're not creating mass. This is probably something to do with the probes right. 239 00:46:47.620 --> 00:46:57.450 David J McLean: But suffice to to say that we thought the qualitative trends were an indication that we're getting to about 100% mass utilization, efficiency. So these are measurements. This is a comparison to qualitative expectation. 240 00:46:58.490 --> 00:47:11.630 David J McLean: All right. Now, here's the real test. This is beam utilization. If the transport or the confinement time is dictated by that bone limit. and Here is the expectation of it's dictated by those classical collisions. 241 00:47:12.470 --> 00:47:14.510 David J McLean: And here's what we found right. 242 00:47:14.540 --> 00:47:26.420 David J McLean: So the beam utilization, efficiency, the confinement of the ion electrons does, in fact, decrease as you go to tired discharge currents, but it's only a moderate decrease. It goes from 81 down to about 70, 243 00:47:26.710 --> 00:47:31.490 David J McLean: and whereas the classical case case, we'd expect a much larger reduction. 244 00:47:34.140 --> 00:47:43.570 David J McLean: And indeed, if we compare the total efficiency which is the project of mass utilization to that beam utilization, we find a result that looks something like this. 245 00:47:43.620 --> 00:47:53.400 David J McLean: So we can now understand from that previous discussion where this optimum comes from, it's a trade off between mass utilization and decreasing beam, utilization, or confinement of the plasma. 246 00:47:53.400 --> 00:48:10.770 David J McLean: But more encouragingly and very exciting result is that we see that the overall drop in thruster performance. It went from its nominal value here about 57 down to about 53% right this is a very small drop in performance, given the fact that we've achieved effectively a 10 X increase in current density. 247 00:48:11.120 --> 00:48:29.480 David J McLean: All right. So this is the key finding, and, in fact, the key finding of the overall work, which is that we in effect, we're able to increase the current density of the system with minimal impact in terms of overall performance. And this goes to show that, in fact, the confinement, as we hypothesize, should exist somewhere between these 2 limits. 248 00:48:29.480 --> 00:48:35.380 David J McLean: and, in fact, the reduction in confinement was sufficiently moderate that this is still an attractive thruster at these current densities. 249 00:48:35.980 --> 00:48:42.150 David J McLean: So yeah, let me just highlight it again. 10 X increase in current density with only 5% decrease in performance. 250 00:48:42.500 --> 00:48:51.350 David J McLean: So let me conclude here with kind of what the implications of this finding are so first in terms of application. Let's return to this graph. 251 00:48:51.510 --> 00:48:59.330 David J McLean: This is where some people think if you follow bone scaling hall frustr's could exist in terms of thrust density. This is where they existed before. 252 00:48:59.390 --> 00:49:07.820 David J McLean: and this is our new data point. So we, in effect have a cheat, and, in fact, in effect, have achieved almost an order of magnitude, increase in thrust density. 253 00:49:08.180 --> 00:49:16.520 David J McLean: So, returning again to this image, where we saw these hall thrusters are getting bigger than a person. Of course we'll have to have 10 on the spacecraft to get to our 4 megawatt level. 254 00:49:16.660 --> 00:49:29.700 David J McLean: We, in fact, have demonstrated that in principle it's possible to achieve comparable and even better performance than the men need to plows dynamic thruster at moderate power levels that could potentially scale to in Mars architecture when you marry, that 255 00:49:29.700 --> 00:49:39.100 David J McLean: with the idea that hall effect thrusters at least in low power, have extensive flight heritage. This technology starts to become competitive, and perhaps even leading contender for a Mars architecture. 256 00:49:39.390 --> 00:49:48.110 David J McLean: Right? So this is a key application, finding What about our understanding of the physics? Well, it's hard to kind of extract 257 00:49:48.160 --> 00:50:06.510 David J McLean: concrete physical insight just from these kind of global performance metrics. But one thing that we do see is that somewhere the confinement lies between this kind of bone limit and classical collision limit, and that in turn might give us some new insight into what the fundamental physics are that are driving the effective electron dynamics in the system. 258 00:50:06.940 --> 00:50:22.510 David J McLean: for example, it has been proposed in previous models. That the confinement time should depend on the inverse square of density, which would be the inverse square of discharge current in this case is illustrated by the plasma frequency, and the results kind of loosely seem to suggest that this should be the case. 259 00:50:22.660 --> 00:50:37.160 David J McLean: Now that isn't to say that we could solve the transfer problem again just from these global performance metrics. But what I will say is that this is: get another clue or breadcrum in our ongoing investigation to try to understand and self consistently predict what's governing the electron dynamics in these systems. 260 00:50:38.100 --> 00:50:38.820 David J McLean: All right. 261 00:50:38.930 --> 00:50:56.690 David J McLean: Last comment is, what are some extensibilities that have immersed in these results, and I'll just touch on this briefly. But one thing that emerges from this mass utilization curve is that again we see that the discharge current increases. The ability to convert plaque neutral gas into plasma improves in terms of efficacy. 262 00:50:57.200 --> 00:51:14.800 David J McLean: A key factor which we had assumed. Wisconsin was this thing which is the cross section for ionization. How easy is it for electrons to hit those those neutral gas particles? Historically, Xenon has kind of been the dominant choice, and the reason why is, you know, just a fat particle. It's very easy to hit it and convert it into a plasma. 263 00:51:15.250 --> 00:51:20.860 David J McLean: However, Xenon is expensive. The price has gone up substantially because of the war in Ukraine 264 00:51:20.860 --> 00:51:38.100 it is increasingly scarce as a consequence, and to a certain extent it's difficult to score store. So there's a lot of interest, both from D space architectures as well as commercial architectures, to explore the possibility of working in hall thrusters on alternative propellants things like water. It easily stores air, krypton, or nitrogen. 265 00:51:38.380 --> 00:51:49.130 David J McLean: The challenge is that all of these alternative propellants have much lower cross sections for ionization, and indeed, if I plot the mass utilization, utilization, efficiency as a function of disc charge current 266 00:51:49.130 --> 00:52:01.980 David J McLean: at that kind of nominal 15 amp operating condition, we see the performance of thrusters on things like nitrogen drops precipitously, as we'd expect theoretically, and that stems from the fact that it's just hard to hit those neutral particles because they're so small. 267 00:52:02.460 --> 00:52:17.080 David J McLean: However, if you operate the thruster at these much higher current densities, it kind of becomes a wash, and the thruster translates from a hall effect. Thrustery, lean, mean Ferrari to a garbage disposal, which is, it can take in all the gas and basically just conver it into plasma. 268 00:52:17.370 --> 00:52:34.420 David J McLean: and to kind of do as a proof of concept. There we contrasted Xen on to Krypton and measured the overall efficiency, and we see that well, Krypton's efficiency is fundamentally lower than Xenon and its maximum point. In fact, as we got to those higher current levels, the efficiency of being on Eclipse Krypton. 269 00:52:34.460 --> 00:52:54.280 David J McLean: and the reason Why, that is the case is because the mass utilization has a fundamental, different, fundamentally different dependence on the the discharge current. But this is all to say that this is very exciting from the perspective of being able to operate these sources long term propellants, and indeed, we have ongoing efforts to look at things beyond Krypton, including air, carbon, dioxide and water. 270 00:52:54.810 --> 00:53:11.010 David J McLean: All right. So here's my concluding remarks, which is, we've done all these interesting things in the past 2 years. Where do we go from here? Well, first is, can we turn it up? 20? Why, why stop at 11 again? Our understanding of the physics is quite limited here. 271 00:53:11.010 --> 00:53:16.250 But hopefully we can glean more inside as we continue to push the hall thruster to these these higher power densities. 272 00:53:16.680 --> 00:53:35.010 David J McLean: hand in hand with, that is always a systems level question. We've demonstrated a new toy, and we've shown that it works in the laboratory for 10 or 20 min. But how do you actually translate that to an operational system. And of course there are a lot of key challenges that need to be addressed both on the thruster side as well as the spacecraft side to make this a feasible technology. 273 00:53:35.010 --> 00:53:48.520 David J McLean: So at a high level, these are kind of the key challenges that we've identified, and we hope to go after in the next couple of years. First, this this is a dinky one. Power supply. We'll just get a bigger power supply. We only stopped at 150 amps, because that's how much power we had. 274 00:53:48.590 --> 00:54:03.880 David J McLean: I'm harnessing we had some issues going to higher volages. Magnetic field. This is really critical to how all thrusters work. We're trying to figure out ways to improve it. Thermal management. This is a big one. Here's a thruster after we turn it off after running at 45 kilowatts. Normally, that inside part is black. 275 00:54:03.970 --> 00:54:12.770 David J McLean: right? So we are definitely glowing and rejecting a lot of heat. So we have some ideas on how we'll channel that heat rejection. And this last one is, how do you test these systems 276 00:54:12.930 --> 00:54:19.120 David J McLean: right? So i'll just briefly say this, which is, despite the fact that our facility is much better than Georgia text 277 00:54:21.040 --> 00:54:41.640 David J McLean: and getting competitive to NASA Glen in case you're listening to Georgia tech. We have great facilities, too. It's still not a good representation of the space like environment. Okay, and that's because these chambers have walls which Don't exist in space, though the Chambers do a great job of pumping, there's still a 1 million times higher density compared to a space like environment. 278 00:54:41.700 --> 00:55:00.720 David J McLean: and it so happens that whenever you fire energetic particles at a wall, some of the material that wall is going to bounce right off, and, in fact, we line our facilities with graphite, and when you take the thruster out after you run it for a couple of 100 h, it's just caked in graphite. So really hard to do long duration tests where you're looking for failure modes if you keep adding material to your thruster right? 279 00:55:00.720 --> 00:55:11.080 David J McLean: So one of the key challenges is, how do you translate space like thruster environment testing on the ground to space. And of course, these problems just get worse as you go up in power. Right? 280 00:55:11.210 --> 00:55:27.330 David J McLean: So with that in mind. This is the institute that was mentioned at the beginning of the talk. I'm. A member of the Joint Propulsion Institute, which has the mandate from NASA to address this question. We're not supposed to do technology development. Under this we're supposed to answer the broader question how to even build the infrastructure to test these systems reliably. 281 00:55:28.430 --> 00:55:50.830 I'll also make a shout to the college here. I I think this is a forward looking. The college is committed to building a new facility which will allow us to build much higher power, thrusters and demonstrate them in higher fidelity environments than ever before. So it's not a glamorous shroud for it right now. We call it the garage, but we're going to put this new facility which we just picked up. It's getting a refurbished and should be coming to campus here in the next couple of months. 282 00:55:51.490 --> 00:56:02.760 David J McLean: All right. So summaries here for the overall tech. First, there are new frontiers and challenges in space. It's an exciting time to be in space exploration, and NASA is seriously considering going to Mars 283 00:56:02.760 --> 00:56:14.400 David J McLean: hand in hand with that fact that Space forces. Defense budget was actually larger than NASA's just in this last year, so the defense is getting pretty interested in Space 2, and they're looking, taking a very hard. Look at electrical ocean systems. 284 00:56:14.890 --> 00:56:27.620 David J McLean: With that. There is always going to be new needs for new capabilities and new mission architectures, and hand in hand. With that, we need the technologies to address those upcoming technical challenges. We need innovations both in the power system as well as the engines. 285 00:56:27.720 --> 00:56:37.860 David J McLean: I want to highlight here that although I spent a lot of time talking about hall thrusters, all of them are good candidates, and all of them should be investigated, and it's part of our responsibility as a university to help 286 00:56:37.870 --> 00:56:58.620 David J McLean: the people who hopefully formulate these missions come out the right answer which will work for their mission architecture Kind of the neat thing about all this, though, is that we ultimately had a physics based somewhat phenomenological hypothesis about how we might explore the limits of one of those Technologies, and we were also able to demonstrate it self consistently and to a certain explain extent explain those Measurements. 287 00:56:58.630 --> 00:57:16.450 David J McLean: And as a last point here. These findings are exciting, not only because of Mars architectures, but they open up new possibilities for operation on alternative repellents and other new mission architectures like that. But if there's one thing I wanted to leave you with in terms of impression for this at the end of the day. Sometimes you just have to turn it up to 11. 288 00:57:16.790 --> 00:57:31.300 David J McLean: And so, with that in mind, let me do a quick acknowledgment. This was the collaborative work of my entire research group. Everyone got together on Fridays, and they came up with different ideas. And even though many of these is not their thesis work, we're just excited about the project. 289 00:57:31.300 --> 00:57:43.250 it would not even remotely have been possible without all of their efforts. I specifically want to call out Lee and Sue, who will be graduating here shortly. Who is the lead of this effort, and wrangled all of us, including myself, and to giving these tests done 290 00:57:43.250 --> 00:57:59.980 David J McLean: advisors. We worked closely at Jpl, who gave us some technical insight into this, and finally Nate Janice to give us some funny for running the facility, and then also Michigan, which I had some internal phone in to allow us to try to push the one of this technology. So with that said, I am going to close, and I thank you for your attention 291 00:58:14.030 --> 00:58:15.060 David J McLean: for Ben. 292 00:58:20.680 --> 00:58:25.410 David J McLean: Any questions for yeah like 293 00:58:25.930 --> 00:58:26.710 that, you didn't 294 00:58:27.320 --> 00:58:33.300 David J McLean: fantastic work. I know you can get a 295 00:58:33.380 --> 00:58:36.320 David J McLean: The meeting was not in the field. 296 00:58:36.510 --> 00:58:40.540 David J McLean: It seems like, you know, the thinking was based on very fundamental. 297 00:58:40.750 --> 00:58:41.920 I you 298 00:58:42.350 --> 00:58:44.050 David J McLean: even those cubs that we showed 299 00:58:44.720 --> 00:58:49.620 David J McLean: i'm not the representative of what is happening right in the 300 00:58:50.060 --> 00:58:52.990 David J McLean: before you 20, we will let it right. 301 00:58:53.300 --> 00:58:57.640 David J McLean: So I was wondering what they 302 00:58:58.430 --> 00:59:01.290 David J McLean: made you think beyond that, so 303 00:59:02.180 --> 00:59:05.320 David J McLean: we all of us know that it's not an ideal device. 304 00:59:09.890 --> 00:59:12.820 David J McLean: Why was the community kind of 305 00:59:12.860 --> 00:59:14.420 David J McLean: Oh, by that, then? 306 00:59:16.510 --> 00:59:33.160 David J McLean: So why was the community bound by that? It depends on how pessimistic you are effectively 91 92, the first Soviet thrusters came to the us and they worked spectacularly well, and maybe there's some change in the design 307 00:59:33.280 --> 00:59:34.060 David J McLean: all right. 308 00:59:34.090 --> 00:59:36.900 Another potential reason is, there's been no need. 309 00:59:37.110 --> 00:59:46.010 David J McLean: There's no real need for this except for things like a Mars architecture. Or if you want to go to alternative propellants. So there just Hasn't been a need to actually push that limit. 310 00:59:46.350 --> 01:00:05.860 David J McLean: Now, in terms of what made me think about this. It was a student when I was teaching electr repulsion, and I said, hall thrusters tend to operate at this 100 millions per centimeter squared, and they raise their head and said, Why? Good question? And, to be sure, there are people who publish papers on it. But all those papers have been informed by this highly idealized theory. 311 01:00:05.940 --> 01:00:18.280 David J McLean: right? So I I am quite familiar with the fact that the idealized theory is far from actually describing the plasma. We still Haven't cracked this problem. So it seemed like a worthwhile thing to try just experimentally. 312 01:00:19.620 --> 01:00:23.530 David J McLean: Yeah. But but even But if you have a 313 01:00:23.610 --> 01:00:36.020 David J McLean: computation. right or something like the whole process that's going to show you a lot more than. and there were no clues there. Good, and 314 01:00:36.050 --> 01:00:45.470 David J McLean: the computational models are not self consistent. The transport has to be tuned. So you end up using these same substitutions that have been 315 01:00:45.490 --> 01:00:47.320 time scales all of those. 316 01:00:49.760 --> 01:00:56.200 David J McLean: the transport, the transfer ones, are the cli. The fact that collision frequencies are assumptions. Yeah. Yup. 317 01:00:58.250 --> 01:00:59.680 David J McLean: Yes, please. 318 01:01:00.440 --> 01:01:07.960 David J McLean: One of the what is the main limitation there? 319 01:01:09.410 --> 01:01:24.930 David J McLean: Fundamentally it's a saturation of the men in the course we use to shape the field structure kind of back of the envelope in the scaling papers. They say you can get up to about a 1,000 Gauss in practice. The many of the fields we can demonstrate and see they'll all thrusters. Go up to maybe a couple of 100 counts. 320 01:01:30.340 --> 01:01:37.640 David J McLean: Yes, but 321 01:01:37.960 --> 01:01:40.210 David J McLean: you know. There you go. But 322 01:01:41.440 --> 01:01:54.160 David J McLean: yeah, I I haven't really seen a trade study on that. But you do have quite a bit of margin right. You can even imagine a twox decrease in efficiency, and it's still translate to maybe 8 or 900 metric tons. So you're on parity with the chemical architecture, you know. 323 01:01:54.340 --> 01:02:00.000 but kind of the rule of time that they use in these studies. You start off with the assumption that you have 50% or higher electrical efficiency 324 01:02:02.860 --> 01:02:06.080 David J McLean: all those please so 325 01:02:07.180 --> 01:02:10.290 David J McLean: great great presentation. 326 01:02:10.800 --> 01:02:17.680 David J McLean: And i'm sorry the very beginning when you're inscribing the some point here. 327 01:02:18.950 --> 01:02:26.620 David J McLean: What is so in this wires architecture that to the 328 01:02:28.170 --> 01:02:32.640 David J McLean: yeah. Do you compare that with how many Lum? You don't have to ask for? 329 01:02:41.590 --> 01:02:42.930 David J McLean: How much is there 330 01:02:43.250 --> 01:02:48.150 David J McLean: it and how in that was things to bring to the 331 01:02:48.370 --> 01:02:51.520 information, and 332 01:02:53.180 --> 01:02:53.760 David J McLean: I 333 01:02:53.800 --> 01:02:57.740 maybe that's that's the number I was I was I was I was. 334 01:02:57.950 --> 01:03:06.630 David J McLean: And then how is that on the you have to do things in with the propulsion 335 01:03:08.020 --> 01:03:12.350 David J McLean: compared to the chemical, the more traditional chemical one. 336 01:03:12.360 --> 01:03:19.520 David J McLean: Where is the breaking point? Where how far you need to get in here in order to 337 01:03:20.270 --> 01:03:29.310 David J McLean: to basically a break. Even there. Think on that we like, I mean, I understand that you need some lounge and sorry for it to and all that, and you have figures and so forth. But 338 01:03:29.910 --> 01:03:32.750 David J McLean: did it take the access to 339 01:03:36.960 --> 01:03:40.540 David J McLean: you? Mean the trade space between electric propulsion and and chemical engines. 340 01:03:40.560 --> 01:03:45.680 David J McLean: Yeah. 341 01:03:45.720 --> 01:03:48.930 David J McLean: I don't believe you. You were trying to launch 342 01:03:49.250 --> 01:04:01.830 David J McLean: we right? And then from that point through that you have to have your each nuclear reactor 343 01:04:01.950 --> 01:04:13.230 David J McLean: that's going to be put in space. Then our decision. This is all on top of the table that we're trying to carry the box. So when you compare these side connections off 344 01:04:15.650 --> 01:04:26.720 David J McLean: from orbit to Mars versus doing the electric for the where, where is how how much do you have to get 345 01:04:26.920 --> 01:04:37.090 David J McLean: in this efficiency. Where? Where is your point? Here, there to basically break even with what? What is your objective function? What do you mean by break, even. 346 01:04:45.230 --> 01:04:55.600 David J McLean: Yeah. But they're different metrics you can use to evaluate Michigan Mission machine 347 01:04:55.870 --> 01:05:13.340 David J McLean: mission feasibility right? There's cost. There's reliability. There's transfer time. So all of those need to trade against each other. So let's take one cost right. I have not seen a definitive cost breakdown, but we can start with an initial point. Do you want to know how much sls cost to launch? 348 01:05:14.260 --> 01:05:15.870 David J McLean: How much? 349 01:05:18.040 --> 01:05:24.940 David J McLean: Okay, and how many did we need? And okay, 20 billiondollars? And if you use an ep architecture 350 01:05:25.210 --> 01:05:31.110 David J McLean: 4 billiondollars for for those launches. So 4 to 8 billiondollars. 351 01:05:31.480 --> 01:05:32.360 David J McLean: the great 352 01:05:35.770 --> 01:05:38.300 David J McLean: What is your so? 353 01:05:42.440 --> 01:05:43.280 David J McLean: Or 354 01:05:52.510 --> 01:05:56.060 David J McLean: you have to assemble in orbit, and then and you go to to Mars 355 01:05:57.060 --> 01:05:59.250 David J McLean: regardless what you do. 356 01:06:00.060 --> 01:06:06.910 David J McLean: You still want to have to take this there not people, not people. 357 01:06:07.380 --> 01:06:21.480 David J McLean: A 100 50 metric tons is the most you can send in the wheel. You want to do a Mars trajectory. That amount of drive mass you can do actual payload masking delivery substantially smaller. Right then you add on top of that. This is 100 metric tons of support it through of 6 358 01:06:21.480 --> 01:06:30.530 David J McLean: maybe support a crew of one. You could scale down that mass a little bit, but it would not be enough expendable to support a person, even one person. That's why you have to do multiple launches. 359 01:06:30.830 --> 01:06:41.870 David J McLean: There's no way right now with existing launch vehicles. You can send one vehicle to Mars from the direct launch. I I think it might be worthwhile to have this conversation offline. Yeah. 360 01:06:43.650 --> 01:06:44.410 David J McLean: yes. 361 01:06:46.080 --> 01:06:46.670 yeah. 362 01:06:48.120 --> 01:06:50.140 Okay. Okay. 363 01:06:50.370 --> 01:06:52.250 David J McLean: yes. 364 01:06:52.720 --> 01:07:11.320 David J McLean: Well, effectively, the air ambient air is too high a density to have good ionization. Among other things, if you wanted to actually process the amount of density of air that you have in ambient levels, the power acquire for the thruster. Be on the order, Gigawatts. Yeah, that's right. But in principle you could run a whole thruster head 365 01:07:11.320 --> 01:07:18.560 David J McLean: atmospheric level pressures. On the other hand, I think that that that case would be collisionally dominated for sure. So your electron confinement would really suffer. 366 01:07:18.670 --> 01:07:19.260 David J McLean: Yeah. 367 01:07:21.510 --> 01:07:24.960 David J McLean: yes. 368 01:07:25.240 --> 01:07:41.720 David J McLean: So this is looking at the different gases that we'll be using. For example, see how team look at you, or and depending on the current, like, what does it in terms, especially with like a hardware architecture compared to chemical rockets, the storage of that gas. That's like how large you would be. So, for example, like those in 369 01:07:42.740 --> 01:07:43.330 Okay. 370 01:07:44.550 --> 01:07:50.460 David J McLean: Is there a majority of that just spending up? Guess for sure 371 01:07:50.520 --> 01:07:53.730 David J McLean: in terms of storage storability? 372 01:07:54.040 --> 01:08:12.410 David J McLean: Xenon, you can store passively. Doesn't have to be cryogenically crude cool, which is in its favor. If you do a locks architecture, you have to cryogenically, cool, it's a soft project, so it's easier if you did a lock cell. H. 2 architecture very difficult. This is actually a problem that has not been solved yet. How do you store liquid hydrogen for the month long trance 7 month long transit to Mars 373 01:08:12.410 --> 01:08:15.720 David J McLean: Methane's easier because it's stores. It's soft. 374 01:08:17.890 --> 01:08:28.760 David J McLean: Yes, please. 375 01:08:28.899 --> 01:08:40.430 David J McLean: like you said you could have a that you did, for example, and I guess it all comes out to 376 01:08:40.779 --> 01:08:53.550 David J McLean: to get the same amount of No, and it all tries to specific impulse and nitrogen can actually achieve higher specific impulse. You need less mass 377 01:08:53.770 --> 01:09:06.200 David J McLean: with nitrogen, because nitrogen is a lighter molecule. and the amount of acceleration you can achieve goes by how much energy you put in, divided by mass. So you put the same amount of energy, electrostatic energy and divided by mass, nitrogen, and accelerate classroom. 378 01:09:07.569 --> 01:09:10.800 David J McLean: The nitrogen is not store as well. The tanks become bigger. 379 01:09:11.830 --> 01:09:16.689 David J McLean: Yes. yeah, a couple of these, like all of classical physics. 380 01:09:20.609 --> 01:09:29.689 David J McLean: So 381 01:09:37.479 --> 01:09:56.660 David J McLean: yeah, that's a great idea. And this is something. We looked at a couple of years ago, so I will say, and I think this is a fair assessment that our approach to closure and turbulence understanding in plasma is not as a I want to say, as well study, I'd say, maybe not as well characterized as 382 01:09:56.730 --> 01:10:06.490 David J McLean: classical fluid turbulence. In fact, if you look at some of the monographs that are written on that they start off by saying like, if you try to apply fluid-based turbulence closures to plasma-based, or you might as well throw up your hands and quit. 383 01:10:06.590 --> 01:10:31.240 David J McLean: So at the end of the day our actually approach to closure these transport coefficients is pretty pretty infantile compared to, I think, the more sophisticated works in in classical turbulence. We have tried analytical closures. We've called multiple equation closures, and we've tried data driven so we've used experiment to try to form data-driven models, and some of them work some of the not work so well, but not as works in a predictive way. 384 01:10:34.370 --> 01:10:40.410 David J McLean: You have a question. Please go ahead. I'm trying to see the what is the picture? Yeah, this would 385 01:10:41.530 --> 01:10:45.560 David J McLean: with me. And you want to answer that one. Okay 386 01:10:46.080 --> 01:10:48.300 David J McLean: it it 387 01:10:48.310 --> 01:10:50.850 David J McLean: through the H 9 muscle, which is 388 01:10:50.970 --> 01:11:03.410 David J McLean: so in cross phones. But it's also a 389 01:11:09.120 --> 01:11:10.290 but they don't do. 390 01:11:13.560 --> 01:11:17.150 David J McLean: Yeah. So it is the muscle car of frustr, and hence the name. 391 01:11:17.350 --> 01:11:29.340 David J McLean: Tony. There's a question from parentheses. What about nuclear propulsion? I I By that I think he means nuclear thermal. Yes, it's actually Shane. So we can enter the conversation. Nuclear thermal gets 900 s. Specific impulse 392 01:11:29.400 --> 01:11:33.710 David J McLean: has its own challenges and advantages. Nuclear thermal. You only have to 393 01:11:33.860 --> 01:11:41.480 David J McLean: holds the system for 5 or 10 min continuously. So you don't have to run it continuously like a lunch repulsion system. Thermal directions much easier for nuclear thermal. 394 01:11:41.780 --> 01:11:52.590 David J McLean: The other hand, you have to store liquid hydrogen for the duration of the mission. So there's a trade space there. No. yes, one more one last question. 395 01:11:54.780 --> 01:11:58.200 David J McLean: but it wasn't included on the chart. 396 01:12:08.200 --> 01:12:11.890 David J McLean: Yeah. 397 01:12:12.360 --> 01:12:13.530 David J McLean: that no. 398 01:12:13.660 --> 01:12:23.130 David J McLean: no real reason not to include it. It's efficiency, for organization is even worse. and and it does weird things like break apart, which takes more energy. 399 01:12:23.390 --> 01:12:34.570 David J McLean: Water also is challenging to in ways that those other proponents aren't because the oxygen that you get from water and associating can react with your thrust. Your body. 400 01:12:34.740 --> 01:12:42.440 David J McLean: and particularly the cathode. Those cafes are heated up to 15 or 1,600 degrees, Kelvin or C, and they start to oxidize like crazy. 401 01:12:42.490 --> 01:12:51.250 David J McLean: So this is a problem that has not been solved yet, and it's one of the major limitations, at least for higher power, that they don't run water thrusters. That being, said 402 01:12:51.320 --> 01:13:06.790 David J McLean: Professor Kazumi, who gave a talk here in the fall, they have this company that builds low power Ion thrusters that run off of water, and the way they're able to get away with that is that they use a cathode that doesn't actually have a thermodynamic emitting surface. But these do not scale well, the high power operation. 403 01:13:08.740 --> 01:13:10.370 David J McLean: Okay, One last question. 404 01:13:12.250 --> 01:13:16.380 David J McLean: I or. 405 01:13:24.100 --> 01:13:35.060 David J McLean: Yeah. Well, there's no secret. There it's. It all lies on this thing called Meedic Shielding, which is a way to shape the many field and the thruster such that you establish electric potential gradients that are directed away from the walls. 406 01:13:35.100 --> 01:13:43.490 David J McLean: This is an innovation that came out of a Aerojet rocket dine in NASA about 10 years ago, and we've leveraged it quite extensively in our 407 01:13:44.730 --> 01:13:50.640 David J McLean: Okay, great in view of time. Let's stand ben for a great talk. 408 01:14:13.350 --> 01:14:14.130 David J McLean: you know.