A huge number of satellites that orbit the Earth in near‑Earth orbits are equipped with so‑called SPTs — stationary plasma thrusters. The designs of these thrusters have not changed for many decades. But to improve them, we first need to develop a model of the physical processes that occur inside the engine. We are discussing how the work on developing such models is progressing and what improvements plasma engines need with Alexander Stepanenko, an associate professor at the Department of Plasma Physics at the National Research Nuclear University MEPhI, and a senior researcher at the Laboratory “Physico‑Chemical Processes in the Walls of Thermonuclear Facilities” at the National Research Nuclear University MEPhI.
- First of all, what is a stationary plasma engine?
- It is a type of electrojet engine that uses plasma and electric and magnetic fields to accelerate the plasma. In fact, it is a propulsion system designed for spacecraft, i.e., for operation in outer space. And the applications of stationary plasma engines are extensive. Currently, the engines developed by Alexei Ivanovich Morozov are mainly being installed; this dates back to the mid‑20th century, when their first prototypes actually appeared. They proved to be very successful, and they began to be installed on satellites on a permanent basis; these satellites are launched into space and used for a very wide range of tasks, but primarily for orbit correction. Although now there are projects for more powerful engines for interplanetary travel.
- Do these older‑design engines have any problems?
- The main issue related to engine operation is that it is necessary to maximize its operational resources. The current requirement for devices equipped with a stationary plasma engine is that the service life of a single engine should be at least 10,000 hours. At the same time, the engines can operate for hours, meaning we are talking about many hours of continuous operation.
- What primarily hinders the long‑term operation of an engine in space?
- The main problem that arises when operating engines of this type is the destruction of the walls of the accelerator channel. The feedstock for the plasma in such engines is usually xenon – an inert gas with a fairly high specific mass. When massive xenon ions are accelerated by an electric field with a potential difference of several hundred volts, they acquire sufficient energy to destroy the walls of this channel.
- What are the walls usually made of?
- Usually, it’s boron‑based ceramic, a boron‑nitrogen‑silicon composite. There’s a rather complex mixture there — it’s called borosil — and it’s a standard material used in modern SPT systems to make acceleration channels. And this material sputters over time. This sputtering is most dangerous precisely where the ions acquire the highest energy — and that’s precisely the engine nozzle. As the accelerator channel deteriorates, the elements that make up the magnetic system also break down. And the operation of this engine directly depends on the magnetic system. Therefore, a change in the topology of the magnetic field lines leads to a change in the acceleration characteristics of the substance, and the engine itself may simply switch to other operating modes. This is the main issue. There is also a range of more specific tasks that need to be addressed in relation to the problem of extending the service life of this engine — and overall, this range is also very extensive.
- What is being done to solve these problems?
- The first, and probably the main, task that research groups around the world are working on is to model the plasma acceleration processes directly in the channel itself as accurately as possible. Because, despite the fact that the plasma engine is called stationary, it turns out that there are no stationary processes in it, and very often the engine itself operates in what is known as an oscillatory mode, when plasma is constantly being ejected from it. Therefore, the plasma acceleration process is highly nonlinear. And today, there are virtually no analytical methods for studying nonlinear acceleration models. Therefore, the only tool available to researchers that allows them to actually peer inside the accelerator channel is the method of computer modeling. Researchers around the world are creating their own plasma acceleration models, and most of them are engaged in computer modeling of SPT. We are also engaged in this modeling, trying different levels of detail and different levels of accuracy, taking advantage of the fact that there is a range of models that can be used. And, of course, we develop our own models tailored to our specific tasks.
- What tasks are you trying to model in the first place?
- First, issues related to the formation of thrust, that is, how the plasma is redistributed and accelerated within the volume of the accelerator channel. Second, the stability of the discharge burning, the physics of the process of plasma interaction with the surface. And we are gradually moving from simpler plasma models to more complex ones. We are trying to build a higher‑dimensional model that will include a broader range of processes.
- What are the ways to solve the problem of wall degradation?
- The main approach is to change the geometry of the magnetic field lines, because the configuration of the electric fields formed in the engine nozzle depends on the magnetic configuration, and the configuration of the electric fields in turn determines where the ions go and how they ultimately end up on the wall. For example, by designing the magnetic system in the right way, it would be possible to ensure that the ions hit the wall as little as possible. However, the problem is that the magnetic field affects not only the nature of ion motion and their interaction with the wall, but also how the plasma discharge burns in the engine channel. Therefore, as soon as we change the magnetic system, we immediately change the characteristics of the discharge. And this, in turn, affects the thrust force, specific impulse, and other characteristics of the accelerator. Therefore, the task turns out to be multifactorial: we constantly have to keep two parameters on different scales. On the one hand, we want to make sure that the wall is less prone to sputtering, and on the other hand, we want the accelerator parameters to be exactly what we need. How to balance these two goals is the most important task.
- What allows you to change the geometry of the magnetic system?
- You can wind the coils in different ways and arrange them differently. You can use different numbers of ampere‑turns, different form factors for these coils, and different profiling. You can change the nozzles. For example, there are nozzles with magnetic shields or inserts made of soft magnetic iron, which, due to being a ferromagnet, affects the structure of the magnetic field lines. There are a lot of possibilities. Different research groups use different approaches. But the task is far from trivial, and, for example, there are still no complete three‑dimensional models of matter acceleration in an SPT channel. This is precisely what makes the task challenging. We have to “keep in mind” a whole host of physical processes. We must resolve them in time, in space, and in terms of the combination of interactions, and we must place all of this in a three‑dimensional configuration; it would also be desirable for this to be resolved in terms of velocities. In other words, we must store not the flow characteristics, but the distribution functions.
- In this work, do you set yourself the task of increasing engine power?
- This is the tip of the iceberg; this is the ultimate goal of our work. We want to bring our developments to the stage where they can be physically embodied in a specific technical product.
- As you worked on this project, did you have to develop special software to model all these processes?
- Our group is developing two sets of models. One set of models was initially developed in the research group of our senior colleague Konstantin Vladimirovich Brushlinsky, who, unfortunately, passed away quite recently. He spent his entire life modeling processes in various plasma devices, including SPT. The second group of models consists of those we build using freely available software. There is a wonderful library called BOUT++, which contains, for example, solutions to problems in thermonuclear energy; it has proven to be very useful for solving our problems as well. We have results for each of these areas.
- Are you modeling xenon plasma, or are you already considering other substances?
- For now, we are focusing on the standard working substance — xenon. Although it is known from the literature that attempts are currently being made to create SPT using other gases, including water — there are even such exotic configurations. Perhaps, over time, if we realize that we are constrained within the xenon setup, we will naturally switch to other working substances.
- Do you assume that these engines will continue to be used for orbit correction, or perhaps the focus will also shift to interplanetary flights?
- Projects are currently underway to develop particularly powerful solar electric propulsion systems, including those that are planned to be used for interplanetary travel.
- Will nuclear reactors be their power source?
- It all depends on the tasks. If a large power output is needed, then, naturally, only a nuclear power plant will do. Perhaps, over time, if we make a breakthrough in the field of thermonuclear energy, then a thermonuclear source may appear. But mostly, it’s some kind of compact energy source. It could be some kind of compact battery, for example, a nuclear battery, or it could be a power source from the Sun.
- How would you characterize the status of your research? Are you currently at the halfway point, and do you already see any intermediate solutions?
- The journey has no endpoint; there is only the process itself. Currently, we can characterize our findings as follows. We have taken an important step by moving from one‑dimensional models to two‑dimensional ones. One‑dimensional models, although they are the simplest models of engine operation, are nevertheless very informative, as they allow us to study the physics of the main processes occurring in engines using relatively few resources. The transition to two‑dimensional models is a certain intermediate step. The ultimate goal is to create a three‑dimensional model in which everything is taken into account. But, as always, there are limitations related to the fact that the larger the model becomes, the more difficult it is to coordinate the work of its individual parts. Therefore, I would say that we are somewhere at an intermediate stage.
