a key link in the Mars race?

After the successful test flight of IFT-5, the next breakthrough change from SpaceX is the fundamentally new Raptor 3 engines. This will not be a test unit for one flight, but the real basis of space expansion: they plan to produce 500 of them per year – more than all other engines for space rockets in all other countries of the world combined.

My name is Sasha Berezin, I am the author of the MTS Digital special projects team. Today I’ll tell you why the parameters of the new engines can provide a real revolution in space exploration. We will also discuss whether Elon Musk will be able to realize their potential.

We all know that the United States landed on the moon first because it received engines with the required absolute power. But their specific parameters were pretty mediocrewhich is why they were not used after the Lunar Race. And the fact that after it, humanity stopped flying to other celestial bodies for more than half a century is also partly to blame for insufficiently advanced rocket engines.

In some ways, this story repeats itself: the ability to return to the Moon, start flying there regularly, and then reach Mars directly depends on the rocket engine. This seems surprising: now is not the time of Korolev and von Braun, what new can be invented in such an area as rocket engines?

Why is Raptor 3 so shockingly different?

To understand this, you first need to compare. The most powerful rocket engine for super-heavy rockets that carried people into space is the F-1. The Saturns of the American lunar program took off on it. Fans of Russian cosmonautics will mention the RD-170 for Energia, but the calculations will not fundamentally change, and the Soviet super-heavy aircraft made only two flights (and both without people). F-1 thrust – 690 ton-force. For comparison: the Soviet “lunar” rocket engine NK-33, on which rockets flew into space even in the 2020s, had only 154 ton-force of thrust. It is not surprising that with just five F-1s in the first stage, the United States was able to launch a rocket into space to fly to the Moon. Compare: the royal N-1 rocket required 30 NK-33 engines in the first stage.

Like Raptor 3 and the Starship based on it look against the backdrop of engines for interplanetary rockets of the last century? At first glance, it’s modest: so far the thrust of the Raptor-3 in tests is about 275 ton-force, and even a promising goal is total 326 ton-force. Much weaker than the F-1 from the sixties.

In addition to quantitative indicators, qualitative indicators are also very important. From the point of view of the “consumer” – the rocket – its engine has two main quality parameters: specific impulse and thrust-to-weight ratio. Specific impulse here is the number of seconds that a rocket engine can maintain 1 kilogram-force of thrust while burning 1 kilogram of fuel. For SI fans, 1 kilogram-force is equivalent to 9.8 newtons, in which the rocket thrust is indicated above. F-1 from the 1960s at sea level had specific impulse of 263 seconds. NK-33 from the early 1970s – 297 seconds. Raptor 3 at sea level – 327 seconds.

That is, the NK-33 is 1.13 times, and the Raptor 3 is 1.24 times more economical than the F-1. This is an extremely important difference: more than 90% of the rocket’s mass is fuel. If you are tens of percent more economical, you will be able to launch much more payload on a rocket of the same mass.

Economy is doubly important for a reusable rocket, the cornerstone of any sound space program of our time. Due to the need to spend fuel to return both stages, Starship in a reusable version will have a payload of only 200 tons, and in a disposable version – 400. That is, full reusability literally doubles fuel consumption.

This is quite insignificant in terms of fuel price: a full refueling of liquid methane and oxygen for Starship costs less than one and a half million dollars. But this is extremely significant in terms of the price of the rocket: a reusable Starship has to be made approximately twice as massive and more expensive than a disposable one. Now I will explain with an example.

Using examples

First, about fuel consumption. Let's imagine that you started driving a car, not always one way, abandoning it at the final point of the route, but returning to your home. Fuel consumption will also double, about the same as that of SpaceX: after all, gasoline for the return journey will not come out of thin air.

But for rocket scientists, the problem will be much more acute than in this automotive example. They will have to make the “machine” twice as heavy and more expensive. In such a situation, an engine that is 1.24 times more economical than its competitors will mean an opportunity to reduce the mass (and price) of the rocket by tens of percent.

There is also a second (extremely important, although less than specific impulse) qualitative indicator of a rocket engine – thrust-to-weight ratio. You can make it quite economical due to, for example, large nozzles. But then its mass will soar that it will begin to “eat up” the payload. For example, the F-1 has a thrust-to-weight ratio of 94 (that is, it can lift 94 times its own weight). And NK-33 – 136. Therefore, the 30 engines of the first stage of the Soviet N-1 “for the Moon” were significantly superior in thrust to the five F-1s of the Saturn-V, but their mass, oddly enough, was less.

Raptors look more than decent in this sense. If Raptor 2 had a thrust-to-weight ratio of only 141.1, then Raptor 3 had a thrust-to-weight ratio of 183.6. That is, it “eats” half as much from the payload as its “countryman” from the time of von Braun would have done.

How to achieve a double gain in performance

In general, it is not difficult to understand why there is such a stir in the rocket world around these engines. Twice lighter per unit of power, but 1.24 times more fuel efficient than the previous lunar engine, Raptor 3 looks like a real breakthrough. But how did this happen?

The key phrase here is one: full-threaded closed loop. An open-cycle liquid rocket engine works like this: kerosene (the fuel may be different) and oxygen burn not only in the combustion chamber, but also in the gas generator. From there comes hot gas that rotates turbopumps that supply fuel to the combustion chamber. This is how the F-1 worked.

The weakness of this approach is obvious: the fuel burning in the gas generator by itself does not provide thrust. It is spent on an auxiliary function – supplying fuel to the engine. And this function is very energy-intensive: in the Raptor 2, more than 650 kg per second are supplied to each engine. Raptor 3 has an even higher number, but until its tests are completed, it is not even indicated. It's a lot of work to move that much liquid quickly through fairly thin pipes.

Therefore, in the 1960s, engine builders began to switch to a closed fuel cycle (it was invented by engine maker Alexey Isaev back in 1949). In such a cycle, part of the fuel after the preliminary combustion chamber produces hot gas, which passes through the turbopump, and then is not ejected, as in an open cycle, but enters the combustion chamber, where it is burned.

There are many advantages to such a scheme. The point is not only that fuel is used more economically here. Another thing is also important: since the hot gas after the turbopumps does not “fly out into the pipe,” then more fuel can be given to supply the turbopumps.

And this will significantly increase the pressure in the combustion chamber. As you might guess, as pressure increases, engine efficiency also increases. This is where the 13% leap in efficiency of the Soviet closed-cycle NK-33 came from.

But the Raptor is not just a closed-cycle rocket engine. It is also the world's first flying engine. full-flow closed cycle (or complete gasification, “gas-to-gas”). The full-flow Raptor 2 has already flown into space during Starship test launches, and Raptor 3 will retain this design feature and has already undergone its first burns on Earth. In this sense, Raptor did what the first engine of this design in history, the Soviet one, could not do. RD-270whose development was stopped at the ground testing stage.

Full flow means that not part of the fuel is converted into gas, as in the examples above, but all the fuel at once. SpaceX's new engine has a pair of pre-combustion chambers. The first burns a small amount of methane in excess oxygen. The second burns a lot of methane with a lack of oxygen.

Both streams of hot gas go through turbopumps. Despite their high temperature, this is not a problem: literally many hundreds of kilograms of gas fly through them per second, meaning heat is carried away from the turbines more efficiently than in other versions of the rocket engine. This means that the service life of the turbines is higher, and this is important for reusability.

After passing through the turbopumps, both flows go into the combustion chamber, where they are burned. What is the advantage compared to a conventional closed loop? It uses only one pre-combustion chamber, and the bulk of the fuel enters the combustion chamber without yet turning into gas. Naturally, it will evaporate in the chamber itself and still burn. But evaporation will consume energy (albeit a little). The burning rate of such an initially liquid fuel will be lower than the burning rate of the fuel initially converted to gas.

As a result, the specific impulse of a full-flow closed-cycle liquid-propellant rocket engine is 7–8% higher than that of a simple closed-cycle liquid-propellant rocket engine. This is the main source of the Raptor 3's fuel economy leap over the Soviet NK-33 from the 1970s.

With all the breakthrough features, Raptor engines are also inexpensive: in the series, SpaceX estimates them at 250 thousand dollars apiece.

How will innovations in Raptor 3 affect Starship?

It must be said frankly: at the moment, the final characteristics of Starship are still in the process of being refined. Only after testing will it be possible to speak about them more or less confidently. And as you know, these tests are hampered as much as possible by US government agencies such as the FAA.

There are voices that SpaceX should be deprived of the opportunity to work altogether. For example, the former US Secretary of Labor from the pages of the press speaks: “Elon Musk poses a clear and present threat to US national security. The sooner the US government revokes his clearance, terminates contracts with him and the entities he controls, and creates its own alternatives to Starlink and SpaceX, the safer America will be.”

Musk made technical decisions about the Raptor’s operation scheme, the choice of fuel for it, the choice of material for Starship and many others personally, often against the wishes of his engineerswhom he then had to convince for a long time. That is, without its founder, SpaceX will clearly create a completely different Starship.

And yet I am inclined to expect that it will not be possible to deprive him of the opportunity to work with space technology. The thing is that the US government does not have and will not have any alternative to SpaceX in the foreseeable future. This is indicated by Boeing's ten-year struggle to create the Starliner manned spacecraft or Blue Origin's quarter-century of work, which has not yet resulted in a single orbital flight.

So, most likely, the expected impact of Raptor 3 on the third version of Starship will still be realized in practice. And it makes a very serious impression: instead of 121–124 meters, like the first and second versions of Starship, the expected Starship 3 will reach a height of 150 meters. Of these, 80 meters will be on the first stage and 70 meters on the second stage-ship.

From applications SpaceX's regulator, the government agency FAA, knows that the third version will have 44 Raptor engines in the carrier: 35 on the first stage and 9 (instead of the original 6) on the second. It turns out that they will cost about 11 million dollars. The payload of the “third senior ship” will increase from 100 to 200 tons in a fully reusable version. The sealed volume of the second stage-ship will be in the region of a thousand cubic meters, that is, larger than the ISS. Actually, in terms of habitability, this will be the ISS, only capable of reaching other celestial bodies and returning back.

What about the competitors?

Traditional space players are not eager to develop new rocket engines, so they are not exactly available in hardware and are not being actively tested. Just about the only one playing even on the same field as Raptor 3 is Blue Origin's BE-4 engine, owned by Amazon owner Jeff Bezos. BE-4 also runs on methane and liquid oxygen, since methane burns with almost no soot, which is important for reusability.

But there is a nuance: Blue Origin has never published any complete data about its engine. The company specified its absolute thrust (250 ton-force) and specific thrust – 340 seconds (without even specifying whether it was at sea level or in a vacuum). It seems to be as good as the Raptor 3 (327 seconds at sea level, 380 seconds in the vacuum version for the second stage). But in the side-by-side photo, it's easy to see that the BE-4's engine nozzle is much larger (estimated to be at least 1.8 meters) than the Raptor's (about 1.3 meters). This means that the BE-4 is definitely heavier per unit of thrust: roughly its thrust-to-weight ratio should be in the region of 80–100 versus 180 for the Raptor 3. It also does not allow for the same tight installation as the Starship: the engine nozzles simply will not fit under the rocket.

As a result, the New Glenn rocket, with a diameter of seven meters, will have only seven BE-4 engines in the first stage, and Starship 3, with a diameter of nine meters, will already have 35 Raptor 3. Since the absolute thrust of both engines is comparable, it turns out that the thrust of the first stage of Blue Origin five times lower.

Because of this, New Glenn's payload – due to launch in November 2024, by the way – will only be 45 tons. But this is not the worst consequence of motor lag: there is also a second stage that is not very reusable. Following Musk's lead in 2021, Blue Origin is thinking about reusability for the second stage as well. But this Jarvis project is so far from being realized that it is not even clear how the New Glenn payload with a reusable second stage will be reduced.

Why did this happen? Why are the parameters of the BE-4 lower than the Raptor 3 to such an extent? The main reason: Blue Origin's owner heavily recruited people from traditional space companies. And he did not impose his ideas on them, sometimes going through the resistance of engineers, but accepted their concepts.

As a result, the BE-4 does not have full flow (just a closed cycle), the pressure in the combustion chamber there is several times lower, and all this leads to lower fuel efficiency and greater weight. It probably shows 340 seconds only in a vacuum, and compared to 380 seconds, this is not enough for the Raptor 3. The BE-4 would have been an excellent engine in the early 1970s, on par with the NK-33, only with methane. But half a century later, compared to Musk's engines, it looks too traditional.

But what about the Chinese, about whom they say that they copy your development faster than you can complete it yourself? Three years ago they really set themselves serious goals to create a full-flow closed-cycle liquid rocket engine YF-215 for the reusable Long March 9 rocket with a 150-ton payload. Just like its Western competitors, this liquid-propellant rocket engine runs on methane and oxygen (and the reason is the same – it needs reusability). But there is nothing special to discuss here yet: if the BE-4 at least flew, then the YF-215 today did not even have a complete burnout of the engine itself (only the preliminary combustion chambers). Considering that the target completion date for Long March 9 is 2033, this is logical.

Will Raptor 3 take people to Mars?

A breakthrough engine, and even the rocket it enables to become truly innovative, is a means, not an end. Musk's goal is known, and it is the same as Korolev's. Just like Korolev designed the N-1 for Marsbut was able to persuade the government customer only to the Moon, and SpaceX, creating Starship 3 and Raptor 3 for Mars, has so far been able to persuade the government customer only to the Moon.

But even here big problems began. First, the government agency FAA allows Starship test flights many times less frequently than SpaceX requests. Let's say the company is ready for the fifth test flight since August, but permission was given only on October 12. And NASA is in no hurry to intervene to help Musk bypass the government bureaucracy. The fact is that the Agency benefits from SpaceX’s delays on the lunar version of Starship.

To land on the Moon, you need not only the Starship HLS, but also a ready-to-fly Orion spacecraft. On it, NASA plans to launch people into space to land on the Moon; a direct launch on Starship is not included in the government office’s plans. The agency motivates this by the fact that there is no emergency rescue system on board Starship.

The truth is that it would be possible to transfer astronauts in orbit to Starship HLS from the Dragon spacecraft. In the American space community, they tend to explain the need for the Orion spacecraft and, accordingly, the SLS rocket by another consideration: each such launch is based only on operating costs costs $4.1 billion. Well, if we also take into account the past costs of developing all this hardware, then these figures are like states American trade press, “easily doubled.”

The overall estimate of NASA's spending on the program for 2012-2025 alone is: 93 billion dollars. This means a lot of money and expensive jobs. By comparison, under von Braun, each additional Saturn V mission to the Moon (operating costs) cost 2.3 billion in modern dollars, that is, almost half as much as in the 2020s.

One Starship HLS flight will not cost even a billion dollars: according to SpaceX’s own calculations, its target cost is somewhere in the region of 700 million. It's too cheap for the Agency to be willing to fly to the Moon on it. This is a fatal and fatal flaw: SpaceX can do a lot, but it can’t do one flight for four billion. Such difficult tasks for players in traditional American space.

As a result, the plan calls for a manned flight of Orion to the Moon (with a crew), a parallel flight of Starship HLS without people, and an in-orbit docking with the transfer of astronauts to the SpaceX ship. Then people will sit on Selena, walk on it and return to orbit.

Can Starship HLS with Raptor 3 provide all this? My opinion: with a high probability – yes. But the problem for NASA is that the Orion ship's heat shield suffered serious damage on a test flight for unknown reasons. A new test flight would cost four billion, but there is no such money. It’s somehow scary to send people without a new test flight. So far, the agency has not figured out what to do with this dilemma.

In addition, NASA does not have spacesuits for landing on the Moon. The Axiom company, which was ordered to create them, has not yet presented prototypes with a working life support system. A more she has difficulty paying her bills and mass layoffs. The photos she recently showed off are in many ways reminiscent of photos of earlier NASA spacesuits from 2019. Collins, the second company that NASA ordered a spacesuit for extravehicular activity, generally said that the task was too difficult and it simply would not do it.

Of course, SpaceX makes its own extravehicular suit. So far he looks so-so: with an air hose and without normal mobility. But the company is developing quickly and, in theory, could have time to make spacesuits. True, hardly by 2026, when the United States is formally scheduled to return to the Moon.

The FAA's inhibition of Starship testing is now beneficial for NASA: this way the agency will not be asked in Congress where its spacesuits are and why the Orion + SLS project costs 40+ billion dollars, but never showed decent performance of the ship's heat shield in tests.

It turns out that it is premature to raise the question of whether Starship with Raptor 3 will be able to deliver people to Mars. The more pressing question is whether there will be any Starship at all in the 2030s, or whether Blue Origin will manage to begin orbital flights this year (and NASA will shift all its attention and cash flows to it).

From the point of view of politicians like the former US Secretary of Labor, the fact that New Glenn will not be able to land people on Mars is not the most significant thing. But the revocation of Musk’s clearance and the termination of all contracts with SpaceX are truly important. If such policies do not succeed, then from a technical point of view, Starship 3 with a payload of 200 tons is quite sufficient to send a ship with people to Mars. And even to organize regular flights there in the next decade. After all, cautiously pessimistic estimates of the cost of launching into space for such a fully reusable rocket are about $100 per kilogram. This is enough to implement Sergei Korolev’s idea of ​​landing a man on Mars within the next ten years.

The key word in this whole situation is “if”. So far, SpaceX's technical success is greater than its ability to get along with the right people. This soft skill must be learned, otherwise Musk’s Martian project will face the same sad fate as Korolev’s Martian project.