Global future space architecture thread

[nativechicken]

If you want to talk about actual physical hardware, Japan also built a VTVL demonstrator. They flew it in 2001.

I do most of my extensive reading in Chinese literature. If a topic isn’t mentioned, or is only infrequently mentioned, in these sources, I tend to overlook it, which can lead to forming incorrect views.

Japan did indeed attempt to make some achievements in the field of reusable vehicles back in the 1990s. I believe what you’re referring to is the part that I had overlooked.

 

[gelgoog]

The usable throttle ratio of modern rocket engines has consistently been around 40-50%.

Because it was not necessary to throttle them down further for their expendable rocket design.

It is trivial to make a deep throttling pressure fed engine for example. Like the Apollo Lunar Lander engine. A Flowmetrics engine is also trivial to deep throttle for much the same reason. It has no pumps just valves.

Please, Log in or Register to view URLs content!

Another rocket engine design that can deep throttle is the RL-10. It was originally an upper stage engine. A lot of them can deep throttle to save fuel.

Please, Log in or Register to view URLs content!

In engines with turbopumps and which need to fly at sea level it is more difficult to make the engines deep throttling but not impossible.

Considering the constraints of Earth’s gravity well, a practical VTVL system essentially requires around ten engines.

Not really. Not if you have a deep throttling engine.

One alternative is like in SASSTO or BETA. You have multiple small engines that you can individually control around a plug nozzle.

Please, Log in or Register to view URLs content!

Please, Log in or Register to view URLs content!

The X-33's linear aerospike also had multiple small engines.

This is another consequence of the N1 failure; it caused the aerospace sector to reject solutions involving a large number of engines operating in parallel.

In the 1960s the computer control was not there. For the same reason you only saw fly-by-wire become common in aircraft in the 1980s.
The N1 also had issues with quality control and lack of testing.

Furthermore, the engines built later by the US and Soviet Union (Russia) were basically around 200 tons of thrust class, while Europe’s and Japan’s first-stage hydrolox engines are all over 140 tons of thrust.

There were small engines like SNECMA Viking and Rocketdyne RS-27. But they were not designed to throttle.

Precise trajectory control with throttling to conserve fuel in flight requires advanced flight computers. Early rockets did not have them.

The reason SpaceX and Blue Origin were able to pioneer this path is entirely because their initial rocket engines had only several tens of tons of thrust. This made the concept of a reusable rocket much more feasible.

I would say these designs only became possible in the 1970s with analog computers, and in the 1980s with digital computers. But at that time most investment in space was over after the glut in the 1960s.

The US and the Soviets only put money into the Shuttle and Energia.

The core reason China needed 10 years to build a Falcon 9-class rocket is that the YF-100’s thrust is simply too large and cannot be sufficiently throttled down. Therefore, to create a truly recoverable rocket, they had to develop new 80-100 ton class engines to achieve Falcon 9-level capabilities.

It is just a slightly modified YF-100 in that CZ-10. Same base engine as in CZ-7 or CZ-5.

The CZ-10 represents a different technological path: one of tackling the problem head-on by solving the deep-throttling technology for a 100-ton class engine. The first goal is to achieve a minimum thrust of 1/3, followed by 1/5, and then to see if they can reach 1/10.

Choosing this technically difficult route was necessary because the CZ-10 project was not initially conceived as a reusable rocket. Its primary requirement was to meet the demands of the moon landing; reusability was a requirement added later. This is why the entire arrangement feels rather awkward.

It will probably work just fine. But LOX/Kerosene fuel is susceptible to coking so the engine will be less reusable than with Methane or Hydrogen fuel.

 

[ENTED64]

In practice, five years is enough time to build a heavy-lift rocket. SpaceX actually took 4-5 years to develop its heavy-lift rocket to first flight (in 2023). So, if China's aerospace sector aims to achieve a first flight in 2030 (with the expendable version), March-April 2026 will be the critical period. If there is no news by 2026, the CZ-9 will certainly be delayed to a 2035 first flight; otherwise, it will fly around 2030. There's no great secret—it's just the spending cycle of the five-year plan.

It's odd to me that there is such a large window, I don't think it would be that tied to the exact timing of FYP. Nonetheless, I guess 2030-2035 is more or less where I assumed first flight of CZ-9 would happen so it doesn't seem like it's delayed too much. I guess I was hoping for it to be a bit faster with CZ-10 being significantly earlier but it was probably unavoidable with all new engines.

 

[Blitzo]

When I use AI translation software, it sometimes has many problems, including partially misinterpreting what I say. It might also misinterpret what you say.

If I have misunderstood your meaning, I offer my apologies here.

That's fine, no problem.

I hope you can understand that it's difficult to have a conversation through AI translation, so ultimately I probably won't have much means of holding a constructive discourse if there is such a high risk of accidental miscommunication.

 

[nativechicken]

It is trivial to make a deep throttling pressure fed engine for example. Like the Apollo Lunar Lander engine. A Flowmetrics engine is also trivial to deep throttle for much the same reason. It has no pumps just valves.

The deeply throttled engines you're talking about are not the same as the ones I'm referring to, which are those currently under development in China.
My definition of deep throttling is basically a 10:1 to 5:1 ratio, with 3:1 barely being considered entry-level.
I am specifically talking about achieving this on engines of over 100 tons of thrust class. The Apollo program proved that achieving 10:1 throttling on a 5-ton class engine was no problem. The Constellation program also worked hard to achieve 10:1 throttling on a 10-ton class hydrolox engine, but it never entered substantive development. As for engines in the 100N to 1000N thrust range, I recall the U.S. has achieved over 20:1 throttling capability.
The throttling I'm talking about must be achieved on a staged-combustion cycle engine, and it involves retrofitting the YF-100, an engine not originally designed for true deep throttling, to meet this standard.
Under the constraints of these three prerequisites, the technical difficulty of providing such an engine is far beyond your imagination.
Currently, the Raptor engine (over 100 tons thrust class) and a host of various advanced Russian engines on paper all claim a throttling capability down to 20-30% of maximum thrust (a 5:1 to ~3.3:1 ratio). But as far as I know, this capability has never actually been realized or put into practical application because there are many problems at both low and high power settings.
In reality, achieving 30% minimum thrust (a 3:1 ratio) in a real engineering application is already extremely difficult. Of course, if it were a gas-generator cycle rocket engine, the difficulty would be much lower. However, the specific impulse (Isp) loss would also be greater (this is the key to the Merlin 1D's success, and why China's equivalent-class engines use a similar design, abandoning the staged-combustion cycle).
In the deep throttling research for the YF-100, it's necessary to comprehensively adjust over 3,000 of the original engine's detailed engineering algorithms and formulas (which should be for the engine control system). China has done a lot of concrete work, conducting preliminary research on 10:1 throttling between 2010 and 2015 (with the earliest work starting in 2008). After 2015, it entered various stages of substantive technical verification, and literature from 2024 basically confirms that 3:1 throttling has entered full engine testing.
This is the real reason the CZ-8R hasn't materialized. For a 120-ton class engine, at least a 5:1 throttling capability is needed to achieve a single-engine landing for the CZ-8R. A 10:1 throttling capability is needed for a dual-engine landing.
Currently, the first stage of Starship essentially uses two engines for the final landing hover. This is because Starship's landing mass is over 240 tons. With two Raptor engines throttled to 50%, the thrust roughly balances its own weight, allowing it to approach the landing gear at a controlled speed of 1-3 m/s. If it used three engines for retropropulsion, the thrust (3x120 tons) would be 1.5 times the force of gravity, which actually increases the landing risk (making speed control more difficult).
The real reason the current Starship second stage recovery doesn't dare to land directly on the launch pad is imprecise speed control, not an inability to control the landing position. To put it simply, the current Starship second stage recovery has only achieved precise positional control on the X-Y axes. Precise speed control is needed on the Z-axis. The second stage also uses two engines, switching to a single one at the final moment, because the second stage's mass should be around 120 tons. If the Raptor on Starship could achieve 3:1 or 5:1 deep throttling, then 5:1 would be sufficient for a dual-engine second-stage landing, and 3:1 throttling would basically allow for a single-engine second-stage landing (enabling a controlled approach to a hover). With a 2:1 throttle, you would need to control the deceleration precisely (the thrust exceeds the vehicle's weight, especially since the second stage needs to reorient itself before landing, which is even more troublesome).
Therefore, looking at the current stage, the Raptor has not yet truly achieved 30% minimum thrust (3:1 throttling), but it has achieved 40-50% minimum thrust (a 2.5:1 to 2:1 throttling ratio).

 

[Daniel707]

The deeply throttled engines you're talking about are not the same as the ones I'm referring to, which are those currently under development in China.
My definition of deep throttling is basically a 10:1 to 5:1 ratio, with 3:1 barely being considered entry-level.
I am specifically talking about achieving this on engines of over 100 tons of thrust class. The Apollo program proved that achieving 10:1 throttling on a 5-ton class engine was no problem. The Constellation program also worked hard to achieve 10:1 throttling on a 10-ton class hydrolox engine, but it never entered substantive development. As for engines in the 100N to 1000N thrust range, I recall the U.S. has achieved over 20:1 throttling capability.
The throttling I'm talking about must be achieved on a staged-combustion cycle engine, and it involves retrofitting the YF-100, an engine not originally designed for true deep throttling, to meet this standard.
Under the constraints of these three prerequisites, the technical difficulty of providing such an engine is far beyond your imagination.
Currently, the Raptor engine (over 100 tons thrust class) and a host of various advanced Russian engines on paper all claim a throttling capability down to 20-30% of maximum thrust (a 5:1 to ~3.3:1 ratio). But as far as I know, this capability has never actually been realized or put into practical application because there are many problems at both low and high power settings.
In reality, achieving 30% minimum thrust (a 3:1 ratio) in a real engineering application is already extremely difficult. Of course, if it were a gas-generator cycle rocket engine, the difficulty would be much lower. However, the specific impulse (Isp) loss would also be greater (this is the key to the Merlin 1D's success, and why China's equivalent-class engines use a similar design, abandoning the staged-combustion cycle).
In the deep throttling research for the YF-100, it's necessary to comprehensively adjust over 3,000 of the original engine's detailed engineering algorithms and formulas (which should be for the engine control system). China has done a lot of concrete work, conducting preliminary research on 10:1 throttling between 2010 and 2015 (with the earliest work starting in 2008). After 2015, it entered various stages of substantive technical verification, and literature from 2024 basically confirms that 3:1 throttling has entered full engine testing.
This is the real reason the CZ-8R hasn't materialized. For a 120-ton class engine, at least a 5:1 throttling capability is needed to achieve a single-engine landing for the CZ-8R. A 10:1 throttling capability is needed for a dual-engine landing.
Currently, the first stage of Starship essentially uses two engines for the final landing hover. This is because Starship's landing mass is over 240 tons. With two Raptor engines throttled to 50%, the thrust roughly balances its own weight, allowing it to approach the landing gear at a controlled speed of 1-3 m/s. If it used three engines for retropropulsion, the thrust (3x120 tons) would be 1.5 times the force of gravity, which actually increases the landing risk (making speed control more difficult).
The real reason the current Starship second stage recovery doesn't dare to land directly on the launch pad is imprecise speed control, not an inability to control the landing position. To put it simply, the current Starship second stage recovery has only achieved precise positional control on the X-Y axes. Precise speed control is needed on the Z-axis. The second stage also uses two engines, switching to a single one at the final moment, because the second stage's mass should be around 120 tons. If the Raptor on Starship could achieve 3:1 or 5:1 deep throttling, then 5:1 would be sufficient for a dual-engine second-stage landing, and 3:1 throttling would basically allow for a single-engine second-stage landing (enabling a controlled approach to a hover). With a 2:1 throttle, you would need to control the deceleration precisely (the thrust exceeds the vehicle's weight, especially since the second stage needs to reorient itself before landing, which is even more troublesome).
Therefore, looking at the current stage, the Raptor has not yet truly achieved 30% minimum thrust (3:1 throttling), but it has achieved 40-50% minimum thrust (a 2.5:1 to 2:1 throttling ratio).

Hello, can I ask you something.

According to your own analysis, how much percentage that the US can beat Chinese Moon landing plan (in 2030) with their current Starship HLS or any other backup plan they come up with?

 

[nativechicken]

Hello, can I ask you something.

According to your own analysis, how much percentage that the US can beat Chinese Moon landing plan (in 2030) with their current Starship HLS or any other backup plan they come up with?

Currently, the only possibility seems to be Blue Origin's Blue Moon MK1 (cargo version) being adapted into an Apollo-class lunar lander. The Blue Moon MK1 should have a BE-7, a 3-4 ton class hydrolox landing engine (thrust comparable to and payload similar to Apollo's lander). The Blue Moon MK2 is a lander with three BE-7 engines.
Furthermore, earlier this year, the U.S. announced that Blue Origin is building lunar landing simulation equipment (similar to what China unveiled in September). Therefore, if the Blue Moon MK1 is well-prepared (specific details are not currently known, but the plan appears absolutely reliable), it indeed has a 50% chance of landing on the moon before China. This is basically America's only realistic option.
Personally, I really like and admire Blue Origin. Its propulsion system portfolio is far superior to SpaceX's. In my view, to see if an aerospace company has a future and possesses profound thinking, you must first look at its propulsion system portfolio. Blue Origin is extremely formidable in this regard. In comparison, SpaceX's propulsion portfolio is highly problematic.
Blue Origin's main publicly known propulsion systems are:
BE-3: a 50-70 ton class hydrolox engine.
BE-4: a 300 ton class methane engine.
BE-7: a 3-5 ton class hydrolox engine (supporting landings on extraterrestrial bodies).
Blue Origin is actually the true pioneer of this current wave of VTVL applications. It's just that its technological path uses hydrolox propulsion, which makes it unfeasible for reusing the first stage of a practical orbital launch vehicle (a 2016 Chinese paper concluded that VTVL must use a hydrocarbon-based propulsion system). So, Blue Origin's current lag is truly regrettable.
SpaceX's problem in the engine sector (what I call abnormal thinking) is its biggest issue: one rocket model, one engine type. In reality, the best solution is for a rocket to have two different fuel types of propulsion systems or two thrust-level classes of propulsion systems.
Having one engine type per rocket isn't a major problem at the Falcon 9 level, but for larger rockets or complex space applications, it is not the optimal choice.
If I were SpaceX, I would lay out three thrust classes for methane engines: a 200-300 ton class, an 80-100 ton class, and a 5-10 ton class (the Merlin would be scrapped, as it's a kerosene engine). At the same time, I would maintain at least one 5-10 ton class hydrolox engine in reserve.
This kind of portfolio is what truly shows the makings of a king (i.e., having the qualifications to dominate). This is where Blue Origin's relatively better portfolio lies.
Additionally, Blue Origin is also participating in U.S. projects for deep-space nuclear thermal and nuclear electric dual-mode propulsion.
Therefore, in my eyes, it is the true heir to the throne of American aerospace. SpaceX is just an upstart, lacking the deep-seated foundation of a true king.
If you know Chinese historical stories, SpaceX is not in the role of a prince or crown prince, but merely a court-backed merchant (like Shen Yishi or Hu Xueyan).

 

[nativechicken]

My confidence in China's future in aerospace stems largely from its "off-the-shelf" portfolio of propulsion systems, as discussed above.
China has established a product reserve across multiple thrust levels in nearly all three mainstream liquid propulsion fields: kerosene, methane, and hydrolox. For each category, there are basically plans for heavy (>200t), large (100-150t), medium (25t-80t), and small (5t-10t) classes. The product line is incredibly diverse.
This signifies that China has completely entered a stage of "propulsion freedom." With propulsion freedom, are rockets still a problem? The great struggle for China's aerospace sector has always been that the propulsion system was the bottleneck. The Long March 2 and 3 series were propped up by the YF-20 family of engines, and the Long March 5, 6, 7, and 8 weren't much better, with a maximum thrust of only 120 tons. The next generation has finally solved all of this.
For advanced equipment, propulsion must come first. (The R&D cycle for advanced propulsion systems is generally more than 10 years, sometimes even 30). And this doesn't even include China's other propulsion systems; various powerplants for hypersonics and solid rockets are all developing at high speed.
My disappointment, distress, and disdain for NASA stem from this very issue. This kind of strategic planning and long-term layout cannot be left to companies like SpaceX and Blue Origin. It is something NASA must plan and arrange for in advance. NASA has failed to consider this, instead simply trying to solve problems by purchasing commercial services. If you have studied how NASA operated in the 1960s to 1980s, you would realize how ridiculous its current "buy commercial services" mindset is.

 

[gelgoog]

The buying commercial services mindset is just fine now that the private companies are large enough. NASA is back to its roots as a place to do aerospace research when they were called NACA.
NASA got sidetracked when the US Army Redstone Arsenal and Von Braun's team were pressed to beat the Soviets in space. That is the current Marshall Space Flight Center.

NASA still has engine testing facilities at Stennis. I expect those to continue.

Blue Origin was founded before SpaceX. If anything they developed at a sedate pace.

All SSTO designs typically use LOX/LH2 in the first (only) stage. But if you want to maximize payload you are better off with a TSTO design with denser fuel in the first stage. Kerosene preferably. Methane for reusability.

LOX/LH2 is less bad than some people think in terms of density because first stages typically run with an oxidizer rich propellant mix. An SSTO with LOX/LH2 has a variable mixture ratio. Oxidizer rich at low altitude and fuel rich at high altitude.