Stoke Space Second Stage Solutions So Sick
So Stoke is pretty popular ay. I remember there was this fun time in like 2022 on discord where about once every month somebody would discover Stoke for the first time and it would be this fun little conversation. That boat has well and truly sailed.
Now onto this trade space. I'll begin with engine ISP as that's the easiest to understand. Now despite what this meme and Stoke might have you believe, the engine isn't that efficient in a vacuum at 430s of ISP. Conventional vac optimised expander bleeds can hit 450s, so they're down relative to that. And against a hypothetical FRSC hydrogen upper stage engine which could hit 470s of ISP (and they are capable of building); they're losing 40s. 430s is what open sea level engines achieve.* Essentially, if it was methalox, the ISP would be ~345s, below that of staged sea level engine like Raptor. This is their primary performance loss in the trade space. So what do they gain for that? They get to integrate 3 capabilities into the one engine, effectively decreasing the system complexity and dry mass of the stage.
*Although worth noting that those sea levels need ~100 bar in the chamber to achieve that ISP, Stoke does it with 40 bar. 40 bar means that they get to do things like skip bi metallic jackets on the chambers and can instead just have a copper alloy print. This combines with the 'positively benign' environment that a bleed expander turbine experiences so their engine doesn't experience a lot of stress; it should have a really good lifespan and be easy to reuse.
So it's kinda more of a sea level engine that gets reasonable performance in a vac for how robust and versatile the engine is. Something that restricts the performance of Starship, especially for HLS, is the fact that they need those sea levels in the skirt. Because they can't fit a gimballing Vactor in there, it means they have to deal with the ISP drop that comes with having to use 350s engines alongside your 375-380s vac engines. (not to mention throttling problems).
As a landing engine it’s fairly nice as it’s much easier to deeply throttle the architecture because you don’t have to worry about combustion instability in the pre-burners/gas generators, nor in the combustion chambers themselves as they can just limit the number of chambers that receive propellant. I don’t know what the theoretical limit is, but the renders do show 10% throttling.
Actively cooled heatshields are to ablative heatshields as liquid rocket engines are to solid rocket motors. (don't think about passive tile solutions, they don't fit in the analogy). This is something that people doubt the technical viability of. However, what I think people need to realize is that propulsion engineers already have to deal with thermal fluxes of ~100 MW/m² on the combustion chamber walls in their engines. The ~10 MW/m² of LEO re-entry is actually a breath of fresh air for them. In this case, the Stoke’s ‘aeroplug’ won’t experience that high of a heating due to the relatively low thrust/area of the engine, so re-entry thermal flux will likely be higher in this case. But the point still stands, it’s well within the propulsion engineers experience; get out of here TPS engineers! (just kidding we still need you for the material science of optimizing the alloy for the heat shield).
What does an actively cooled heat shield achieve? It’s no secret that the heat shield is one of the weaker elements of Starship’s design currently, with 18,000 tiles and a secondary ablative heat shield taking a significant portion of the mass and maintenance budget. It’s more robust than I would’ve thought; well at least the vehicle is. However for a fully and immediately reusable orbital heat shield, there’s definitely a lot to be improved upon. Active cooling has a much better shot of getting there, because effectively we’ve already gotten there previously with thrust chambers on reusable rocket engines. Monitoring the heat shield for damage should be easier, given that it’s inherently easier to place sensors and watch the flow of propellant compared to analysing 18,000 tiles. I’m not sure it would be easier to repair per say, but if it is permanently damaged, you could just treat the heat shield as a Line Replaceable Unit (LRU) and slot another one in.
Now comparing Starship flaps to Stokes lifting body heat shield is interesting, because the flaps do provide greater control. The mitigating factor here is that Nova doesn't need the greater control. Because SpaceX has a tile solution and enters on the broad side in an passively unstable state, they need the flaps to provide that flight control during re-entry and landing. Nova is passively stable with its base heatshield, so with the offset to provide a degree of lift they have enough trajectory control.
Chines are just in there for the meme of it (storage space). If they do get rid of the chines on Super Heavy, I’ll definitely miss them. They add so much to the look of that vehicle.
So what about hydrogen? To get something out of the way, I'm tired of people treating hydrolox as a curse on price. Yes the hardware (valves/seals) to support it does cost a % more, but it's not like you're forced into hundred million $ stages as a result. It's like going, 'oh well Shuttle proved reusability could never be cost effective, so Falcon 9 reuse will fail.' You're blaming mixed technical trades when the ultimate driver is managerial and organizational culture.
Hydrogen is fairly important to this architecture. A likely significant reason why methane wasn’t competitive with passive tile solutions for Starship was that by mass, it requires 4.5x as much methane to extract the same amount of heat as hydrogen. Now volumetrically this is less than the hydrogen, but mass is the driver at this stage of the mission.
I will say that VTVL first stages prefer high mass second stages as staging earlier decreases delta v + thermal flux to get to the landing site. In Falcon 9 case; replacing the current stage with a hydrogen stage of the same proportions would decrease performance to LEO by like ~3 tons for the reusable missions. This is why the higher impulse density propellants like methalox/kerolox are nice in this case and why Starship in particular.
From a long term perspective, hydrolox is better for the Moon, methalox is better for Mars. I will say that in order to produce methalox via ISRU on Mars, you can produce hydrolox. However impulse density and storage conditions make it preferable for methalox on the red planet.
From a Mars and a LEO perspective, the Starship architecture likely does provide better performance. Given that those are the 2 biggest theoretical market's for SpaceX, it makes sense that they optimized around that space. However the presumably more effective reuse and performance in the cislunar space, likely give Stoke's design the edge there. (Obviously there's a degree of separation in that one is a 3 ton to LEO vehicle and the other is a 100 ton to LEO vehicle)
