Jump to content

sevenperforce

Members
  • Posts

    8,984
  • Joined

  • Last visited

Everything posted by sevenperforce

  1. Yeah, I don’t think “orbital-class” is at all nitpicky. It’s not a distinction without a difference. A F9 booster is the first stage of an orbital rocket; the New Shepard booster is the first stage of a suborbital rocket. If Bezos wanted to replace the New Shepard capsule with an orbital rocket stage just to show that it can be done, then perhaps Elon would welcome him to the club. Otherwise it is more like a sounding rocket than anything else. A Falcon 9 booster goes nearly twice as high, I believe.
  2. It's also variable given that you'd have to choose a hypersonic L/D ratio and those aren't necessarily clear. There's an asymptotic curve but even it is based on assumptions.
  3. Nah, it made it more interesting! We may be able to get an idea of how they're doing the ogive once we get a look at the opposite sides of these nosecones. These are clearly the leeward side given thruster placement.
  4. It took me SO long. Yes, agreed. But the fact that they're willing to do curved tiles means they are clearly willing to use a wider range of tile types. Doesn't necessarily answer how they're doing it, only that they'll probably be using custom tiles. Curved tiles per se won't fix the ogive problem. They can use flat tiles for the ogive; they just either have to taper them or they have to figure out a way to reduce the number of tiles per layer without leaving huge gaps: Using tapered tiles (on the left) can produce a smooth ogive without any larger gaps, but it requires the greatest number of custom tiles (one new tile type per layer) and rapidly becomes difficult as the tiles get narrower and narrower. Using a handful of custom tiles to "delete" one tile from a row, essentially creating a seam, will have more large gaps but will allow you to use the same tile types all the way up. The curved nose will be interesting. I am wondering whether they will go with a single monolithic nosecone at any point. That area is going to take the greatest heating and so it makes sense to use custom tiles but it means one catastrophic failure point instead of many small, inconsequential failure points.
  5. The windows thing is meh. The Mercury capsule had crappy windows; what of it? The escape system is more significant to me. And the fact that you're not actually going above the Karman line. But making a whole infographic to whine about it seems pointless.
  6. So there are. I suppose that settles that question.
  7. **looks up, ten pages into a 1975 NASA treatise on ideal impeller speed and its relation to flow density** Apparently the problem is most serious with liquid hydrogen. LOX and denser fuels can operate well at speeds as low as 3,500 rpm but hydrogen impellers work best at speeds as high as 40,000 rpm. One of the early Titan II main engines used a single turbine with a central driveshaft that was geared independently to both the fuel impeller and the oxidizer impeller. That allowed them to spin the main turbine at whatever speed was ideal for the exhaust itself and then gear that in one direction for the fuel and in the other direction for the oxidizer. A later Titan II main engine used an inline fuel pump and a geared oxidizer pump. The F-1 engine used a common shaft.
  8. No, peak Isp is a combination of three main things: propellant choice/mixture ratio, power cycle/chamber pressure, and expansion ratio. All three are significant. Go take a look at the comparison of orbital rocket engines page on Wikipedia. I just added a column for all power cycles so you can make comparisons. Power cycles are EXTREMELY important. The RD-58 (and virtually all Russian or Russian-derived kerolox engines) achieve high vacuum specific impulse by using staged combustion. The staged combustion cycle is more efficient than the gas generator cycle. You will NEVER be able to reach staged-combustion levels of specific impulse merely by adding a larger nozzle to a gas generator engine. You would not be able to reach 348 seconds of vacuum specific impulse by adding an altitude-compensating nozzle to a sea-level Merlin 1D engine. At most you could reach around 346 seconds. Additionally, your engine would weigh almost twice as much, and it would be much larger. If SpaceX could simply snap their fingers and add an altitude-compensating nozzle to the Falcon 9 lower stage to increase TSTO, why wouldn't they have done so already? It doesn't make sense. In reality, the Merlin 1D already uses an altitude-compensating nozzle. It is overexpanded at sea level, just like the RS-25. If SpaceX made it any bigger, they wouldn't be able to fit as many of them underneath the Falcon 9, so they would lose payload. My mistake, I was thinking of the YF-77. You are correct that the vacuum specific impulse of the RS-68A is 412 seconds. Yes, the performance of the RS-68 would be radically improved if it had a vacuum specific impulse of 470 seconds. You could achieve this by replacing the RS-68 with a completely different engine that is capable of reaching 470 seconds. The RS-68 is not capable of reaching 470 seconds of vacuum specific impulse, no matter how ridiculously large of a nozzle extension you give it. It is a gas generator engine and so it will never be able to exceed 440-450 seconds. Consider, for reference, the LB-5 engine, which is also hydrolox and is also a gas generator. It already has a vacuum nozzle and it pulls just under 450 seconds. You can get higher vacuum specific impulse on an engine by giving it a larger nozzle extension. However, this produces diminishing returns. You cannot get more energy out of propellants than the engine power cycle provides them. There is no free lunch. The RL-10 is a closed expander cycle. Of course it gets a higher peak specific impulse than an engine that dumps turbine exhaust through a gas generator. By analogy: wearing brass knuckles makes punching do more damage. But wearing brass knuckles does not make you punch harder. No matter how large a set of brass knuckles you give to a ten-year-old, they will not be able to punch as hard as Mike Tyson. They will do more damage than if they didn't have the brass knuckles, but they still have a limit.
  9. Does anyone know what type of staged combustion the ISRO's CE-7.5 engine uses? I'm presuming fuel-rich but I don't know. I can't imagine that it's full-flow.
  10. I spoke too broadly; not all engines have gearing. Raptor doesn't need gearing because it uses inline impellers on both its turbopump/preburner assemblies, and the both the Merlin 1D and the RD-180 use single-shaft dual-impeller turbopumps run by the turbine. The RL-10 does need a gearbox, however, because it uses dual-shaft turbopumps operated by only one turbine. The hydrogen expansion through the single turbine drives the hydrogen turbopump, with its drive shaft geared to the separate drive shaft for the oxygen turbopump. You can get away with a single-shaft dual-impeller turbopump (Merlin 1D, RD-180) when the density of your fuel and oxidizer are close enough to each other to allow it, like with kerosene and LOX or with hypergolic propellants (when you're using a power cycle other than pressure-feeding). But with a hydrolox engine, the turbopump impellers need to spin at significantly different speeds to pump propellant most efficiently, so you're more likely to need gearing. The RS-68 and YF-77 avoid gearing by using dual turbine-impeller assemblies, each fed from the same central gas generator. The RS-25 avoided gearing by having two separate fuel-rich preburners: one for the high-pressure fuel turbopump and one for the high-pressure oxygen turbopump. However, one could very easily imagine a different version of the RS-25, with just a single fuel-rich preburner geared to the two turbopumps. Now that I think more about it, I don't believe they'd even have blade pitch. All modern turbopumps use centrifugal impellers so they shouldn't have variable pitch. They can just reduce turbine RPM if they need to downthrottle.
  11. Random question, but do turbopumps have transmissions? Obviously there is gearing between the turbine and the impeller but does anything in that ever need to “shift” in some way? I wouldn’t imagine so, but I am curious.
  12. In KSP, most engines have very limited gimbal range, so you need fins. Three minimum, four if you want fine pitch control on ascent. Anything more is probably overkill; if you want better stability, just use larger fins. in the real world, engines always have sufficient gimbal range to maintain pitch and yard, so you never need fins unless your vehicle is intended to fly at some point without active gimbal control.
  13. The RD-0124 uses oxygen-rich staged combustion to get a whopping 359 seconds of vacuum specific impulse, but it does so at the expense of an engine which (a) cannot be fired at sea level, (b) cannot be throttled, and (c) has a rather poor T/W ratio of only 52.5 at its highest-thrust configuration. You cannot go full-flow staged combustion with kerolox because of coking issues, so 359 seconds is probably the best you're going to get, and of course you don't get that just by slapping a different nozzle on the Merlin 1D. The RS-25 and YF-90 are also staged-combustion engines which share a propellant type but the RS-25's altitude-compensating nozzle means it loses about 2 seconds of specific impulse compared to the vacuum-optimized YF-90. It also has a sea level specific impulse 19% lower, which also means sea level thrust that is 19% lower. So if you tried to build an SSTO using a version of the RD-0124 with an altitude-compensating nozzle, you'd end up with a sea level specific impulse of around 289 seconds, a vacuum specific impulse of 357 seconds, and a sea level T/W ratio of only 42.1, which means you're REALLY gonna have trouble getting off the ground. SSTOs need high thrust at launch...you wouldn't want to go any lower than 1.2 at the very minimum. So your engine is going to be at least 2.85% of your GLOW. If you're aiming to match or exceed a 3% payload fraction (as @Exoscientist said upthread), then that means nearly 6% of your mass is just payload and engine, nothing else. Let's also assume, just to simplify the math, that the first 30% of the burn takes place at the fixed sea-level specific impulse of 289 seconds and the latter 70% of the burn takes place at the fixed vacuum specific impulse of 357 seconds. Next, let's assume, ad arguendo, that your tanks literally weigh **nothing**. m0 = 1, m1 = 0.718, mf = 0.0585 dV1 = 939 m/s, dV2 = 8778 m/s, dVf = 9.72 km/s That's barely enough to get into a usable orbit, and probably won't be once you factor in the higher gravity drag from your lower T/W ratio. And we haven't accounted for the weight of actual propellant tanks. The tanks and structure of the Atlas D (not including engines) equaled 4.2% of the total propellant mass. Even if "carbon composites" can magically cut that by 40% to just 2.37% of GLOW, you've already gone from a 3% payload fraction to an 0.63% payload fraction. And for what? Just so you can say you built an SSTO? What possible reason could there be to not simply stage it and deliver 5X as much payload for the same GLOW?
  14. Missed this before, but the RS-68 is a gas generator cycle. You're not going to get anywhere near 465 seconds or higher...more like the YF-75's 438 seconds, if that. The RS-68 already gets 430 seconds of vacuum specific impulse.
  15. How in the world would a sea-level Merlin 1D get 360 seconds or more of specific impulse via “altitude compensation” when the actual Merlin 1D Vacuum only gets 348 seconds?? Plugging in nonsense numbers yields nonsense answers. But (if true) how is this relevant? If you have access to “maximum Isp engines” and “weight optimized structures” then just use them to build a smaller TSTO that delivers your desired payload more cheaply than an SSTO.
  16. Going to nine Raptors instead of just six would also nearly solve the pad abort problem. That’s a T/W ratio of 1.4 instead of barely 1. It could also make single-stage ballistic hops feasible.
  17. And that would never be an issue because you would downthrottle or use separate landing engines. Does your fiction need ocean launch and landing for a plot point? If so, make it so. If not, don’t. I’ve never seen an ocean freighter two miles long, but I’ve seen a “land-ship“ that long or longer. If you’re landing at sea on a platform, the platform will be moving, which makes it waaaaay harder. If your plot requires a vehicle that launches and lands on water, make it so. Otherwise don’t.
  18. Note that the larger your vehicle becomes, the higher T/W ratio you can squeeze out of a nuclear thermal rocket, because bigger reactors mean more surface area to transfer heat to the propellant. There's really no risk of "nuking" the launch site; just use a water deluge on the pad. Use separate thrusters for landing. No big deal. I will also note that this has already been done. This is the Nuclear DC-X, proposed back in 2004. About the same size as Starship+Superheavy, but fatter to hold the lower-density hydrogen. A single aerospike nozzle fed by five large NTRs and LOX injection at launch to add thrust. It used the same forward stabilator design as Starship to effect a biconic re-entry with a metallic heatshield. Notionally used steam thrusters for RCS and landing. 100 tonnes to LEO, fully reusable. You could probably make it more efficient by adding Raptor engines for added liftoff thrust and the landing burn and using three different propellants: methane, hydrogen, and LOX. Max thrust would involve running methane through the NTRs with full LOX afterburning alongside all the Raptors. As weight dropped during the climb, you'd shut down the LOX afterburn flow, then shut down the Raptors, then finally transition from methane to hydrogen in the NTRs. Use methgox hot gas thrusters for RCS on orbit and use the NTR in hydrogen mode for the deorbit burn. Raptors for the landing burn. Here's a cross-section of the engine cluster so you can get a better idea of what it looks like: No reason why this wouldn't work. The fixed control surfaces even act as passive radiators on orbit so you can use the reactors to produce electricity via the Brayton cycle. And while the Nuclear DC-X would have used a heat shield, you're not the first person to propose going without one. If you want to go even bigger, you can ask nicely for the nuclear lightbulb design and do a 1000-tonne-to-LEO Liberty Ship with 15 km/s of dV, enough to go up and come back down. You **would** want to launch that from water or from a Pacific atoll somewhere, not necessarily because you'd torch the launch pad, but because you want a wider berth for abort modes if something goes wrong on launch or landing.
  19. If you want a ship that is designed to land on unprepared surfaces, then design a ship with a set of landing legs capable of landing on unprepared surfaces. If you do not want a ship that is designed to land on unprepared surfaces, then do not design a ship with a set of landing legs capable of landing on unprepared surfaces. You would want to run the LANTR in pure NTR mode, and downthrottled as well, which doesn't really make sense in the context of a very short landing burn. Methalox engines would do more aggressive surface excavation than a downthrottled nuclear thermal rocket, anyway. Again, if we are talking about nuclear engines, 700 seconds is really not that much. Of course it will make lava. Any modern rocket engine will make lava. The combustion temperature of the RS-25 is greater than the boiling point of iron. Exhaust from rocket engines is much, much higher than the melting point of rock. But no one leaves rocket engines firing at rock long enough to make lava because that would be a big waste of time. You would not be the first person to have asked this. No, you can land pretty much anywhere you want, if that's how you design it. Again, it's a design question. Start with what you want and design around that, and then figure out what kind of engines you'll need. Otherwise you're just spinning in circles endlessly.
  20. Like I said, this set of requirements is completely within the range of near-future scifi. 700 seconds of isp is not that much. It can land on Raptors on any prepared concrete surface, and with these kinds of energy budgets you can have as beefy of landing legs as you want. It's all very simple and straightforward if you just get a nice powerful LANTR bundled in between your Raptors. You're starting with the wrong problem. If you have a world where you're flying an LANTR-based SSTO multiple times per day, material engineering is not going to be a problem for you. Materials can survive all this easily.
  21. No, that's not how rockets work. You can always use engine clusters or a low-thrust mode. Besides, a reusable two-stage solution requires that the upper stage also come back down and land, so you have the same problem. A two-stage solution also requires a first stage with more powerful engines than the second stage, unless you are merely looking to do a chemically-launched hop to a few km. If you want a vehicle that can deliver a payload to low Earth orbit, loop once around the planet, and then propulsively lower its velocity to the point that it can make re-entry without a heat shield (kind of a weird waste, but whatever), then you need about 15.8 km/s. You'll need a little more for landing which is kind of slop at this point but let's kick it up to 16 km/s just to give us margin. Note that I didn't say anything about "100 tonnes" at all, because the payload amount is immaterial. Above a certain size, the specific impulse requirements involved are going to be on the same order regardless of the size of the payload. Starship+Superheavy has a dry mass of around 400 tonnes and carries 4600 tonnes of propellant. With a 100 tonne payload, that's a m0/mf of 10.2, so to squeeze 16 km/s out of that, you'd need an engine with a specific impulse of about 700 seconds. You can multiply or divide your payload, dry mass, and propellant mass by whatever number you want, and the specific impulse you need remains at 700 seconds. Probably well within the reach of a good near-future-scifi LANTR design, perhaps with methalox landing engines which would also provide added liftoff thrust.
  22. Do we know if the RB-series engines are already uprated for more thrust, or are they merely fixed and unthrottleable?
  23. But this is simply not true. It just isn’t. If your notional SSTO and notional TSTO both have access to the same high-performance engines and lightweight tanks, and both have the same GLOW, the TSTO will carry more payload to orbit. More as a fraction of GLOW and more as a fraction of total vehicle dry mass. But it can’t. Even with boostback penalties, it can’t. You can set engine performance, tank mass ratio, and heat shield/recovery mass **wherever** you want them, and a TSTO configuration will **always** carry more payload to orbit than an SSTO of equal GLOW. Without fail.
  24. Even F9v1.0 had an LEO payload of 9 tonnes. There’s no way that any modification of a F9B5 first stage would be able to loft anywhere near 9 tonnes as an SSTO. Maybe 1-2 tonnes max. But if you only need to lift 1-2 tonnes to LEO to begin with, you’re not going to be buying a Falcon 9 flight; you’re gonna be getting a rideshare or going with a much cheaper smallsat launcher, so the comparison makes no sense. And that’s without even factoring in reuse. It’s not lower-cost to throw away an entire reusable first stage just to avoid throwing away a much cheaper second stage.
  25. Trying to use more advanced materials to make an SSTO use case close is like adding spoilers to a UPS truck in hopes of making deliveries faster. Literally any advancements that could make an SSTO closer to workable would make a TSTO architecture significantly more workable.
×
×
  • Create New...