Single Stage to Orbit spacecraft and spaceplanes a reality?

[Nik]

Mustard & Skylon

Mustard, IIRC, was a UK design with three almost-identical 'lifting bodies', the outer pair acting as fly-back boosters. It had the great advantage that you built the sub-orbital units first, and flight tested them before going for LEO.

Skylon, with its 'deep cooled' Sabre engines, is one of those 'once in a lifetime' near-genius notions, combined with the persistent bloody-mindedness to find a singular, practicable solution to outrageous engineering problems. If their filament heat-exchanger scales, they'll have an arm-lock on LEO for modular loads.

Notice, however, that UK.Gov didn't support Reaction Engines Ltd until hard-headed Euro-backers coughed up a cool million to take the work further...

Back in the mists of time, IIRC, UK.Gov decided that it could not afford a space-faring facility of its own. That's after orbiting one (1) satellite. They decided that a European consortium was better value. So, IIRC, a succession of UK Black Knight boosters worked as designed, but the French second stages all exploded or were destroyed due misfires / veering. The 'S' word, 'Sabotage' was mooted, but never proven: Remember, this was the height of the cold war, and paranoia bloomed like mushrooms on manure. And, yes, it was 'Rocket Science'. Even the fastidious Japanese found that launching stuff is much harder than it looks, while the Brazilians lost an entire pad-crew when a solid rocket booster caught a static flash and lit-off in their vehicle assembly area...

The US focus on space capsules was, IIRC, primarily due to the problems with the X-15. Those early flights set new records, even qualified several pilots as 'astronauts'. Unfortunately, the later, 'faster' attempts had horrendous problems with heat damage and crashes. In parallel, the small 'lifting body' approach showed that there were dire control issues, especially during landing. Remember the 'Six Million Dollar Man' intro ? That crash & burn was *real footage*, the pilot was very badly injured...

IMHO, the other problem with US reusable designs was the pi$$ing contest between the USAF-supported Dynasoar and the 'civilian' program. Those issues returned to blight the Shuttle, as the cargo bay had to be big enough for massive spy-satellites and their Earth-focused cameras and antennae. Had a smaller Shuttle been built, it would have piggy-backed on cargo pods and had wing-loading light enough to avoid the severe heating issues that forced use of those uber-fragile 'tiles'...

Incidentally, that's why Skylon doesn't need 'Shuttle' tiles...

 

[Shevek23]

I have a question. If you use a TSTO setup with an airplane as the first stage, could the plane get high enough to use an NTR as the second stage? Use the plane to get above most of the atmosphere, so you can do a horizontal burn with the NTR.

Aside from getting high enough in the atmosphere that your nozzle performance will be close to vacuum performance, so you design for the latter and get good efficiency all the way, and a tiny reduction in the delta-V you need due to the plane's own velocity, does this do any good? Even with optimal nozzle performance the NTR still has the liability of low thrust/weight ratio. Perhaps if the airplane can pull up into a steep climb and then launch, sacrificing some of that horizontal delta-V for "buying time" for the rocket thrust to do its work?

Taking Mach 1= 300 meters per second (which, full disclosure, is the cheat I just about always do, rather than try to figure it out based on temperature) and supposing you're talking about a subsonic airplane, let's say it's going at 283 meters/sec straight and level when the pilot pulls up with a 3 g pull; it should be climbing at 45 degrees when the plane has climbed a further 600 meters, so what we need is for the atmosphere to be thick enough for the wings to sustain 2.7 Gs of lift (force 2.7 times the weight of the plane that is) at some angle of attack the plane can sustain, and thick enough so the plane's engines can provide sufficient thrust to handle whatever drag this lift angle implies, plus .7 G on the craft. Which, remember, includes our rocket with NTR stage! Well, heck, we can always have a big solid fuel JATO unit on the plane, it only has to burn a few seconds...

So now we are releasing the nuke rocket to fire (hopefully the launch plane has some way of ducking past the exhaust!) while it's been lobbed up and forward along its launch track at 200 meters per second in each dimension.

Now obviously if the NTR has such a miserable thrust/weight ratio that we can't afford to make it even generate 1 G initially, we are still pretty screwed! We have to reach a speed of something like 8000 m/s, starting from 200, and we only have 20 seconds to do it before the thing starts to fall if we daren't use any of it vertically...it would have to thrust at 40 Gs and if it could do that (never mind that we'd mash any crew to strawberry jam) we wouldn't be fussing around with the airplane launch, would we?

If initial thrust, before we start using up fuel, is in the ballpark of just 1 G, and we are using an NTR in the first place because it has superior specific impulse--say it has an effective exhaust velocity of 4000 meters/sec, then 6/7 of its initial mass needs to be reaction mass, meaning that a 1 G rocket at the start of the burn is thrusting the remnant mass at 7 Gs at the end of it! Ouch! Clearly there are other considerations than just how much thrust/weight can we wring out of the thing, for safety and survival of the payload we need a rocket that initially can only manage a half G!

With that specific impulse, starting and maintaining that thrust, it takes 686 seconds to read orbital velocity assuming we thrust full horizontal, but meanwhile of course the vertical initial velocity of 200 meters/sec has long ago been checked and reversed and the ship has been crashing to the ground...Suppose that contrary to our concerns about thrust/Weight ratio, we could go ahead and make the engines 6 times bigger and simply throttle them back for a steady 3 G push? Well, then we surely wouldn't fall because even maintaining exactly 3 Gs horizontally, after a bit of burning we'd have spare thrust available to angle downward and maintain height and soon even gain as much climb as we need. At 3 Gs sideways we only need 260 seconds of acceleration to do the job, our initial 200 meters/sec buys 20 of those seconds during which it becomes possible for the rockets, by angling downward a bit, to provide 1.8 G of thrust vertically while maintaining 3 G horizontally. Of course this adds to the total delta-V our reaction mass must provide for.

But such powerful NTRs could clearly provide for direct launch from the surface, even if we suppose their nozzles would be less efficient there.

What if the specific impulse is really fantastic, like 8000 meters/Sec, so the mass ratio is e? But we are restricted to a rocket that delivers only say 2 Gs of thrust on burnout, meaning only a bit under 3/4 of a G initially. Then it would take 700 seconds to reach burnout and again, the rocket not even being able to deliver a full G within 20 seconds of firing, clearly the thing crashes. Again we need a decent amount of acceleration initially.

So, what good does it do us really, to drop it from a big airplane? If our initial velocities were a lot faster than sound, they could help, by seriously reducing the necessary delta-V, by lobbing the thing up at a faster speed to buy more time.

For instance, if the plane could go at Mach 5, and then do that 45 degree climb (which is much more problematic because the turn radius is 25 times as great meaning a climb through major density changes of the atmosphere) then we'd have something like a kilometer/sec initial velocity in both directions, buying us 100 seconds and 50 km of altitude increase before it starts falling, and reducing the necessary sideways delta V to 7000; using the 8000 m/sec impulse rocket above, the mass ratio is only 2.4 and assuming again a rocket that can only give us 2 G at burnout, we now need 560 seconds to reach orbit, but after 100 seconds the thrust available is almost a full G's worth. Clearly we are still in trouble but there might be some way to finesse things a bit; for one thing we could have started with more powerful engines, which would have thrust up to 3 G at burnout, then we'd clearly have some margin at this point to keep accelerating while holding gravity at bay. Just as clearly, this fix would not have saved us with the slower launch.

Restricted to subsonic, it seems to me we'd do just as well to have launched the thing on a chemical fueled booster (one optimized for the lower atmosphere it spends most of its time thrusting in) to the same altitude and velocity. For instance, even allowing for a very slow and low-thrust chemical booster that doesn't accelerate the craft to sonic speeds until it is 10 kilometers high, which is obviously very wasteful, it shouldn't add more than 40 percent of mass to the nuclear stage of the rocket--I'm guessing that's much lighter than any airplane that could haul said rocket to that altitude and speed. And if we aren't afraid to hit Mach 1 much sooner in much denser air below, we can get it below 20 with the same kinds of rockets.

So yes, I was aware that there are air-launched systems, but unless the planes can reach serious multiples of the speed of sound, I can't really see why!

I ask because I've dug up some proposals for absolutely gigantic aircraft from the 60s, including one big enough to airdrop a Saturn-V.

On the other hand, I would be really thrilled and grateful to be able to see these proposals!

I'm especially interested to know--does an airplane capable of lifting and dropping a Saturn V have to weigh as much as its payload does? That's a rule of thumb I've been going by. I'd like to see if I'm right to do so. (Rule of thumb being, useful lift tends to be about half the total lift capacity of any aircraft, fixed structure typically being about half the capacity.)

 

[asnys]

Aside from getting high enough in the atmosphere that your nozzle performance will be close to vacuum performance, so you design for the latter and get good efficiency all the way, and a tiny reduction in the delta-V you need due to the plane's own velocity, does this do any good?

...

So yes, I was aware that there are air-launched systems, but unless the planes can reach serious multiples of the speed of sound, I can't really see why!

Gotcha. That's why I asked. And I very much doubt the blended wing body designs NASA was throwing around were ever going to go supersonic.

On the other hand, I would be really thrilled and grateful to be able to see these proposals!

I'm especially interested to know--does an airplane capable of lifting and dropping a Saturn V have to weigh as much as its payload does? That's a rule of thumb I've been going by. I'd like to see if I'm right to do so. (Rule of thumb being, useful lift tends to be about half the total lift capacity of any aircraft, fixed structure typically being about half the capacity.)

See my post above. The proposals went up to 10,000-ton takeoff mass, although I've only found detailed designs for up to 4,000 tons.

If I recall correctly, the 10,000 tons is about 4,000 tons of payload and 6,000 tons of plane, so you could theoretically fit a Saturn-V and 500 tons of launch cradle on board. Using this for air launch was something I thought of; NASA wanted to use them as giant commercial transports.

You could get a better ratio using a design specifically optimized for this - in particular, if you want to go totally atompunk, you could leave out the chemical takeoff fuel and use a shadow shield rather than full-360 shielding. But, speaking as a total nonexpert, it seems unlikely you could get as low as the mass ratios you talked about above for chemical rockets.

 

[Workable Goblin]

Shevek, you forgot about the possibility of throttling That 1-G initial NTR doesn't have to thrust at the same level through the burn--it can throttle down towards the end, just the same way the Shuttle's engines do.

Since the idea of a carrier plane acting as the first stage of a launch system has come up a number of times in proposals, and actually been implemented IOTL (with the Pegasus), I would be inclined to say that it has some value. One that immediately comes to mind is basically being able to choose your launch vector from the start--unlike a fixed pad, you're not limited to being able to launch in one direction, and having to do maneuvers with your rocket after ignition to go somewhere else (that's where the pitch and roll maneuver comes from). Especially if you were basing, say, out in the Central Pacific, so you had a huge range and could launch in whichever direction you liked. Similarly, you could launch from *wherever* you liked in that range, so weather wouldn't be as big a deal as it is with fixed pads--if the weather's bad, go somewhere else.

Another is the possibility of using (relatively) ordinary runways and facilities for your vehicle, rather than needing specialized, expensive facilities. This is particularly obvious with the Pegasus, again; theoretically, you could probably launch from just about any airport that could accommodate an L-1011. That seems to be a lot of the reason for HTHL systems, which naturally need to include aircraft-type stages.

A third is the possibility of reusing the same plane again and again. Sure, in your example a chemical stage might be cheaper to boost second stages above the troposphere for any *individual* launch, but a plane's cost can be amortized over possibly dozens or hundreds of missions, whereas a new booster need to be manufactured each time for the chemical scenario. Since aircraft are actually reusable, too, this isn't just a theoretical savings like with the Space Shuttle.

 

[Shevek23]

I've managed to see your links for the various nuclear aircraft studies.

You know a Saturn V as used for Apollo massed 3000 metric tons, right? An airplane of 1000 tons (even!) is not going to cut it! You'd need to go up nearly an order of magnitude more, to 6000 tons, or the mass of 20 Boeing 747s.

Can it be done? I suppose it could. And by the way, nuclear power would be a separate issue--obviously if we could come up with nuclear engines for a 6000 ton airplane we could come up with fuel-burning turbofans instead. For this mission, given that discussion elsewhere has established fuel cost is a small component anyway, and given that the mission is to take off, climb to ceiling and maximum speed, drop or toss the rocket, and if one escapes the exhaust blast, immediately return to base--fuel does not have to be anything like a full load for long-range cruising since the plane, though admittedly putting its engines through a high-demand wringer, has a very brief flight profile, of takeoff, climb, quick drop, descend and land. So a conventional fueling of the plane even with a margin for safety would be light and allow more lift available for payload, meaning we can design a somewhat smaller hence cheaper plane. So for this particular mission we'd want the chemical version even if nuke versions become standard, since most of the weight of the fission system would be dead weight in this case!

Anyway it still isn't clear to me the Saturn V would be better off being tossed at subsonic speed in the stratosphere. Well clearly there would be some margin of improvement for the payload, but would it be dramatic? (It might be, since payloads are such tiny fractions of the all up mass, so a single percent improvement could translate into very large percentages of the payload). Another advantage I can think of--Apollo, and Shuttle, indeed all rocket launches, are vulnerable to bad weather at launch, and the big rockets have the additional liability that they need to be slowly hauled from the assembly building to a launch pad, vertical. (At least that's the way Americans do it; the Soviet system is generally to design the rockets to lie flat and only tip them up vertical at the pad, which has some penalties but avoids the problem I am about to mention, and simplifies assembly.) So vertical rockets on a moving crawler are vulnerable to winds in particular, and so Saturn launches involved major efforts of weather forecasting to predict whether there was any risk an excessive wind might blow up during the many days it took for the rocket to make this trip.

Well obviously a version of the Saturn V that could be lifted up and dropped from a really big airplane would have to take horizontal storage. For that matter it would have to take other stresses like sudden surges of acceleration. Perhaps the fuel shouldn't be loaded in until just before launch (that was the practice anyway I think) meaning the plane needs to have fuel tanks for everything, with reserves to allow for evaporation. These tanks of course could be designed to contain the fuel surges using methods that would be costly on the rocket, then the plane starts loading in fuel once it has climbed to cruising height and is straight and level. Depending on how long this takes the mission could be a lot longer than I guessed, maybe we need the nukes after all. (Between committing to the nuke plant which probably takes up a third of the empty weight, corresponding to the quarter of fully loaded weight fuel takes on a typical airplane that maximizes range times payload, and various auxiliary functions like the fuel tanks with their special provisions, I'm thinking now the airplane takes off at more like 12,000 tons! We've got a flying cruiser here...) But, at least such a plane can probably take off in a much greater range of weather conditions than the rocket could risk being out in, and loading the rocket onto the plane would probably be much quicker than hauling it out to the pad so the weather needs to be considered only over a much shorter, much more predictable period. Once airborne a typical modern high-speed jet flies at altitudes where weather is much less of a concern, and if the flight computers are up to it, can launch from a wide range of locations thus dodging the worst of it. So that's a major operational consideration.

Still the modifications to make the rocket robust enough to stand being stressed by typical takeoff and climb maneuvers, not to mention any fancy footwork unexpected winds might toss the pilot's way, and of course the actual launch maneuver (where the rocket goes from being an aerodynamically faired horizontal cargo to a more or less vertical angle and gets tossed aside) might well offset the advantages we can expect from a subsonic launch and an atmosphere much closer to deep space conditions. (One way we might have taken advantage of launching from high up might have been to make the rocket assembly shorter and broader, since punching through air drag is much less of a problem now, but that would tend to make it harder to fit the squat version into the fuselage of an airplane!)

I've long held my own wacky notion that NASA should have developed a high-altitude aerostat launcher, a big fat lenticular airship. It would have to be huge of course, and the lift for the rocket itself had better not come from helium, because venting it would be incredibly expensive. (Oh well, that's actually par for the NASA course). I wouldn't want it to be hydrogen so that leaves steam. The idea would be, haul the aerostat over the VAB, lower a bunch of lines, and haul up a gantry to which the rocket is secured. Leaving this assembly hanging below the aerostat, rise under static lift to the desired altitude--the higher we go, the bigger the balloon needs to be of course. As with the plane, we might have the fuel tanks in the aerostat rather than in the rocket until we get pretty high. Here we can certainly go with a shorter, squat rocket if we are rising to thin air! We fire the rocket when it is hanging below; as the thrust rises to the point the weight is neutralized and it starts to rise, we rip open seams holding the central body of the airship's upper and lower surfaces; the ones below fall down (as does the gantry which falls away to the side, maybe splitting into two or three pieces for symmetry) the ones above are lifted both by the escaping lift gas (steam, or expensive irreplacable helium) aided by some balloons attached to the panels; suddenly the lift that recently lifted the rocket is gone, freely rising ahead of the rocket and diffusing (or in the case of steam, condensing) into the atmosphere. The rocket has a clear path above it to rise right through the now-toroidal airship, whose fixed mass is lifted with peripheral gas bags.

The rocket is obviously rising in zero wind at launch, since the aerostat would have been drifting with any wind there was; the drift velocity is carefully measured and factored into the launch computers. The geometry of the situation is such that I believe all parts of the aerostat are distant enough from the exhaust flame. It rises through and past the steam layer and into clear undisturbed atmosphere and onward, reaching a maximum pressure that is lowered by whatever degree the rise of the aerostat lowered it. The aerostat has monitoring equipment that closely watches the rocket launch until it has gone some distance and then it vents excess gas and descends to be retrieved and moored for maintenance.

A variation on this theme, if we consider the notion of a free-flying aerostat too extreme, is to use one simply to lift most of the weight of the rocket to the launch pad. If we bring the simplified version over to the VAB as before and suspend the gantry, using a much lighter and more agile crawler merely as a stabilizing weight, and haul both aerostat and rocket out on lines, we can probably proceed much more quickly; if the rocket is disturbed, it is now a hanging pendulum instead of an unstable pencil balanced on a finger so we can risk much faster movement. Once the gantry is secured on the launch pad we can then haul the aerostat away again. In this way we could greatly accelerate pad operations and better avoid inclement weather.

 

[Shevek23]

Shevek, you forgot about the possibility of throttling That 1-G initial NTR doesn't have to thrust at the same level through the burn--it can throttle down towards the end, just the same way the Shuttle's engines do...

No, I was responding to the idea that air launch could somehow overcome the liability of NTR rockets someone brought up earlier, that they have poor thrust-to-weight ratios. If we are finding their thrust so low relative to the weight of the engine itself that gravity losses are eating up the benefit of higher specific impulse, we can't very well afford to be throttling them down, can we? If we could install enough thrust at the beginning to move it well above 1 G, then that problem you brought up seems to be a nonissue, especially if the engines are recoverable (and I'd think nuclear engines had better be!)

But throttling back later in the boost also will stretch out the time gravity is fighting us and thus demand more delta-V overall. It's inevitable of course if you have mass ratios like 7 to 1 and don't want to subject the astronauts or other delicate payload to 10 G's at the end of the boost.

And that, to get back to the OP, is another reason we use staged rockets--the upper stage engines have lower thrust than the lower stage ones, so even without throttling the average boost stays in the same ballpark.

Also, isn't throttling of rockets a rather tricky operation? Giving the SSMEs that capability was part of what made designing the things a long-drawn-out process. And even so they can't throttle over a full range--something like 105 percent down to a third, and they can't go below that without shutting down.

Another option is to have a cluster of many rockets and shut some of them down while others continue to fire; that, combined with a rather narrow range of throttling, will give a smoother ride.

But obviously the easiest way to shut down an engine is to run its tank dry and then drop it and its tank--in other words, staged operation.

I guess by the time I post this you can see where we agree on the virtues of airplane launch; I certainly think it's good to be able to choose vector and to jink around the launch operations zone to find the good weather!

But again if we are talking moon mission scales of spaceship, as opposed to itty bitty designer satellites suitable for something as "small" as the Pegasus (which IIRC uses big jetliners for its launch), we are obviously talking about really gigantic airplanes. Which I'm obviously willing to talk about forever! But they aren't going to be operating from any run of the mill international airport!

I haven't had a chance to actually read the pdfs yet; it took a while to download them. (I typically have my browser way overloaded so the fact that they crashed it is not that significant. But they did.) Do they really talk about vertical takeoff for a 1000 ton jet?!? Seaplane, that's obviously the way to go with such monsters. Or their really big brother that could really lift up a Saturn V!

 

[asnys]

You know a Saturn V as used for Apollo massed 3000 metric tons, right? An airplane of 1000 tons (even!) is not going to cut it! You'd need to go up nearly an order of magnitude more, to 6000 tons, or the mass of 20 Boeing 747s.

I know. There are discussions of 10,000-ton aircraft in there, although they focus on "lighter" aircraft - only 1,000 tons! - for obvious reasons.

Do they really talk about vertical takeoff for a 1000 ton jet?!? Seaplane, that's obviously the way to go with such monsters. Or their really big brother that could really lift up a Saturn V!

Yep, they really were talking about VTOL, I guess so they could serve non-coastal areas. Lockheed had one design, part of a separate program, that massed 5,500 tons and used, IIRC, 180 turbojets for takeoff and "flight maneuvering." Seriously.

Can it be done? I suppose it could. And by the way, nuclear power would be a separate issue--obviously if we could come up with nuclear engines for a 6000 ton airplane we could come up with fuel-burning turbofans instead. For this mission, given that discussion elsewhere has established fuel cost is a small component anyway, and given that the mission is to take off, climb to ceiling and maximum speed, drop or toss the rocket, and if one escapes the exhaust blast, immediately return to base--fuel does not have to be anything like a full load for long-range cruising since the plane, though admittedly putting its engines through a high-demand wringer, has a very brief flight profile, of takeoff, climb, quick drop, descend and land. So a conventional fueling of the plane even with a margin for safety would be light and allow more lift available for payload, meaning we can design a somewhat smaller hence cheaper plane. So for this particular mission we'd want the chemical version even if nuke versions become standard, since most of the weight of the fission system would be dead weight in this case!

Fair enough. My original thinking was that nuclear power would be a way to get these giant aircraft in the first place. Since the shielding mass dominates the mass of the engine, and scales sublinearly to the power of the reactor, your P/W ratio improves as the mass of the plane increases. In theory, there's a point somewhere between 1 and 5 million pounds where the nuclear airplane outperforms the chemical airplane economically. The exact point depends on mission range, reactor power density, insurance and interest charges, how much chemical fuel you carry for takeoffs/landings and emergencies, and who's doing the calculation.

Of course, there's no reason we have to stick a reactor in the airframe once we have it, but the nuclear engines could justify building the airframes in the first place.

 

[Shevek23]

If you weren't going SSTO, and just wanted reuseable, would designing the intermittent stages to be piloted and landable be acceptable?

I.e. a Saturn 5m is launched (m for manned). The first stage is jettisoned, and the pilot controls its descent back to earth, so it lands near the launch site. Ditto for 2nd and third stages. This means it will have to launch vertically rather than horizontally, but it also means it can be launched from anywhere in the world without worrying about who is east.

This will require additional mass per stage for the cockpit, landing equipment, aso, but it means the Saturn 5m can be launched over and over. The engines and tanks will have to be designed to be reused rather than expended, but it gives you a 50-ton LEO craft (original Saturn was 119 tons, I am halving it to account for the other stuff).

The other stuff being either parachutes, which means you have to chasing the components and fishing them out of the ocean, which is how the Shuttle's solid fuel boosters were supposed to be renewable; or wings so the lower stages can actually be flown back and landed.

That was an idea that was considered for the Saturn series. (And winged V-2s were tried by the Germans, as a means of giving the missiles more range). Obviously wings mean not just more weight but more drag at launch, but yes, the idea was to fly them back to base and refit the stages for reuse. I don't know if the lower stages were meant to be manned, I doubt it very much. I'd think it would be sufficient to remote control them. Though I suppose remote control landings are tricky. I thought the plan was to fly them to a water range very near the base (Kennedy and Vandenberg of course are both coastal) and recover them that way.

Of course a spent rocket stage is not the most aerodynamic thing; it's the wrong proportions and it has a messed up nose. And tail for that matter; rockets designed for atmospheric operations include the effect of the supersonic exhaust gas trail in their figuring of drag; the exhaust is a virtual part of the structure as far as air flow is concerned. Once it's shut off, the tail is no longer aerodynamic at all. So you're basically putting wings on a brick (well, a wad of tinfoil, an empty rocket stage isn't very heavy for its bulk). This is why the simpler parachute option might be more attractive, especially because with a Rogallo type hang-glider sort of parachute one has some hope of guiding it to a desirable crash-landing spot of one's choosing. More or less, depending on winds!

I think the point of fixed wings actually was that even a lower stage will burn out pretty high and going pretty fast, so it has a long distance in thin air to cover to get back to the base. But not so long given the poor aerodynamics.

The first stage should have been fairly easy, as its burnout speed would not be that great, so its structure probably would not suffer too much from free-falling back toward sea level. Not so sure about the second stage though; it should have been burning out at very high altitudes and very high Mach speeds, and to recover it would have to operate at high temperatures. This is not such a problem on the way up because the air is thin by the time these speeds are reached but coming down it will speed up as it falls, except insofar as the air drag slows it down, which it does by heating it up!

So we have recoverable first stage, a second stage that is harder to make recoverable and doing so probably means increasing its weight and cost a lot, and still it will take a beating and not be reusable as often as the first. And what of the third? Its goal is to get the payload into orbit or on its way to the moon, and to reuse it means either a reentry from orbital (or effectively interplanetary, for a moon orbit) speeds, bearing in mind the orbit it is in when it burns out might not return at all or come in in a most inconvenient way so it has to be reoriented, or to leave it in space and reuse it as either structural material or, with refueling (from what?) as some kind of orbital tug.

The manned recovery module can be designed for reuse of course, but on a typical Apollo mission one was leaving both pieces of the Lunar module and the Service Module to drift, crash, or burn up somewhere. I believe all SMs wound up burning up in the atmosphere again because their final mission was to position the Command Module for reentry, which meant they were on the same collision course with the atmosphere. To enhance that recoverability, one would have to, at some considerable cost, extend the heat shielding of the CM to include the SM, engine nozzle and all--or I suppose one could eject that and replace it, along with other bits of bric-a-brac.

 

[Workable Goblin]

Also, isn't throttling of rockets a rather tricky operation? Giving the SSMEs that capability was part of what made designing the things a long-drawn-out process. And even so they can't throttle over a full range--something like 105 percent down to a third, and they can't go below that without shutting down.

It's not easy, but then nothing is, really, with rockets. The SSME had problems because they were basically trying to do 6 or 7 things at once--they wanted a reusable (never before done), high pressure (never before done), staged-combustion hydrolox (never before done), throttable engine which would have good performance from sea-level to vacuum (never before done), including a good ISP (quite a bit better than the J-2's, but not so good as the RL-10). It's not particularly surprising that when they're trying to do that many things that have never been done before simultaneously, especially when some of them conflict (reusability and high-pressure in particular) that they would have trouble. But in the greater scheme of things, the SSMEs actually weren't too much trouble--they only took about as long to develop as the F-1s, for instance, and those pushed their limits much less in many respects. Of course, they have also proved very reliable in flight, so clearly they did that right.

Throttle range for the SSME is 109% (emergency/abort power levels, increases engine failure rate considerably) to something like 66% or so. They probably could redesign it to throttle lower, but that isn't and wasn't necessary for the types of mission the Shuttle can fly. RL-10s were modified in the early 1990s (in connection with SEI and other Moon-landing related things) to throttle from something like 10% to 100% or maybe more of rated thrust. So throttling is just fine.

But throttling back later in the boost also will stretch out the time gravity is fighting us and thus demand more delta-V overall. It's inevitable of course if you have mass ratios like 7 to 1 and don't want to subject the astronauts or other delicate payload to 10 G's at the end of the boost.

The end of the boost would be the part where you're going horizontally and gravity losses aren't an issue anyways. The problem comes early in the boost when you're trying to get above the sensible atmosphere so you can tip over and focus totally on speeding up. Throttling down at the end is fine, especially if (as seems likely) such a high T/W ratio NTR would be LOX-augmented, so doing so would be both somewhat less tricky than with a pure NTR vehicle and necessary for fuel economy in any case. (Of course, a highly-augmented NTR has an ISP no higher than a good hydrolox engine like the RD-0120 or the SSME, anyways, which begs the question of why, exactly, you're bothering with the "N" part)

 

[e of pi]

It's not easy, but then nothing is, really, with rockets. The SSME had problems because they were basically trying to do 6 or 7 things at once--they wanted a reusable (never before done), high pressure (never before done), staged-combustion hydrolox (never before done), throttable engine which would have good performance from sea-level to vacuum (never before done), including a good ISP (quite a bit better than the J-2's, but not so good as the RL-10). It's not particularly surprising that when they're trying to do that many things that have never been done before simultaneously, especially when some of them conflict (reusability and high-pressure in particular) that they would have trouble. But in the greater scheme of things, the SSMEs actually weren't too much trouble--they only took about as long to develop as the F-1s, for instance, and those pushed their limits much less in many respects. Of course, they have also proved very reliable in flight, so clearly they did that right.

However, it does sound like a lot on the face of it. If anyone here has read Eyes Turned Skyward's first post, this boundary-pushing is the reason that commentators in our ATL are so skeptical about the SSME.

RL-10s were modified in the early 1990s (in connection with SEI and other Moon-landing related things) to throttle from something like 10% to 100% or maybe more of rated thrust. So throttling is just fine.

Don't know about those experiments, but I do know about the CECE (Common extensible cryogenic engine), which was similar for the Altair lander and is notable for being able to throttle from 104% to merely 8%.

 

[e of pi]

The manned recovery module can be designed for reuse of course, but on a typical Apollo mission one was leaving both pieces of the Lunar module and the Service Module to drift, crash, or burn up somewhere. I believe all SMs wound up burning up in the atmosphere again because their final mission was to position the Command Module for reentry, which meant they were on the same collision course with the atmosphere. To enhance that recoverability, one would have to, at some considerable cost, extend the heat shielding of the CM to include the SM, engine nozzle and all--or I suppose one could eject that and replace it, along with other bits of bric-a-brac.

SpaceX's Dragon is essentially a reusable recovery module, and the only major systems not inside the Dragon capsule proper are the solar arrays and cargo fittings in the trunk--everything else comes home. Now doing that, but also finding a place for a big orbital maneuvering engine like the one on Apollo or Orion...that's trickier, especially if you want that engine and its tanks to come home too.

 

[Workable Goblin]

Don't know about those experiments, but I do know about the CECE (Common extensible cryogenic engine), which was similar for the Altair lander and is notable for being able to throttle from 104% to merely 8%.

I was misremembering based on some comments Zubrin put in Case for Mars about requirements for SEI lunar landers--in actuality, there were some experiments done in relation to the DC-X which showed throttling as low as 30%, which is still pretty low.

 

[Shevek23]

OK, e of pi, Truth is Life, Asnys--this mafia of yours has been drawing me over to the Dark Side of thinking seriously about nuclear rockets!

So poking around the Net some more I've found references to Project Timberwind (part of SDI in the 1980s, to provide the heavy lift for the systems envisioned--and the Teller-Strangelovitis just keeps getting creepier)

Ahem. Anyway, these things were as far as I know entirely on paper. But say they could actually work, and Come Over to the Dark Side and forget about worrying like a Nervous Nelly who'd have been out campaigning for Jesse Jackson at the time (ahem, Yo!) about what happens if the damn things crash...

Thrust to weight ratios in the 30:1 ratio, ISP 1000 in vacuum, and nearly that at sea level to boot. Can this yield a Single Stage to Orbit?

Even among those gung-ho types at SDI, that wasn't exactly the plan. Apparently the Timberwind engines were to be used in "Timberwind Titan," and were meant to be disposable eek I guess because the very intense power releases involved used up the fissionable material very fast--which is good in a way, because that would allow a lighter reactor. (Perhaps they skimped on shielding too, because this was meant for unmanned lift and with a typical staged rocket arrangement the engines are down at the bottom and the payload is up on top, so distance and the intervening mass of fuel (well, until you've used up a lot of it anyway) help cut down on the radiation.

But let's just guess the engines are man-rated, and adequately shielded, and even if the given burn times pretty much expend the reaction and after that you've got a mass of nuclear waste, essentially--having done some back of envelope work, I believe that you could use them to make something like a Shuttle but massing around 500 tons on the launch pad--which is a quarter the all-up mass of the STS.

I assumed throttled (yet another wild guess, it isn't clear that one can throttle a Timberwind--you'd have to simultaneously throttle both the reactor and the nozzle) thrust tailored to maintain a 3 g thrust throughout, and a profile that would first lift up at effectively 2 Gs acceleration straight up, essentially, to reach a target speed of about 1620 meters/sec (mostly vertical, with about 500 m/sec speed downrange toward the eventual orbital vector) over 80 seconds, rising thus to around 60 km altitude, then switch over to a steady 3 g burn completely in the tangential, orbit-seeking direction for about 240 seconds--I figured that would leave it precisely in circular orbit at 200 km altitude at burnout.

There's plenty of time in the burn times given to do the job.

But first of all-given the somewhat lowered thrust at sea level (not really that drastically lowered though, a matter of 13/15 or so, not the pathetic reduction a NERVA would suffer!) and that we are spending 80 seconds fighting gravity head-on, we'd need something like 24 of the bigger Timberwind 75s. But far fewer of them as we approach final orbital insertion.

Second, we'd need a huge fuel tank. One would think that with the fuel massing something like half of the mass of what fills a Shuttle fuel tank (that's around 720 tons) we could get a more compact tank, but actually, now all of our reaction mass is hydrogen--with the STS, most of the reaction mass, about 6/7, was oxygen, so actually a tank big enough to hold 350 or so tons of hydrogen only would be bigger than the Shuttle's.

We've got rid of the solid fuel boosters to be sure. But trying to enclose that vast tank volume in a single hull with the orbiter's essential structure and all 24 Timberwind engines (I draw the line at disposing of them intentionally! Even if they are spent, we can't seriously intend to burn them up in the atmosphere, we probably can reprocess them for quite a lot of fuel, and we have to dispose of the rest of the waste responsibly. So decrees the Space Hippie!) in a heat-resistant coating of some kind and flying the whole thing aerodynamically in--well, the sheer emptiness of the volume helps in the sense that it's less dense than the OTL Orbiter so aerobraking is that much more effective, but structurally such a big hollow volume is going to weigh more than a compact one.

So we are quickly led in a different direction, away from strict "single-stage" toward something much more like the STS we know. We should obviously have a separate fuel tank for most of the fuel, expand the orbiter only for a small amount of it, put half the engines on the bottom of the fuel tank--indeed, something like 1/3 of them there, 1/3 built in to the Orbiter, 1/3 as parallel boosters we drop (but recover!) much earlier in the burn. We have to recover the engines on the tank, so we might consider trying to recover the tank as a whole separately.

But we'd still want at least 2 stages! They might all be fully recoverable (well, the nuke engines might need to be replaced, but we'd surely recover a lot of useful material from each one).

But even with this quite fantastic sort of nuclear engines, the logic of staging is still overwhelming.

Apparently for single stage, either we have to use the atmosphere for propellant, which involves its own inherent nightmares, or have something even better by far than these Strangelovian Timberwind engines. Something, I guess, like fusion engines. (Good ones I mean--lightweight. Quite fantastic!)

Or an Orion. (The Taylor-Dyson kind, not the cancelled recent NASA kind).

Of course it was searching the Net for Orion-related stuff that disclosed AH to me in the first place, in the form of the naysaying essay I'm told was written by Ian himself...

So there we are. Stages, learn to love them.

 

[e of pi]

Shevek,

I actually find your inclusion of me in your "nuclear mafia" strange, given personally I don't like nuclear rockets much, except maybe NEP (nuclear-electric propulsion) and even then the mass advantages over solar aren't critical, it's more the freedom from insolation variation over the solar system.

As you point out, even a nuclear rocket isn't enough to make a reusable SSTO practical, and a non-reusable SSTO is just code for "bad ISP and horrific accelerations at the end of the burn". To make an RLV really practical, I think two stages is really optimal--one built for sea-level and relatively low-speed recovery via boost-back, the other orbital and optimized nearly for vacuum. Essentially, the first stage just serves to give the upper stage enough of a push that it can act like an optimal SSTO RLV.

Personally, I'm currently a big fan of depot and EOR-based schemes, and my impressions of working with Truth is Life on ETS stuff is that he roughly agrees. If Skylan works, then maybe a SSTO RLV is practical, but we'll see in the next few years if their practice can keep up with their theory. I hope it does, it's a cool concept, but...we'll see.

 

[Shevek23]

Shevek,

I actually find your inclusion of me in your "nuclear mafia" strange, given personally I don't like nuclear rockets much, except maybe NEP (nuclear-electric propulsion) and even then the mass advantages over solar aren't critical, it's more the freedom from insolation variation over the solar system.

Sorry, I guess the way y'all typically pop up on the same threads saying generally similar things has tended to blend everyone together in my head. Come to think of it I don't recall anyone thumping hard for any sort of nuclear launching system.

Does anyone know if engines of Timberwind type capabilities (ISP over 900, thrust/weight in 1 G gravity as great as 30, minimal deterioration of thrust efficiency in sea-level atmosphere) even are on the margins of possibility?

Not that I actually want them!

{dark side whispers seductively...}

 

[Ravaun]

Reality Check

Sorry, I guess the way y'all typically pop up on the same threads saying generally similar things has tended to blend everyone together in my head. Come to think of it I don't recall anyone thumping hard for any sort of nuclear launching system.

Does anyone know if engines of Timberwind type capabilities (ISP over 900, thrust/weight in 1 G gravity as great as 30, minimal deterioration of thrust efficiency in sea-level atmosphere) even are on the margins of possibility?

Not that I actually want them!

{dark side whispers seductively...}

The Timberwind engine WAS tested. You will not find any document verifying that...until about 2030 or 2040. They were used to launch spy sats into orbit. To watch OUTWARD not inward. By that point in time (the early '90s) the USSR had effectively collapsed, leaving nothing but external threats.

It requires 12.5 tons of shielding to man-rate a Timberwind 250 engine.


SSTO: The USAF worked on a study (Chrysler based) during the Shuttle era. NASA kept telling Chrysler NO, then ended up giving them another study contract. They did everything but model-test, based on currently released info. It was called (Project) SERV (Single-stage Earth-orbital Reusable Vehicle). In general shape it looked like the world's largest Apollo capsule. Saturn II tooling was used for the 4 propellant tanks. It had 12 - 16 engine "modules," based on the SSME (Space Shuttle Main Engine) and 20 - 40 air-breathing, turbojet engines (for initial lift-assist to take-off, obviating the need for more than a flat concrete landing pad and at around < 3,200 m/sec, as re-entry heating decreased enough to re-open the jet doors). The documents: "SERV Feasability Study" and "Phase A Space Shuttle Study" (Contract NAS8-26341, MSFC-DRL-214, DRD MA-077-U3, July 1, 1971) are 195 and 128 pages long. While well-written, they're filled with tech-speak, though not as bad as the TSTO HLLV and MLLV docs generated by Boeing during the same time period.

The USAF continued studies in this vein. Their existence has (since about '97, as I recall) been partially declassed. That study was called the VTOVL and was a North American Aerospace project. For something to be classified for 25+ years as NAA VTOVL was, that generally means it was actually BUILT. The Timberwind Project was classified until 2003/2005 and only sparse info. is avail. on either program. IAW the NRC (Nuclear Reg. Commission), the only reason something would be only partly declassed when in involves nuclear reactors/engines/weapons is when it's actually BEEN BUILT. <-- That's a paraphrase from an actual NRC released doc. re: yet another program, the NAA Manned Bombardment and Control Vehicle, a TSTO using a sled-launched astro-plane (orig. design) which was redesigned to a balloon-launch vehicle equipped with 4 drop-style nuclear weapons, being phased out of the battlefield at the time of their deployment. It's widely believed this led to the collapse of the USSR as they KNEW the U.S. had the higher ground.

http://www.astronautix.com/craft/naahicle.htm


I've spent the past 8 - 12 years researching in this particular area.


Best poss. option for a TSTO would be a Saturn V-class (Saturn I-C) first stage, using the LOX/Propane engine developed (and tested) to a fraction of the 2 Million lbs-force required for the baseline NASA Solar Power Satellite launch shuttle. It's LH2-cooled (but not burning) and requires no maintenance, aside from de-coking on return from orbit. Launched vertically, landed horizontally. Would've required a landing field about 1.5 - 2x the length of the Shuttle runway at KSC.

Second best poss. option for a TSTO would be an Air-Induced, Hybrid solid rocket. The hybrids people are looking at things all wrong, just trying to figure out how to STORE the fuel, prior to launch. Instead, they need to gather it up, chill it down--liquify it (the chill down/liquification process is designed to auto-sort for LOX vs. other gasses, which are bypassed just as in a high-bypass jet). The estimated max. poss. ISP for this design, using LOX/Aluminum Powder-Paraffin Wax, is 360 sec (vacuum).

The GE CF6-50 is well-suited for pulling in atmo, in the lower layers. A MIST-based/SNAP-50 reactor would handle fractionation. Hydraulics, Avionics, Comm and Lights (including exterior structure) based on the Boeing 2707 design, with modern updates. Total mass for a 260" (inch) diam. Full Length Air-Induced Hybrid Al Powder-Paraffin rocket: 143,656.3346 kg. On-board propellant mass of 1,446,651.41 kg. Total propellant: 5,641,940.5 kg.

Est. remaining (unburned) propellant: 87,450 kg. Est. I(sp): 295.586 | 326.63496518 sec* optimal. As with most engines, expect up to a 2 - 3 sec variance between design and actual execution.

** The above design is ocean-recoverable **


A good (interim) TSTO booster would be similar, using a 10% Niobium + 100% 18Nickel2400 maraging steel (restricted transport due to nuc. non-proliferation treaties) canister with up to a 30% Niobium nozzle. Propellant mix would be Trinitramide/Al Powder-Paraffin.
Stats:
[1,765,348.3] + 131,960.3453 kg I(sp): 243.13669 | 268.676259 sec
(27,362.9 kg unburned)

** The above design is ocean-recoverable **


If you want an SSTO design and you don't care about the environment (e.g., you could prob. use in on the Moon or Mars, no where on Earth) a Dumbo-nuclear engine design would do the trick. They're a bit heavier than the SSME but have an est. performance in the 960 sec range. These were proposed same time as NERVA but the more conventional designe (NERVA) got the funding. According to OTL, the Dumbo Nuclear Rocket then quietly fell out of history, apparently never spoken of again...

 

[Kevin Renner]

Haven't heard Dumbo spoken of in a long time*. IMO one idea I would likke to see tried is Black Horse. Take off from a runway with a light fuel load,meet a tanker and light the candle. Lower weight at take-off means less massive landing gear which are a big problem with horizontal take-off SSTOs. But theory and studies are one thing. Until somebody cuts metal or carbon fiber and starts flying it is all moot

 

[Emperor Norton I]

Babies born during the last post have become full blown toddlers. It's been years. Naughty, naughty.

 

[CalBear]

The Timberwind engine WAS tested. You will not find any document verifying that...until about 2030 or 2040. They were used to launch spy sats into orbit. To watch OUTWARD not inward. By that point in time (the early '90s) the USSR had effectively collapsed, leaving nothing but external threats.

It requires 12.5 tons of shielding to man-rate a Timberwind 250 engine.


SSTO: The USAF worked on a study (Chrysler based) during the Shuttle era. NASA kept telling Chrysler NO, then ended up giving them another study contract. They did everything but model-test, based on currently released info. It was called (Project) SERV (Single-stage Earth-orbital Reusable Vehicle). In general shape it looked like the world's largest Apollo capsule. Saturn II tooling was used for the 4 propellant tanks. It had 12 - 16 engine "modules," based on the SSME (Space Shuttle Main Engine) and 20 - 40 air-breathing, turbojet engines (for initial lift-assist to take-off, obviating the need for more than a flat concrete landing pad and at around < 3,200 m/sec, as re-entry heating decreased enough to re-open the jet doors). The documents: "SERV Feasability Study" and "Phase A Space Shuttle Study" (Contract NAS8-26341, MSFC-DRL-214, DRD MA-077-U3, July 1, 1971) are 195 and 128 pages long. While well-written, they're filled with tech-speak, though not as bad as the TSTO HLLV and MLLV docs generated by Boeing during the same time period.

The USAF continued studies in this vein. Their existence has (since about '97, as I recall) been partially declassed. That study was called the VTOVL and was a North American Aerospace project. For something to be classified for 25+ years as NAA VTOVL was, that generally means it was actually BUILT. The Timberwind Project was classified until 2003/2005 and only sparse info. is avail. on either program. IAW the NRC (Nuclear Reg. Commission), the only reason something would be only partly declassed when in involves nuclear reactors/engines/weapons is when it's actually BEEN BUILT. <-- That's a paraphrase from an actual NRC released doc. re: yet another program, the NAA Manned Bombardment and Control Vehicle, a TSTO using a sled-launched astro-plane (orig. design) which was redesigned to a balloon-launch vehicle equipped with 4 drop-style nuclear weapons, being phased out of the battlefield at the time of their deployment. It's widely believed this led to the collapse of the USSR as they KNEW the U.S. had the higher ground.

http://www.astronautix.com/craft/naahicle.htm


I've spent the past 8 - 12 years researching in this particular area.


Best poss. option for a TSTO would be a Saturn V-class (Saturn I-C) first stage, using the LOX/Propane engine developed (and tested) to a fraction of the 2 Million lbs-force required for the baseline NASA Solar Power Satellite launch shuttle. It's LH2-cooled (but not burning) and requires no maintenance, aside from de-coking on return from orbit. Launched vertically, landed horizontally. Would've required a landing field about 1.5 - 2x the length of the Shuttle runway at KSC.

Second best poss. option for a TSTO would be an Air-Induced, Hybrid solid rocket. The hybrids people are looking at things all wrong, just trying to figure out how to STORE the fuel, prior to launch. Instead, they need to gather it up, chill it down--liquify it (the chill down/liquification process is designed to auto-sort for LOX vs. other gasses, which are bypassed just as in a high-bypass jet). The estimated max. poss. ISP for this design, using LOX/Aluminum Powder-Paraffin Wax, is 360 sec (vacuum).

The GE CF6-50 is well-suited for pulling in atmo, in the lower layers. A MIST-based/SNAP-50 reactor would handle fractionation. Hydraulics, Avionics, Comm and Lights (including exterior structure) based on the Boeing 2707 design, with modern updates. Total mass for a 260" (inch) diam. Full Length Air-Induced Hybrid Al Powder-Paraffin rocket: 143,656.3346 kg. On-board propellant mass of 1,446,651.41 kg. Total propellant: 5,641,940.5 kg.

Est. remaining (unburned) propellant: 87,450 kg. Est. I(sp): 295.586 | 326.63496518 sec* optimal. As with most engines, expect up to a 2 - 3 sec variance between design and actual execution.

** The above design is ocean-recoverable **


A good (interim) TSTO booster would be similar, using a 10% Niobium + 100% 18Nickel2400 maraging steel (restricted transport due to nuc. non-proliferation treaties) canister with up to a 30% Niobium nozzle. Propellant mix would be Trinitramide/Al Powder-Paraffin.
Stats:
[1,765,348.3] + 131,960.3453 kg I(sp): 243.13669 | 268.676259 sec
(27,362.9 kg unburned)

** The above design is ocean-recoverable **


If you want an SSTO design and you don't care about the environment (e.g., you could prob. use in on the Moon or Mars, no where on Earth) a Dumbo-nuclear engine design would do the trick. They're a bit heavier than the SSME but have an est. performance in the 960 sec range. These were proposed same time as NERVA but the more conventional designe (NERVA) got the funding. According to OTL, the Dumbo Nuclear Rocket then quietly fell out of history, apparently never spoken of again...

Well, thanks for stopping by.

You can exit via the door to your left. We are fixing the screen you got in through. Do let us know when the shuttle lands. Oh, and don't forget your tin foil hat.

Banned

We divorce you.