Reusable Rockets to Orbit

GW Johnson · 2013-08-27 10:23:22

Hi Bob!

I saw your post on this yesterday. I quite agree that SSTO is feasible, especially at around 5% inert-weight structural fractions. It's all in the price reduction for 1 stage vs the cost increase for 2-stage vs the rather-sharp unit price increase for the lower available payload fraction with 1 stage. In point of fact, that's what NASA is doing with its Senate-mandated SLS design. Theirs is a one-stage core (with some SRB's) that essentially pushes a huge mass to LEO. The upper stages they are looking at are mostly for departure from LEO. Or for GTO work. It's just that there are no commercial payloads yet that are big enough to need a rocket that size. The next commercial step is Spacex's Falcon-heavy at 53 metric tons to LEO out of Canaveral.

I don't think 5% inert fraction rockets will ever be practically reusable with the materials we have, despite what Spacex is attempting with its Grasshopper test vehicle and reusable-Falcon concept. To date, I know of no liquid-propellant stages that have ever survived the tumbling airloads on striking dense atmosphere supersonic/hypersonic. That's why Spacex has never yet recovered a Falcon-9 first stage in a salvageable condition: just too fragile not to break up. They did try, at least initially. And that doesn't even consider ocean impact loads, no matter how big the parachutes.

Whether Spacex can control the attitude of a Falcon stage for rear-end-forward thrusting recovery is problematical, and even they admit that. The combination of thrust and airloads entering like that is far beyond anything the lightweight stage was ever designed for in its ascent. That is one tough row to hoe. I wish them well, but I am not holding my breath about success. Neither is Musk.

I rather doubt that survivability can be obtained at 10% inert fractions. I think minimally-survivable (relatively short service life) inert fractions will look a lot more like 20+%, and that's using composites to-the-max everywhere the aeroheating won't damage them. Done all-metal, such inerts would fall in the 30-40% range. The longer the design service life, the more structural weight is required. The X-15 was right at 40%. Most high-speed bombers are around 50%, with Navy birds pushing 60%. These materials are all only so strong. You have to employ enough of them to take the punishment your mission environment dishes out. That's heavy, no way around it.

This situation is alien to most of the rocket designers out there today, SRB's excepted (and they just "lucked out", they didn't really know). No one has ever seriously tried to survive reentry with liquid booster stages before. The usual rules-of-thumb for rocket design don't apply, it's a completely different environment, and way far more challenging. The SRB guys sort-of survived parachute recovery at sea (a lot of these segments were too damaged to reuse) only because those were 900 psi pressure vessels, of inert weight fractions 10+%.

GW

Last edited by GW Johnson (2013-08-27 10:28:24)

GW Johnson · 2013-09-06 08:43:52

Has anybody considered hybridizing this reusability concept? A combination of reusable and throwaway components?

How about a reusable first stage core and strap-ons? Don't worry too much about recovering the second stage. Entry is less challenging at 3 km/s than it is at 7 km/s. That 3 km/s figure would be typical for the first stage of a TSTO vehicle, as they are currently designed. Use the strap-on concept to help make up for the extra inert weight in the reusable core, since that weight hurts you worst near takeoff.

Use a "pointy-ended" first stage core and a "Russian-style" truss interstage, cheap enough to throw away. The pointy end is the heat shield to protect the tankage, and coming back streamline avoids broadside crush load breakup. The engines at the rear are protected from airloads and entry heating better.

Stream out a series of chutes/ballutes of increasing size out the rear to slow to subsonic, then swap ends suddenly by clever chute redeployment, for a rocket-braked touchdown on land. Here on Earth with our dense atmosphere, one can use chutes like that to reduce the amount of rocket-braking propellant.

For launch sites like Canaveral, first stage touchdown will be at sea. I'd land by chute alone, pointy-end first. Better chance of not breaking up on ocean impact that way. No rocket-braking in that scenario. You can keep the truss interstage. If it survives entry at all, it can act as a sacrificial spike to help open the cavity into the water upon ocean impact.

Either way, the tankage is going to see some rather heavy airloads, and some of them will be lateral, as you swap ends. Even subsonically, that is tough to do. But we have a history of heat-shields, including hitting the sea successfully with them. Ocean impact loads are really enormous, too, even with really big clusters of really big chutes. Anything over about 30 mph sees loads not unlike those striking hard Earth. Ask any waterskier about that.

Just tossing out some ideas that seem feasible to me.

GW

Terraformer · 2013-09-06 11:14:46

I thought you wanted to avoid saltwater?

That was kinda where I was going with my comments about a hybrid first stage. Tops out at 3km/s, launches a disposable upper stage with a dV of 5-6km/s to put a small payload in orbit. Lands back at the launch site, checked over, refuelled, payload fitted, and relaunched.

Terraformer · 2013-09-11 09:15:56

I was thinking of something kind of like Black Colt, except bigger, with a lower top speed (3km/s rather than 4km/s), no aerial refuelling, and using a ramjet. Call it Rainbow Dash instead.

It really doesn't seem to be that difficult to make a reusable lower stage spaceplane for a TSTO system. I reckon it could be developed for under $200 million easily. For comparison, see the Falcon rocket ($300 million) and SpaceShipOne ($40 million). Such a system could dramatically lower the cost of launching small (<1 tonne) payloads into orbit. Add in automated on orbit docking, and you could launch some pretty decent probes and satellites. Maybe later we can up the capability to 2, or even 3, tonnes, so we can maybe use it for crew launch...

GW Johnson · 2013-09-12 08:27:36

Hmmmmm. There is a concept for HTO TSTO (two-stage orbital craft with a winged airplane for a first stage) that I haven't looked at before. In all the descriptions above, The staging occurs at M6 (1.8 km/s) at 20 km altitude with a ramjet airplane. It is beneficial to pull up to about 40 degrees at staging, which requires rocket thrust be added to the ramjet thrust.

New idea: go back on rocket thrust at that point and thrust to a bit higher staging speeds in thinner-to-no air a bit higher up. The penalty is a larger first stage airplane, because it has to hold more propellants, and more of its acceleration trajectory is powered at (lower) rocket Isp, which compounds how big it grows to be. The advantage is reducing the delta-vee required of the second stage below the 5.9 km/s in the prior discussions, letting such craft be more potentially reusable while having higher payload fractions (for better economics).

The ultimate limit on payload becomes more like that with an all-rocket TSTO airplane design (somewhat like the original concept for the shuttle, except HTO, not VTO). There is a practical limit to the size of the airplane we can build, because it starts to look like a water balloon supported on sharp nails, when parked on its landing gear waiting to launch. Square-cube scaling law at fixed material stress-strain limits. And wing-loadings are limited by practical takeoff and landing speeds. The bigger the wing the heavier, by roughly the 0.58 power of the wing area.

Unexplored trade study territory. Interesting. Hmmmmmmm, again.

GW

ps -- I'd resist going VTO with the winged first stage, because that way "traditional" rocket-launch huge logistical tail-type thinking gets into it, making it hugely expensive. Those government-type launch rocket guys never think like missile/weapons guys, because the missile/weapons guys are all thinking one-shot by their tradition. Spacex and ULA are the commercial exceptions that prove my point: lower costs with reduced logistical support. If it behaves more like a traditional airplane, then traditional airplane-type thinking goes into it, and that is aimed at extremely-low logistical support requirements. That means far lower costs.

Last edited by GW Johnson (2013-09-12 08:34:40)

Terraformer · 2013-09-12 10:48:24

A HTOL first stage also has rapid transport as a possible revenue source. Plus, if you can carry 10 tonnes of payload at 3km/s, you can rapidly carry a couple of dozen special forces guys, equipped with their own personal heat shields, to anywhere on the globe, ala project Hot Eagle. The Marines would probably buy several.

At the moment, I'm thinking of a system that can launch 1 tonne into LEO, using an expendable upper stage. Something that would compete with Falcon 1, but sold as a space launch system to interested organisations. If NASA wants to run their own fleet, then they can. If Britain wants to own one, it can. Let Uganda have one, if they can pony up the money. Obviously, the system would have to be developed and manufactured outside the United States...

Terraformer · 2013-09-12 14:31:27

Indeed, if we have a payload for the first stage of just enough for a Star 48 - so, about 2.2 tonnes - we could put about 150kg into LEO. Could we build a small demonstration craft capable of that, that would have a GLOM of about 25 tonnes? Call it the Scootaloo. Dry mass of 6 tonnes, say...

Terraformer · 2013-09-12 14:36:54

Or rather, another off the shelf stage that can actually provide enough delta-V...

JoshNH4H · 2013-09-12 21:27:24

3 Km/s is around Mach 7. I don't think there are too many aircraft flying at those speeds. You might get away with refitting a fighter jet with ramjet engines, but that seems like a pretty expensive approach.

Terraformer · 2013-09-13 03:21:32

What? Where did you get that idea from? I'm talking about building the vehicle new, because anything else would be too heavy.

JoshNH4H · 2013-09-13 08:29:51

From you talking about a test article made from existing components. FWIW, I like the idea but I'm not totally sold on the rocket-ramjet lower stage concept. It's just too complex, even if the rocket portion is flyback liquid rocket boosters. On the other hand, if your ramjet stage is boosted by the rocket stage, which then shuts off (throttles down?) until the ramjet stage is done that's another issue entirely.

I'm still holding out for an SSTO with a mass ratio of 5, but seeing as that necessitates an Isp of 590 s, I don't think I'm likely to get it any time soon.

That is the approximate exhaust velocity of a solid core NTR running on Ammonia. Other than the massive nuclear reactor running at 3000 K, that seems like a piece of cake to design

Terraformer · 2013-09-13 09:37:06

The idea as I'm imagining it has a dual engined first stage suborbital spaceplane that takes off under rocket power, boosts to ramjet take over speed, then flies under ramjet power up to it's max speed, finally switching back to rocket mode to boost it's altitude to say 100km and it's speed to 3km/s, at which point the upper stage is deployed. After this, it cruises back under ramjet power to the launch site, to be checked over, refuelled, a new upper stage and payload added, and then relaunched.

Of course, whether such a thing is plausible will require more than just speculation on a forum. Can a ramjet/rocket combo, running Methane and LOX, reach 3km/s@100km altitude with a mass ratio of 2.5? I'm thinking of 15 tonnes of inerts (30%) and a 5 tonne upper stage + payload (10%) for putting 500kg into orbit. That's the Scootaloo one I'm talking about.

GW Johnson · 2013-09-13 10:09:00

The real question is aeroheating damage to the airframe at speeds above Mach 6 (1.8 km/s in the stratosphere). This was catastrophic shock-impingement damage to the aft fuselage of the X-15 at Mach 6.67 with a scramjet test article on the ventral fin stub. The compression cone shock wave hit the underside of the X-15's tail. Had the flight lasted seconds longer, it would have cut the tail off the aircraft and caused a fatal crash. Above M6 in sensible air, the flight configuration has to be very "clean" so that shock impingement does not occur: basically a simple dart with stub fins, no nacelles. There are enormous pressure gradients through the wave, too. You have to be structurally extremely tough, as well as proof against ridiculously-high temperatures.

On the other hand, if you leave the sensible air before accelerating past M6, then you can avoid this issue. Shuttle's side-mount configuration avoided this by leaving the sensible air (about 80,000 feet or 25-ish km) at only about Mach 2. There's no sharp boundary here, the shock-impingement aeroheating problem can be significant at about M3, depending upon the materials you are using. X-15 had superalloy Inconel-X skins, which pushed the critical speed to about M6. They go into trouble by (1) going faster and (2) adding a nacelle.

Maybe Terraformer has the right idea just above: take off on rocket, climb then pullover-and-accelerate on ramjet, then relight the rockets and climb/accelerate to stage more-or-less in vacuum at 3 km/s. The booster airplane is a bit bigger because of the high altitude burn of very significant delta-vee, but the payoff is a second stage with a lower mass ratio requirement, so that its payload fraction is much larger. That reduces second stage size, but probably not overall launch size.

The portion of the first stage trajectory that can be powered by the ramjet shrinks in that scenario. For a supersonic design, takeover speed is about M 1.6-to-2 (0.5-to-0.6 km/s). The max speed capability will depend more on vehicle drag than the engine design, but with missile shapes falls in the Mach 4-to-6 range (1.2 to 1.8 km/s at stratospheric speeds of sound). The numbers I have run indicate too little thrust margin over drag for the weight, at altitudes over about 60,000 feet (20 km), because the air is getting too thin. Acceleration times become impractically long, raising vehicle weight again for all the extra fuel. Every design is different in detail, of course.

It's awfully hard to imagine a two-stage airplane design that (1) does not involve "effective nacelles", and (2) does not have higher drag because it is a cluster. Both are leading toward lower peak speeds from ramjet alone. If you solve that dilemma, pushing peak speeds toward Mach 5 or 6, then you run up against the extremized form of the shock-impingement aeroheating problem. This will be inherent, because the staged aircraft configuration will almost certainly have a side-mounted second stage, in effect a "nacelle" that sheds the offending shock wave, and perhaps also gets hit by one from the first stage. (THAT is also why I see no point to using scramjet for launch.)

Spaceplane design is not easy.

GW

Last edited by GW Johnson (2013-09-13 10:12:07)

Terraformer · 2013-09-13 15:21:56

That's why, at the moment, I'm not thinking of a two spaceplane approach. Just a reusable first stage spaceplane that deploys an expendable upper stage that doesn't have to go through the rigours of launch - the payload bay will protect it. Maybe later we can fit a lifting body spaceplane into the payload bay. If you're staging high enough, drag shouldn't pose a problem for "drifting" it out of the bay, as in the Black Colt design. I don't know how high you have to reach for that to be possible though - almost certainly in space, but will 100km do?

I'm not thinking of a design that gets more than a 1% payload fraction. Maybe later it could be pushed to 2%, with better materials and engines. That would allow the "Rainbow Dash" (100 tonnes GLOM) to service the small LEO satellite range. That would be the equivalent of the Pathfinder version of Black Colt.

JoshNH4H · 2013-09-13 20:53:26

Don't forget about on-orbit fuel depots!

Terraformer · 2013-09-14 12:38:49

Never. Celestia forbid it! If we can get to the coveted $100/kg, and put 2 tonnes at a time in orbit, then even one supplied from Terra starts to make sense if bigger launches are sufficiently more expensive (but when I checked, they'd have to be several times more expensive, unless you're already going for a depot anyway). That might actually happen though, so it may well be economical. That would require a high performing upper stage, perhaps even one using hydrogen.

At 2 tonnes at a time and refuelling, you might actually be able to start making some pretty nifty spacecraft by docking components on orbit. 2 tonnes should get you a propulsion and fuel tank module, and you could possibly dock an actual manned capsule (in two parts?) to it. The Mercury capsules didn't mass that much, after all, and the Gaiashield mission suggested by Zubrin gives 4 tonnes for the actual structure of the craft, before provisions, life support, and furnishing - that's two docked modules. If we have somewhere to actually do the work, say a large inflatable model launched on a Falcon Heavy...

It would allow some pretty cool probes using docked components, and possibly for a Lunar sortie. Though you might as well launch much of that as an upper stage.

JoshNH4H · 2013-09-14 19:41:10

Even if we decide that it's not feasible to build anything of significant size out of components that mass 2 tonnes or less, in terms of actually going somewhere we could reduce total launch costs by a factor of ~3 given that fuel itself costs $100/kg left in depots-- Or stored as ice in LEO to be electrolyzed by spacecraft

Terraformer · 2013-09-15 12:36:38

It would probably make the design too complicated and overly heavy, but... could we work an air-augmented rocket into the design as well? Perhaps by adding in some system to add the rocket exhaust into the ramjet, so that we can get much improved Isp shortly after takeoff... could we do it without modifying the ramjet at all (save for routing the exhaust into it's combustion chamber) and still get the improvement?

JoshNH4H · 2013-09-15 13:28:33

That's an interesting idea. My initial guess is an extremely unhelpful "maybe"-- So long as the Fuel:Oxidizer ratio in the engine is really, really bad, such that the exhaust velocity doesn't exceed mach 5 (1.7 km/s, 175 s). What could be done would be to fire a Hydrazine monoprop rocket into the ramjet inlet, then augment it with external oxygen. It seems possible that, if properly expanded, the pressure differential that could be caused by further expanding the exhaust jet would be enough to suck in oxygen even at zero velocity. I don't think that this system would actually represent an improvement over H2/LOX, though.

Terraformer · 2013-09-15 14:28:55

Well, considering that we're going to be having a fuel line to the ramjet combustion chamber anyway... what if we add in an oxygen line as well, and start using it with additional oxygen soon after launch?

I don't think it would be necessary to make the system work, but it would improve it somewhat if it could be done without too much trouble. Later additions could include a system to refuel the oxygen tanks whilst cruising hypersonically, so that the first stage can get to 4-5km/s instead. But that would be to investigate later, after the first system has been operating for a while.

JoshNH4H · 2013-09-15 15:45:40

But you still have the issue of hypersonic input into the ramjet inlet, don't you? I don't know if that's the major issue with scramjet design (as opposed to the actual supersonic combustion dynamics), but if so I suppose extra remass is always a good thing

GW Johnson · 2013-09-15 16:04:38

What Terraformer is describing in post 243 above is really an ejector ramjet. My friend Joe Bendot in LA is the premier expert in ejector ramjets. I was the expert in ordinary ramjets. Joe is in his 90's now, if he's still with us. I'm 63, so none of us will be around much longer. Ejector ramjet never got weaponized, while ordinary ramjet did, mostly by the Russians.

See also my post this date under the 600 sec thread, this division. It discusses combined-cycles vs parallel-burn separate engines. I come down firmly on the side of parallel-burn as something we could do right now.

GW

Terraformer · 2013-09-15 16:28:17

What about what I describe in post 245 - adding an oxygen line to the ramjet combustion chamber, allowing for it to be ignited soon after launch and provide thrust (I have no idea in the slightest whether that would have any shot at working - I'm not even studying aeronautics, let alone ramjets)?

GW Johnson · 2013-09-15 19:11:39

As best I understand, injecting oxygen into an otherwise-designed supersonic inlet-type ramjet would not sensibly affect thrust, but it would lower Isp. What it really does is stave off the low pressure-induced high-altitude flameout problem, simply because the oxygen concentration is higher. It makes the chemistry a tad more vigorous, when it otherwise doesn't want to be.

Whether that effect is actually worthwhile, is more than a little problematical. At the high altitudes where that becomes significant, the air is so thin that frontal thrust density is already rather small compared to frontal weight density. Thrust margin over drag divided by weight is your acceleration, which at low-"density" aero forces and high-"density" weight force, is small indeed.

If your acceleration is low at high altitude (and it will be, oxygen injection or not), it takes large numbers of range and time to accelerate to some desired speed. Especially for a vehicle that must also fly back to launch point, that adds greatly to weight, for the extra fuel that the long burn time requires. It's still a killer, even for a one-shot missile. You are way better off, if you can accelerate abruptly, rather than gradually. That simply takes thicker air (i.e., lower altitude). True for turbine, too.

Nothing absolute about this, it's a trade-off, different for every design. But, forced to generalize, I'd say stay under 60,000 feet (about 19 km) for "best" results with launch accelerators. Although, the design cruise altitude for ASALM-PTV was 80,000 feet (about 25 km). That was a ramjet designed for a Mach 4 cruise up there.

Like I said, every design is different in detail.

GW

GW Johnson · 2013-09-24 20:05:56

I ran a little bounding-type trade study for SSTO launch. Results are posted over at "exrocketman", in the 9-24-13 article.

GW

New Mars Forums

Announcement

#226 2013-08-27 10:23:22

Re: Reusable Rockets to Orbit

#227 2013-09-06 08:43:52

Re: Reusable Rockets to Orbit

#228 2013-09-06 11:14:46

Re: Reusable Rockets to Orbit

#229 2013-09-11 09:15:56

Re: Reusable Rockets to Orbit

#230 2013-09-12 08:27:36

Re: Reusable Rockets to Orbit

#231 2013-09-12 10:48:24

Re: Reusable Rockets to Orbit

#232 2013-09-12 14:31:27

Re: Reusable Rockets to Orbit

#233 2013-09-12 14:36:54

Re: Reusable Rockets to Orbit

#234 2013-09-12 21:27:24

Re: Reusable Rockets to Orbit

#235 2013-09-13 03:21:32

Re: Reusable Rockets to Orbit

#236 2013-09-13 08:29:51

Re: Reusable Rockets to Orbit

#237 2013-09-13 09:37:06

Re: Reusable Rockets to Orbit

#238 2013-09-13 10:09:00

Re: Reusable Rockets to Orbit

#239 2013-09-13 15:21:56

Re: Reusable Rockets to Orbit

#240 2013-09-13 20:53:26

Re: Reusable Rockets to Orbit

#241 2013-09-14 12:38:49

Re: Reusable Rockets to Orbit

#242 2013-09-14 19:41:10

Re: Reusable Rockets to Orbit

#243 2013-09-15 12:36:38

Re: Reusable Rockets to Orbit

#244 2013-09-15 13:28:33

Re: Reusable Rockets to Orbit

#245 2013-09-15 14:28:55

Re: Reusable Rockets to Orbit

#246 2013-09-15 15:45:40

Re: Reusable Rockets to Orbit

#247 2013-09-15 16:04:38

Re: Reusable Rockets to Orbit

#248 2013-09-15 16:28:17

Re: Reusable Rockets to Orbit

#249 2013-09-15 19:11:39

Re: Reusable Rockets to Orbit

#250 2013-09-24 20:05:56

Re: Reusable Rockets to Orbit

Board footer