What is the future of reusable launch vehicles?
It should be clear by now that it’s hopeless to expect the government to do it.
12 thoughts on “Space Transports”
Comments are closed.
What is the future of reusable launch vehicles?
It should be clear by now that it’s hopeless to expect the government to do it.
Comments are closed.
Yep, loss of government involvement is no (net) loss.
I’m curious if all the proposals used winged flyback and landing, or if some tried alternate approaches.
*** long boring description of a bizarre idea on a re-usable booster follows ***
I’m stlll looking at the oddball re-usable VTOL (non tail-sitter) launch/landing concept and different paths to implementation, and the numbers don’t look too bad (although at current engine costs the reduced pad support still probably wouldn’t beat a tail sitter until the vehicle size is very large). For those who missed a previous discussion about this idea, it’s simply to hurl the rocket into the air from a horizontal resting position, taking about 15 to 20 seconds to lift and rotate it.
For an initial stab at the idea, my thought is to only replace the two booster stages on a Falcon 9H with horizontal launch/landing capable versions. They’re longer (with lots more intertank sections), extending to the start of the payload shroud, and then include a four-foot diameter nose truss that extends along the payload shroud for horizontal support. They’d have horizontal engines scattered along their length, reducing bending moments, (and requiring the extra inter-tank sections), and instead of trying to lift the mass of the core stage via a highly torqued connection between two cylinders, they instead extend additional horizontal engines underneath the core stage, swining out much like landing gear, so the core stage mass is carried by mating plates, with the engines underneath the plates.
The extendable engines could hinge on a pair of large ball valves, one for the fuel line and one for the oxidizer line, and the fuel and oxidizer lines could double as the engine struts (though the main thrust loads are transfered directly to the outside of the core stage). The ball valve/hinges prevent even the remote possibility of firing an extended core-lift engine unless it’s deployed in the core-lift position, since when they’re retracted and stowed for vertical flight they’d be aimed inside the booster.
The importance of the core-lift engines is that they can spread the lift force along the length of the core stage from underneath, to avoid having to beef it up for horizontal bending loads or significantly modify its structure, so the core doesn’t suffer a performance penalty and need not even be re-usable in the first iterations. So version 1.0 of the hoizontal lift boosters are simply a way to horizontally lift and rotate a Falcon 9 1.1, and then act as a pair of conventional strap-on boosters with an easier flyback capability.
The mass penalty is from roughly doubling of the number of engines (about a 1% increase if lift-off weight) and adding landing gear on the boosters, which based on aircraft weight estimation should come to about 3% of lift-off weight, a number that can probably be cut way down given the reduced requirements on touch-down vertical velocity and high landing speeds. So the total mass penalty is a reduction in fuel fraction of about 0.05 of the entire assembly, or about 0.08 fuel fraction of the booster sections (where all the additional hardware is concentrated), although the extra intertanks might increase the mass penalty over those numbers. Given LOX/RP-1 boosters, that means their performance should be comparable to a conventional SRB (not including flyback), so the payload would be much larger than a Falcon 9 1.1 but quite a bit less than a Falcon 9H.
If the extendable core-lift engines also include landing struts to support the weight of the core on the ground, the assembly could also de-rotate and land while fueled, giving an intact abort/return capability that a tail-sitter can’t match, since the most likely reason for an abort is a problem with the vertical engines, which are required for safely landing a tail-sitter. The intact abort/return capability would last almost until booster seperation if the still-fueled core stage could cross feed back into the booster tanks for the landing phase.
Other flyback benefits are a more forward center of pressure from having the engines and landing gear spread along the booster section instead of being only in the back, vastly increased control authority instead of rear-steering, and no need to derotate from the horizontal return flight to touchdown.
For horizontal landing the vehicle could use fairly conventional landing gear (tires don’t like flames), which can have lots of shock absorption without affecting the stability of the vehicle, whereas a tail sitter really needs of have the entire stage acting as an oleo strut sliding in a footpad to avoid tipping. And of course tail steering a vertical stage is like balancing a broomstick, whereas a horizontal stage should be as easy as landing a Harrier. You could go with wings and an aircraft style landing, but then you can’t land a fully fueled booster, much less the entire stack, so you’d lose most of the inherent intact abort capability that comes from already having the engines required for a horizontally-oriented vertical launch.
Unfortunately the whole idea requires a heavy modification of a conventional booster, including landing gear (possibly taken from bone yards for version 1.0), the retractable core-lift engines, horizontal booster-lift engines in the intertank sections, flight software, and lots of hover and rotate testing, so it would not be a cheap booster to develop, being an all-up VTOL rocket with retractable gear, big engines on the back, and other expensive features. But once working it would be very cheap to use, assuming very reliable, low maintenance engines (which is a big assumption that probably pushes the idea into the distant future), eliminating all vertical operations and a launch tower and allowing higher flight rates, and the concept would work for just about any existing launch vehicle.
If may not be financially viable for Falcon 9’s or EELV’s, but for very large vehicles where the the crawler, tower, and vertical processing costs skyrocket, it might make a practical architecture, especially if you combine it with a re-usable core first stage.
I’d be a bit further along in my number crunching but a wasted a couple days figuring the weight penalty for going with a pressure-fed system for the horizontal engines, which was significant even though they only need to run at full throttle for about 10 or 15 seconds, and maintaining the abort capability of course made the tank mass penalties much worse.
*** end of long boring description of something that nobody would build ***
You seem to be putting all the mass on the rocket. Why not go with a launch sled? The twisting torque does have to be that much if you power down while doing it and power up after (putting all the velocity change back in line.) Given enough initial velocity there may not even be a need to power down first.
(I like reading your boring stuff, George.)
doesn’t, I meant.
Well, a launch sled that carried the structural loads through the lift and rotate maneuver was one of the early consideration I considered, and it would certainly result in less structural mass during the subsequent ascent, but it doesn’t retain the intact abort capability. When I saw that the mass penalties in horizontal launch weren’t all that big, it made sense to evolve the sled into a pair of re-usable strap on boosters.
A configuration that’s even lighter than a sled is to have the base of the rocket sitting in a rotate cradle and lift the rocket into position with solids, which is probably the minimum engineering that could accomplish the task (even simpler than using a crane), but there’s no recovery option if there’s a problem, and if the first stage doesn’t ignite the astronauts won’t have a way down, so that’s out, too.
My thinking in this whole design exercise is based on figuring out what rocket operations would look like if we were actually good at it, by which I mean if the equipment was extremely reliable and flights were almost as routine as airline flights. We’re not there yet, but eventually we’d like space launches to be simple.
As an example, imagine what has to happen once SpaceX returns and vertically lands a first stage tail-first. The stage is back but it’s still a hundred-foot tall multi-million dollar piece of aerospace hardware, so a large team of engineers and technicians will go out to the landing site with specialized equipment like crawlers, lifters, and a crane.
They’ll have to carefully interface the rocket to a lifter just to move it, using big clamps or bolts to make sure it remains stable. Then they’ll have to lower it to the horizontal position and lay it down on a specialized transport truck, so they’ll be out on the pad for hours. Then they’ll haul it back to an assembly building so they can mate it to another second stage, and once that’s all done they’ll have to haul it back out to the pad and erect it, again using specialized cranes and towers.
In airline terms, it’s taking a couple thousand man-hours of highly trained specialists to get the plane from the runway to the terminal and then back out to the runway. Whenever you have to use huge cranes that can erect or lay down a flight vehicle, you’re going to have a gaggle of engineers and supervisors to oversee the operation, and that will never be cheap.
With horizontal landing, ideally you just send forth a couple of high-school graduates you hired from Miami International to hook the returned booster stage to a vehicle and tow it back to the hanger, or you even leaving it sitting on the runway and just tow the new stages out to it for mating where it sits.
Following along with that concept, mating boosters to a core stage has to be something almost as simple as backing a semi-trailer up to a loading dock, shifting it into mating position with some built-in feature, and locking it together automatically (instead of carefully torquing a bunch of explosive bolts – and having a supervisor verify the torques, etc). The concept is to add complexity to the flight vehicle to eliminate complexity in its turn-around, even at a large performance penalty, because any required ground operations have to be performed correctly every time it flies.
Some might think launching a highly complex rocket without a ground-support crew numbering hundreds of people is impossible, yet we’ve all seen it done. Neil Armstrong flew one off the moon. You just have to make it a design requirement.
Anyway, another thought on using the side-boosters is to build them with chines that snuggle up to a cylindrical core so the core-lift motors don’t have to rotate out from the booster. Think of parking two SR-71 fuselages edge-to-edge with a rocket resting on the chines. You still use the core-lift engines, now embedded as close to the edge of the chine as you can get them, with fairly short nozzles. Short nozzles means each individual lift engine is small, so a whole lot of would run all along the chine, which should have enormous structural benefits in lifting the core. Putting the booster-lift engines toward the inside of the chines, along with the landing gear, wold eliminate the extra inter-tank section in the booster.
The chines would also certainly improve the fly-back performance and having chine-mounted engines would provide much greater roll authority throughout the booster’s flight envelope. I’d also be tempted to use several turbopumps to feed all the small lift engines, either with a centralized plumbing system so all turbopumps can feed any the motor, or feeding them so the loss of a turbopump kills motors distributed all around the vehicle, so there’s a loss of overall lift instead of a loss of all motors in a particular area.
The only thing I can’t figure out is how you mate the two chine boosters to the core while the core is still supported on its own landing gear, which somehow has to get out f the way so the boosters can slide in underneath. An overhead gantry would make the operation trivial, but I’d prefer to avoid using one if there’s a simple and clever way not to.
I sort of disagree with the article. The X-37 is no less reusable than Buran was. Honestly I’ve always been a bit skeptical of fly back reusable booster concepts. Since both SpaceX and Blue Origin are working on reusable VTVL launch I don’t see any reason to complain.
I wouldn’t be surprised if the DoD still has some HTHL near orbital concept being worked around somewhere. I somewhat doubt Blackswift was the end of it. This has been worked around since WWII and development never really stopped.
They will emerge when the market justifies their existence. Currently the market is not large enough to do so at the launch prices that would be required to provide a ROI on their both their operation and development.
No mention of Skylon…
The part I like the most about SpaceX’s technology, is that you don’t have to be 100% succeful to get something out of it. With the Falcon Heavy, landing ability would be cost reducing even if all you get back is the side boosters. Getting the 2nd or 3rd stages back would be a bonus.
“One thing is clear: in an era of constrained budgets and competing priorities, there seems to be little appetite within government for another large RLV development program as has been attempted several times over the preceding decades.”
People keep saying stuff like this yet our spending keeps going up.
Not the kind of spending you or I would prefer though, I imagine …
One thing I think people overlook with the airplane analogy is regular maintenance. It’s not just fuel and go. Time between maintenance is measured in hours.
With existing launch vehicles you have to use a calendar. I think the fastest turn around on a Shuttle was two-months, and that was for a fuel-and-go carrying the same payload due to an incomplete mission. If we could get a re-usable launch vehicle up to what the Air Force or Navy would consider a high-maintence, low readiness jet aircraft, we’d be doing great. The SLS is being designed to “sit happily” on the pad for almost six months prior to launch. It’s like it’s being designed as a decoration instead of a flight vehicle. The Air Force usually waits till a plane is retired before they stick in a pole at the main gate.