Carrier Aircraft for SLV
Nov 12th, 2008 by johnhare
guest blogger john hare
My propeller concept was pretty well destroyed in comments. Several major flaws in the concept were pointed out. This one is more conventional, or at least as close as I get.
A custom aircraft for air launching spacecraft is an attractive option, just an expensive one. The current aircraft used for the purpose are converted from other purposes. Even the high performing Scaled Composites custom aircraft are optimised for high altitude capability rather than grunting a large spacecraft to altitude. That is how they can use them for other revenue purposes and spread the costs over more than the occasional space launch.
A relatively unsophisticated air vehicle could get a spacecraft to 50,000 plus feet if it focused more on a blend of technologies than sheer aerodynamic excellence. Adding a rocket to a conventional aircraft has been discussed many times for permitting a zoom climb to spacecraft release. The problem is that all of these existing vehicles have been designed to do something else from carrying passengers to carrying cannon. A passenger aircraft is optimised for cruise and high flight rate, which doesn’t necessarily make it happy with abrupt acceleration and zoom climbs. Fighter aircraft carry a lot of baggage in terms of supersonic and maneuverability requirements that don’t always match the best performance for an SLV carrier.
We need something strong, light, high payload capacity, economical to develop, and with a carrying mount area that works well with the upper stages that are intended to be the main job of the vehicle. We can trade off good economy cruise as well as high altitude level flight capability and maneuverability for the qualities we need.
Light weight and strength can come from a compact layout without long roomy fuselages and high aspect ratio wings. Low aspect ratio wings are much stronger per square foot area and we don’t need all that interior volume for exterior loads. Induced drag rises with low aspect ratio which means it will take more power and wing area per total weight of vehicle. Wing area increase is acceptable with the much lower weight structure and more engine is cheaper than more engineering. Since we don’t need interior fuselage volume except for the aircrew, with the fuel in the wings and exterior payload, a flying wing design design really suggests itself. Just put a real tail section on it for both strength and control. If it has to have fly by wire to be feasible, it is too complicated aerodynamically for the job at hand.
One of the comments on toroid tanks suggested connecting cylinders instead of the complex saucer shape and donut tanks in that post. While I think the torus and saucer will not be that bad, the connected cylinders do work well for a flying swept wing. By tying two strongly swept wings to a tail in an equilateral triangle, it is possible to have a very strong structure with external cargo capacity in the middle of the triangle. With the high horizontal tail, the spacecraft can roll in under the tail and behind the wing for a roomy carry position without extensive ground facilities. Landing gear can be relatively short as ground clearance is not an issue for the nested spacecraft.
With jet engines under the wings in airliner fashion, and a pair of rocket engines in the aftwing/tail area, more than enough power can be available to get off the ground and zoom climb to 50,000 plus feet. A vehicle that has the job of launching an SLV from 50,000+ feet at near vertical attitude, does not necessarily need to be good at high altitude cruise. High altitude cruise requirements could actually hinder the vehicle from doing a better job of launching vehicles. This vehicle is two wings, a tail, a cradle, a cockpit, and a bunch of engines in a strong framework.
The less efficient aerodynamic layout of this vehicle compared to others used for the purpose is a trade for weight and strength. If cruise must be done at 20,000, or even 10,000 feet instead of 50,000 to reach the launch area, it will just cost a little time and fuel. When it reaches the designated launch area though, lighting the rockets can bring it past 50,000 feet in a steep, high subsonic climb in about a minute. After releasing the SLV already pitched up for launch, this carrier aircraft has plenty of strength to recover from the high altitude peak without excessive structural concerns.
There is no need to push the margins of aerodynamic design and structural skill to get an aircraft to deliver an SLV to a high altitude launch point in the right attitude. There is nothing wrong with a high altitude aircraft to do high altitude aircraft work. There is something wrong with accepting serious cost and performance compromises by a vehicle designed for the wrong job. The only thing that makes using those high altitude efficient aircraft good for space launch, is that they are making their investment costs back by doing something else. A simple rocket will do in sixty seconds a job that takes an excellent wing an hour. If a company is in the rocket business, then they already have rockets with no need to push the boundaries of aerodynamic capability for the launch business.
Over a period of time, this launch vehicle carrier could incrementally raise the release altitude to 100,000 or so feet with the considerable propellant capacity of the structural tank wings. Follow on versions could eventually raise the staging velocity beyond supersonic in the low dynamic pressure of very high altitude. A million pound lifter of this configuration would still fit commercial airports.
For the early units lifting under a ton, a home built type aircraft would work. With all the wing sections identical, and a straight tail section, it should be less difficult than some of the high performance pressurized composite planes.
After a number of analysis iterations and a bit of head butting with Scaled, we’ve found that it is generally better to move the propellant for a gamma turn rocket from the aircraft into the launch vehicle; I can’t say they have come around as yet, however.
Also, a major reason for wanting good cruise is that one of the principal benefits of air-launching is the ability to fly cross track in order to match ascending nodes with you rendezvous target on orbit, so that you can fly direct ascents and not have to phase. That requires fairly efficient cruise, on the order of 1000 km for any-day launch.
I like carrier aircraft where the wing is not wet; much simpler to design and build. Fuel can be in podded tanks or booms.
Gary C Hudson,
Considering your knowledge and experience in this field, and my lack of both, I would be a fool to debate with you. I guess I’m a fool.
After a number of analysis iterations and a bit of head butting with Scaled, we’ve found that it is generally better to move the propellant for a gamma turn rocket from the aircraft into the launch vehicle; I can’t say they have come around as yet, however.
From the outside looking in, it seems to me that Scaled is the best aerodynamics and composite construction company in the world. Their carrier aircraft just seem too good as aircraft relative to the exact purpose of launching spacecraft. From a systems viewpoint, it seems to me that it would be better to build a good average aircraft to launch from say 25,000 feet, than a real work of art that can cruise at 50,000. It looks to me like the development money would be better spent shifted to the spacecraft itself.
Also, a major reason for wanting good cruise is that one of the principal benefits of air-launching is the ability to fly cross track in order to match ascending nodes with you rendezvous target on orbit, so that you can fly direct ascents and not have to phase. That requires fairly efficient cruise, on the order of 1000 km for any-day launch.
Is it really necessary to cruise at 40,000-50,000 feet though? Down in the thicker air it just doesn’t seem to require as much aircraft to get the job done.
I like carrier aircraft where the wing is not wet; much simpler to design and build. Fuel can be in podded tanks or booms.
Your experience certainly trumps my guesswork. Simple design and build is certainly the target of this post. I will have to rethink some of my opinions.
I wonder if there’s any merit in using the same basic structure as a “cruise booster” for the RLV, rather than a separate carrier aircraft. Rather than building a complete aircraft, build a simple wing/fuel tank that you can attach to the LV to give it cruising range. Piloting and (possibly) propulsion come from the LV, while lift and range come the wing. There’s a case for putting airbreathing engines on the wing, but putting them on the LV will give you simpler control systems and flyback capability. I suspect the optimum lies somewhere in the middle; with low specific thrust engines on the wing for cruising, and high specific thrust engines on the LV for cruising, launch, and flyback. Once the LV separates from the wing, the wing could fly to base as a UAV; or if it were cheap enough, expended (I prefer the former, although its clearly more complicated). Its something for the trade space anyway.
One final thing. You’ve drawn the aircraft with both an elevator and ailerons. With a split elevator to provide roll, you might be able to do away with the ailerons.
John,
*Never* assume I’m right and you’re not; in fact, I really like a number of your engine ideas (perhaps since I’ve played with similar concepts for a decade or so). The way of progress is that we all question everything and from that questioning, innovation emerges! And being “right or wrong” in engineering is oft times tied to the assumptions with which one begins the debate.
I think an analogy I could use is to point out that it is harder to make a “just OK” sort of carrier aircraft than a good one. It’s a bit like building an expendable pump-fed rocket engine. You can’t build a “just OK” one that is good for only a flight or two, because in order to build it properly to survive that one cycle with high reliability, you end up putting in margin that yield relatively longer life. That’s why the RL-10 can handle a few hundred cycles right out of the box even though it is “expendable”.
As for altitude, I don’t think one needs 40-50K feet. We found the knee in the curve is about 25-30K. Higher is somewhat better, but once again, it is best to put the weight into the rocket than add altitude performance to the aircraft, once you have reached about 30K ft.
One of these days I might write this up in more detail. Meanwhile, you can find a partly relevant paper at the AirLaunch website. It’s off line today as we transition to a new hosting service, but when it comes back I’ll give a proper pointer. Or I could send the paper to you or John by email.
1) John, your idea reminds me of a similar concept proposed in the mid 80’s for ALS launch (Advanced Launch Systems). It was a simple, almost skeletal, aerodynamic “cradle” with turbojet engines. I don’t recall the contractor or much else about the proposal (including whether it was sub- or supersonic, although I believe it was to be low supersonic), but do recall that it was shown in “Aviation Week” magazine.
2) Tim’s idea made me think of another concept: adapting the aerodynamic characteristics of Spaceship 1 (and probably also 2) for a flyback booster. It would have to be an off-the-pad vertical launch, but the booster would have a similar wing configuration to Spaceship 1, or at least use the same stall and feathered descent technique for the flyback.
3) I assume Gary was relating his experiences from his stint at T-Space. I wonder how T-Space is doing these days?
I have CAD design of blended wing body now. Google millennium academy for early prototype images. X-Plane and models fly OK, building R/C model next. Hope to show new airframe tech at Oshkosh. Power by Bill Colburn hybrids and PW jets.
I have CAD designs for a blended wing body spaceplane now. Power by Bill Colburn’s hybrid rockets, and PW turbines with new ariframe tech. Building a prototype R/C model and hope to display at Oshkosh soon. My web site (above) shows obsolete prototypes aimed at the X-Prize. Previous models and X-Plane sims indicate potential mission capability.
oops…Blackberry DID work…lol
“3) I assume Gary was relating his experiences from his stint at T-Space. I wonder how T-Space is doing these days?”
Both t/Space and AirLaunch LLC analysis contributed to my opinions.
t/Space is not doing anything at the moment; our experience with NASA has not been positive. Once Griffin is gone then maybe we can revisit matters but I don’t plan on holding my breath for real change…
Gary,
*Never* assume I’m right and you’re not; in fact, I really like a number of your engine ideas (perhaps since I’ve played with similar concepts for a decade or so). The way of progress is that we all question everything and from that questioning, innovation emerges! And being “right or wrong” in engineering is oft times tied to the assumptions with which one begins the debate.
I have to assume you are better informed than I am. I’m not in the business, and it will be at least two years before I will be able to afford to try to get back in again. One thing that concerns me is drifting into the Geatano/Mook mode where I assume I have all the answers and expect all to agree. That is dangerous to me and possible, partly because I could easily reach a point of not comprehending that I don’t even understand the argument I’m lecturing others about. My turbine based ideas lean heavily on the roton concept you and Bevin pioneered, so mostly I’m probably reverse engineering you.
I think an analogy I could use is to point out that it is harder to make a “just OK” sort of carrier aircraft than a good one. It’s a bit like building an expendable pump-fed rocket engine. You can’t build a “just OK” one that is good for only a flight or two, because in order to build it properly to survive that one cycle with high reliability, you end up putting in margin that yield relatively longer life. That’s why the RL-10 can handle a few hundred cycles right out of the box even though it is “expendable”.
Perhaps it might be better to suggest that the carrier aircraft should be optimized for heavy lifting and medium range, with low development costs. Higher fuel consumption and lower service ceiling being acceptable trades to hold development costs down.
As for altitude, I don’t think one needs 40-50K feet. We found the knee in the curve is about 25-30K. Higher is somewhat better, but once again, it is best to put the weight into the rocket than add altitude performance to the aircraft, once you have reached about 30K ft.
My assumptions seem to be mostly based on the Scaled high altitude units. It just seems to me that they are doing some things the hard way. I’m also influenced by the Space Van concepts Len wrote up in the late 1990s, especially the economic arguments.
One of these days I might write this up in more detail. Meanwhile, you can find a partly relevant paper at the AirLaunch website. It’s off line today as we transition to a new hosting service, but when it comes back I’ll give a proper pointer. Or I could send the paper to you or John by email.
I went through the half dozen or so AirLaunch papers I saw a few months back when Jon linked to them. They influenced my thoughts here. If there are more, I would like to read them. rdnktech@tampabay.rr.comspam
Roderick,
You reminded me of another booster idea a few of us discussed a few years back. I’ll add it to my list and post it sometime.
I sent you a newer t/Space paper from September. It was really written a year ago but we didn’t want to release it in 2007 for competitive reasons.
Without revealing Scaled details I can’t talk about, it is generically true that one can trade fuel weight down and range down and increase payload weight, up to a point where there are Cg or structural load issues on the airframe. Making such trades will oft times increase payload as much as 50%, we’ve found. And in the end, everything comes down to economics. Much of the sizing we did on custom aircraft was driven by what was available on the surplus market in the way of engines, landing gear, and actuators or other subsystems. The only affordable way to build a new custom aircraft for the space launch mission is to reuse that sort of item from a certificated airframe.
Gary, extremely fascinating stuff!
What about launch aircraft speed versus its payload fraction? (And launch aircraft cost…)
Cause I think on a higher level look supersonic air launch doesn’t make much sense (assuming a new carrier aircraft and a significant space payload, not just a nanosat from an F-15 or some such), but that’s just a hunch on my part, not having much high speed aircraft knowledge.
With rockets adding or reducing delta vee and calculating the effects is pretty straight forward. Not with airplanes.
One thing to note about cruise altitude, is that being able to cruise at high altitude means you can avoid a lot of weather. Getting to 30Kft gets you over a lot of it; but 50Kft gets you over pretty much all of it. As long as you have a weather window over your launch runway; environmental condtions no longer preclude you from launching.
That said, if you need to get over 30Kft to avoid weather, you’re probably dodging thunderstorms, which is pretty unhealthy to begin with. So the earlier comment about a ‘knee’ in the calculations at about 30Kft is either related to this, or conveniently close.
“Cause I think on a higher level look supersonic air launch doesn’t make much sense (assuming a new carrier aircraft and a significant space payload, not just a nanosat from an F-15 or some such), but that’s just a hunch on my part, not having much high speed aircraft knowledge.”
You are right. As I recall the sensitivities, the most important attribute is altitude, followed closely by flight path angle and then airspeed.
John:
I’m wondering if analysis of carrier aircraft includes the possibility of taking off with both the carrier aircraft and the rocket with a minimal fuel load. Refueling has become more commonplace, with C-130’s, 767’s, and Airbus offering refueling capability. Whether custom or off the shelf, the carrier aircraft can increase the size of its payload if it takes off with 50,000 pounds less fuel. The rocket stage, if it uses aviation fuel, can be much larger if it takes off empty. Mitchell Burnside Clapp proposed loading LOX while airborne, I suggest that the carrier aircraft could be tanked as well. Considering the comments about cross range capability, and your notion that the carrier craft can fly to its launch point at lower (denser air) altitude, I would guess that the payload would see a large increase if the takeoff load was lightened.
Charles,
It does seem that in flight refueling of the carrier and SLV should be early on the list of improvements. I believe some of the heavies are nearly 50% fuel on long haul flights, so the 50,000 pounds you mentioned is probably very conservative.
Did Mitch suggest LOX instead of peroxide in one of his papers? I seem to recall that he made a very good case for peroxide only in flight transfer. Non cryogenic and around 85% of total propellant loading.