The Shuttle Legacy

When Project Apollo was shut down, the most tragic aspect of it was all of the useful technologies that were lost as tooling was destroyed and experts were reassigned to other programs. The mighty F-1 engine was relegated to museums, ceding the kerosene-engine market to the Russians. The demise of the J-2 engine has led to an expensive development program for the new J-2X that will be used on the Ares launchers. Even the exact formulation and processes for creating the Apollo capsule's ablative heat shield were lost to time, complicating the effort to develop the Orion heat shield.

As the Space Shuttle winds down, it appears that the same mistake is not being repeated, at least not on the same scale. In taking stock of the program's technical accomplishments, many of them are being preserved or leveraged for the Ares and Orion systems. Of the ones being discarded, they have served as lessons for ways that a reusable launch vehicle should not be built.

When NASA transitioned from Saturn to Shuttle, significant propulsion elements had to be re-developed. While Space Shuttle Main Engine owes a lot to the J-2 program, it is a much bigger engine with higher specific impulse and thrust, a more complex staged combustion cycle, and built-in reusability. The solid rockets were a massive undertaking in many ways, eclipsing any solid rocket with flight history up to that point in time.

Some elements of the propulsion system will remain relevant for the Ares generation. The shuttle solid rocket boosters will be leveraged for the boosters on the Ares rockets. While the new boosters are a leap beyond the current SRB, it's not as far of a leap as the original SRB was when compared to its predecessors. Additionally, the shuttle's maneuvering engines are being re-used for Orion. This may be the only element of the shuttle system that is reused with no changes.

For other shuttle developments, they will best serve as lessons learned in the development of equivalent systems. Perhaps the shuttle's most remarkable achievement was its main engines. Nevertheless, the SSME's taught us a lot of the ways not to design an engine for producibility or reliability. The high chamber pressure and staged combustion cycle ensured high performance, but required lots of ground support equipment. Thousands of tiny welded tubes in the nozzle and chamber for cooling? It was state of the art for the 70's, but channel-wall cooling is the preferred method nowadays. The RS-68 benefited from the lessons of SSME, sacrificing specific impulse in favor of producibility. Its gas-generator cycle, lower chamber pressure, and channel-wall chamber with ablative nozzle make for a much cheaper engine. Its only drawbacks when compared to SSME are specific impulse (which can be increased with a redesigned injector and regen-cooled nozzle) and lack of reusability. In fact, a channel-wall nozzle was planned as a shuttle upgrade until the program's 2010 retirement was announced.

At the same time, a lot of the shuttle's pioneering achievements in the field of re-usability are being discarded as dead-ends which taught us how not to build a reusable launcher. Case in point is the shuttle's thermal protection system. While the blankets will likely find use on the cooler surfaces of a future reusable launcher, the other heat shield materials will likely be dismissed. The ceramic tiles still are remarkable, but they form a complex system that is difficult to maintain. Reinforced carbon-carbon had incredible abilities to stand up to high temperatures on the shuttle's nose cap and leading edges, but they were too brittle to reliably ensure safe reentry. A future reusable launcher will likely be a "fluffier" design along the lines of X-33, which can get by using a robust, metallic thermal protection system.

Overall, NASA and the industry are taking a wiser approach to the end of the shuttle program than was taken at the end of Apollo. Many critical technologies are being reused, albeit in expendable rockets. The clear succession from Shuttle to Ares is mainly in the field of propulsion, where breakthroughs during the shuttle's development have reduced the risk for Ares. The enduring challenge from the shuttle program is to learn the correct lessons from the reusability concepts that proved so difficult to implement on the operational shuttle.

Engineering a Spacefaring Society

The goal of Project Constellation is to take humanity to the moon, Mars and beyond. That is a noble step on humanity's drive to spread out beyond the earth, explore the universe, and preserve our society beyond the Earth's inevitable death. But it's also clear that there's a massive gap between Constellation's goals (a small moon base and Mars sortie mission) and our eventual goal of spreading out beyond earth. The end-state is what I like to call a "spacefaring society."

In some ways, I think that Project Constellation ignores the technological needs of a spacefaring society in the name of budgetary and schedule expediency. Transportation to the moon and beyond will be achieved with brute-force, by launching large rockets from the earth. But if Constellation is truly a marathon (rather than a sprint,) it should focus on the long-term development of the essential technologies which will enable human exploration of the solar system.

My "technologies for a spacefaring society" list is nothing new. Many of the all-time great visionaries in the space business have said the same things that I'm saying here. Nonetheless, the admission of things that require development is an admission that humanity is not ready to truly leave the cradle for good.

The following is a list of eight technologies that I feel are essential to human exploration of our own solar system. Rather than a "moon-first" focus, I'm beginning to feel that Constellation should be redirected towards developing these technologies before we return to the moon.

1) Space Nuclear Power
Albeit a controversial technology, portable nuclear reactors have the capability to make bases on the moon and Mars sustainable without being hostage to the sun (or Martian weather.) While they would require periodic replenishment from earth, space surface reactors are the way to truly power long-term human bases on other bodies of our solar system.

2) Large-Scale Electric Propulsion
This is not a prerequisite for lunar travel, but they certainly make for a more fuel-efficient cruise to Mars, asteroids or beyond. Current electric thrusters have put out very tiny amounts of thrust. We need much larger thrusters for human interplanetary missions.

While solar power is a possible power source for electric propulsion units, the large array size needed to drive the electric thrusters would make a spacecraft more vulnerable to micrometeoroid impacts. For human missions, nuclear reactors are the preferred power source for driving electric rockets.

Electric propulsion isn't useful for human travel to the moon due to their low thrust. My preferred lunar transport architecture uses nuclear-thermal rockets. Without losing too much thrust over chemical engines, they offer almost twice the specific impulse of hydrogen-oxygen rockets.

3) Artificial Gravity
While we don't understand everything there is to know about long-term exposure to weightlessness, we've seen enough to realize that zero-G is largely detrimental to the human body. We still don't fully understand what rotation rates the human body can tolerate while subjected to an artificial-G system. We also need lightweight materials that can build strong structures for artificial-G spacecraft. An open question is whether it's better to subject a person to a slow, constant spin on a large-diameter wheel spaceship, or if an astronaut could spend short periods of time in a very fast centrifuge to counteract the effects of weightlessness.

4) On-Orbit Fueling
On-Orbit refueling has so much to offer for spacefaring societies. It allows us to launch massive spacecraft from earth into orbit, as long as those craft are launched with empty propellant tanks and refueled on-orbit. Propellant stocks make for cheap payloads and increase the demand for earth-to-orbit transportation.

Beyond their use in earth orbit, propellant depots on the moon and Mars will enable reusable transportation to ferry astronauts between space transportation hubs and the lunar/Martian surface. Whether it's a space station at EML1/2 or on the Moons of Mars, it will serve as a useful staging ground for landings and for inbound or earthbound astronauts.

5) In-Situ Resource Utilization
The ability to make propellant on the moon or Mars will save the expense of launching so much propellant mass into earth orbit. The Sabatier reaction and electrolysis on Mars can produce methane, oxygen, and water by simply using Martian atmosphere combined with hydrogen feedstock. Future missions can make use of water and other substances we find on the moon and Mars.

If astronauts are to survive with little or no resupply from earth, they will need to adopt a "live off the land" philosophy. Just as the pioneers of the American frontier learned how to be resourceful with the things they found in the lands where they settled, so too will the astronauts who settle the moon, Mars and beyond.

6) Closed-Loop Life Support Systems
Unless astronauts can carry massive amounts of consumables with them for long space voyages, they will need to close the life-support loop. Practially everything will need to be recycled. That even means having to find a way to recycle the astronauts' poop. It's a dirty job, but NASA or somebody will need to develop a "biodome" capable of sustaining life with a minimum of mass that will need to be replaced.

7) Aerobraking & aerocapture
The ability to use the atmosphere of Earth or Mars to brake large payloads will save much propellant mass in an earth-Mars transportation system. It's essentially like getting a free ride, as long as we can build heat shields and guidance systems that can make aerobraking effective.

8) Reliable, routine transport to earth orbit
This is a major sticking point for a lot of the space pundits. Many people can't get past the idea that expendible rockets are so wasteful. But it's also true that reusable launchers are more expensive to develop and more expensive to operate. In the current paradigm, the best way to provide human transportation for earth to orbit is with simple, throwaway rockets and simple capsules.

As we build simpler rockets and capsules that can reliably increase the demand and ability to put humans in orbit, we'll get closer to the launch rates that will make reusable rockets cost-effective. It may take 50 or more launches per year to make economic sense of a reusable spacecraft. That day may not come in my lifetime, but it eventually will come.

My current thinking on reusable launchers is that a scramjet-powered first stage would be required for a manned spaceship. It would need an assist from the upper stage's rocket engines for takeoff, but it would then accelerate to Mach 12 (approximately) before releasing its "spaceliner" upper stage. The scramjet-powered mothership would be capable of airliner-like operations, as opposed to the relatively maintenance-intensive upper stage. In the system I envision, there would actually be fewer motherships than spaceliners in order to meet the demand for spaceflights.

Epilogue--Growing Up in the Cradle
As Tsiolkovsky wrote, "Earth is the cradle of mankind, but man cannot remain in the cradle forever." While his words still ring true today, it's clear to me that we are still mere babies in this wide universe. We're trying to stick our hands out of the cradle, but we just don't have the strength to pull ourselves up and out. We will only grow up when we invest time, money, brainpower, willpower, and patience (most importantly) in the tools that will make us strong enough to rise from the cradle.

Say It Isn't So

If all goes according to plan, NASA will soon re-award roughly $175 mil that was originally supposed to fund Rocketplane-Kistler's COTS entry. The loss of this contract is likely to leave RpK in dire financial straits.

On the flip side of the coin, Jeff Foust reports that RpK is redesigning its XP space tourism vehicle to make it more competitive in the suborbital marketplace. The logic behind the change is sound. But the fact that RpK could make that change indicates to me that they were never as far along in the development of XP as they publicly claimed. I haven't seen much info on the engine that the team plans on using (which is allegedly based on the old Atlas sustainer.) It's also been stated that many of the team's personnel, like the legendary David Urie, have been released due to the company's focus on keeping the COTS program alive.

It's been pointed out that, during the late 90's, the Kistler K-1 essentially killed the original Rocketplane (the Pioneer Pathfinder) by winning a key investment by Northrop Grumman. Ironically, the Rocketplane company and its current rocketplane may be killed by their acquisition of the Kistler albatross.

RpK is currently fighting tooth-and-nail with legal appealsl, hoping that NASA will reconsider its decision to re-award the COTS contract. I don't blame the company, but I think a smarter way ahead would be to drop K-1 entirely and try to turn a profit with XP. The K-1 hardware could be sold, although I'm not completely certain who'd be interested in buying it. Orbital Sciences might be interested in buying the NK-33 engines that RpK owns, for use on their upcoming "Taurus II" rocket (hopefully there will be a better name for it soon.)

At one time, I was very enthusiastic about the K-1 concept. By now I realize the the project has been a costly money-pit. I look at the "Launch Assist Platform" first stage and realize that the propulsive return to the launch site is a waste of propellant mass. While the lessons of K-1 could eventually help RpK to build a smaller reusable spacecraft (utilizing a rocketplane first stage and a ballistic second stage,) it has become an albatross that weighed down and sank a once-promising newspace firm.

Fly Away Home

One of the challenges in building a fully reusable launcher is the recovery of the spacecraft's lower stage (or stages.) Analytically, it's easily demonstrated that a reusable launcher will probably have two or three stages. (A single-stage vehicle would place a horribly tiny fraction of its liftoff mass into orbit, and probably isn't even feasible with contemporary structures and propulsion technology.) But the booster recovery problem is much more difficult than it initially seems.

The most mechanically simple technique is to recover the boosters with parachutes. They'd either splash down in the ocean, or make an airbag-assisted landing on terra firma. The space shuttle SRB's have demonstrated this technique over a hundred times. Our operational experience with this method has taught us much about why we need to find an alternative. The cost of the recovery fleet is considerable. So is the cost of tearing each booster down between flights, shipping it to Utah and back, and reassembling it.

Parachute recovery shouldn't be completely rejected, although there's admittedly a lot riding against it. Rocketplane-Kistler hopes to avoid a lot of the shuttle SRB's problems by bringing the first stage of their K-1 to a landing on dry ground. Assuming that you have a large, open area in your predicted drop zone, I can see the Kistler approach working fairly well. Instead of relying on NASA's SRB recovery flotilla, Kistler could lift the booster with mobile cranes and haul it away on a flatbed truck. It also solves the problems associated with protecting the delicate parts of a liquid-fuel rocket from corrosive seawater.

The next best approach is the flyback booster. Buzz Aldrin's company, Starcraft Boosters Inc, has done a lot of work in this field. One of their important conclusions is that a booster staging at speeds under Mach 3.3 has enough energy to glide back to the launch site; a booster burning out at Mach 6 would require turbofan engines to return to base, but its aluminum heat-sink airframe could survive the expected heating.

In the 1960's, Max Faget's "DC-3 Shuttle" concept made use of a single, very large, winged booster. His proposed craft would have a manned crew, and retractable turbofan engines. The booster would exceed Mach 10 before burning out and falling away. It would have required heavy heat shielding based on its mission profile. The booster could be recovered downrange, then refueled with kerosene and flown back to the launch site like a conventional aircraft.

One idea I've been toying with is a very large booster with a very high staging velocity. How high, you ask? Well, high enough to fly an "Abort Once Around" trajectory and return to the launch site after a single revolution of the earth. The structural requirements and payload inefficiencies of such a booster would be nearly as bad as for a rocket that went to orbit with a single stage. The associated second stage/orbiter would be very tiny in comparison to the orbiter. Yet it has some operational advantages, such as the ability to land without turbines.

How do we proceed with the development of an RLV from here? I think that the StarHawk concept from Starcraft Boosters is a logical first step. The Air Force, under the banner of "Responsive Space Access," should take the lead on this program. While Aldrin's company doesn't have the ability to take lead in building the booster, the Air Force should keep them onboard in a consulting role, and contract the detail design and construction of the booster to the established aerospace firms. StarHawk should even be designated as an "X-Plane."

Building a modest flyback booster like StarHawk will give the US government and industry a lot of experience in designing, building, and operating a reusable lower stage booster. That experience can be shared with the industry and leveraged to build even bigger flyback boosters for rockets that will truly open the space frontier.

Running Down a Dream

Benson Space and SpaceDev have unveiled the third design for their "Dream Chaser" manned spacecraft since the concept was announced in 2004. First they started with a clone of the X-34, as recommended in a paper by Marti Sarigul-Klijn. In November 2005 they changed to the HL-20, which was better suited for orbital spaceflights. They even went as far as negotiations with United Launch Alliance for use of the Atlas V 431 rocket. Yesterday, they unveiled this version of Dream Chaser, based on the X-1 and X-15 rocket planes.

I must say I was a bit disappointed to see the HL-20 design discarded. To be honest, I think Benson/SpaceDev's current design would be better for suborbital space missions, but the HL-20 clone would be better for orbital flights. If Benson Space succeeds in breaking into the suborbital tourism business, they should commission SpaceDev to design an all-new craft by the time they're ready to get into the orbital spaceflight market.

It's all a question of what direction Benson Space wants to go with its space tourism efforts. Do they want to go suborbital first, then get into orbital flights as the technology improves? Or do they want to leapfrog over the competition and go straight to orbit? While the latter strategy could play out in the long run, it would appear that the jump straight to orbital flights would be "a bridge too far" for most new-space companies.

In a way, the decisions facing the New-Space firms are similar to the choices made by US and Soviet space officials in the late 50's. The US committed to a brief suborbital program using the Mercury Redstone. The Soviets had planned on suborbital manned flights with a V-2 variant, but cancelled that program based on the success of the R-7 and Vostok rockets.

Solid Support?

In announcing its memorandum of understanding with United Launch Alliance, SpaceDev revealed that it's willing to fly as many as three solid rocket boosters on the Atlas V that could possibly take the Dream Chaser spacecraft to orbit.

SpaceDev is not the first firm to advocate the use of an Atlas V with solid rocket boosters for manned spaceflight. TeamVision also proposed an Atlas V with four SRB's as a booster for the Orion spacecraft, eliminating the need for an expensive Ares I development program.

Lockheed Martin seemed to have ruled out the use of SRB's when conducting its man-rated Atlas V study. The reasoning was that NASA's specs for safety factors in human spaceflight were exceeded when the additional thrust and impulse of the SRB's was accounted for. Lockheed Martin concluded that only the single-core Atlas V's were suited for manned spaceflight.

The SpaceDev vehicle gets around the NASA safety regulations because it's developed and operated by a private entity. It remains to be seen whether the FAA will hold an orbital space tourism firm to the same standards as NASA.

At the same time, the NASA safety factors should be open to debate. As the TeamVision report argues, the safety factors on an Atlas V 551 were sufficient for an expensive, unmanned spacecraft like New Horizons. Is there any reason why a manned spacecraft with a robust abort system couldn't fly on the same launch vehicle? I suspect that the Redstone, Atlas and Titan II rockets would not have met NASA's current safety standards, either. Yet they worked well when they were employed in a total of sixteen manned spaceflights.

Spaceplanes of Gossamer Wings?

About a month ago, beloved space-pundit Jeff Bell predicted disastrous safety problems for upcoming space tourist craft like SpaceShipTwo and Rocketplane XP. I will be the first to admit that rocket powered aircraft traveling into the exosphere will almost certainly be more dangerous than conventional aircraft. That being said, I also think that Jeff Bell overestimates the dangers of suborbital spaceplanes.

Most of the problems noted in Dr. Bell's analysis of the data show that engine problems accounted for a vast majority of catastrophic rocketplane incidents. Four of these were attributed to the use of ullmer leather, incompatible with liquid oxygen. The second X-15 had an in-flight engine failure, and the third X-15 had an engine explode on the ground. While this is a problem inherent in all rocketplane projects, it must be noted that the X-15 program was conducted at a time when large liquid rockets, especially complex ones like the XLR-99, were in their infancy. As time has progressed, rocket engines have become more reliable. Spaceplane engines need not be any different.

It should also be noted that all of the projects cited by Dr. Bell were research programs that "pushed the leading edge" of flight research. The X-15A-2 suffered structural damage during a high-speed run, the X-2 was lost due to inertial coupling, and the third X-15A spun out of control during re-entry. Flights of commercial spaceplanes will avoid these problems by flying well within the type's flight envelope instead of shooting for extreme speeds and maneuvers that are the norm in a flight research program. I especially think highly of SpaceShipTwo's "feather" maneuver, which takes much of the risk out of reentry.

Dr. Bell mentions the grounding of three X-1 series aircraft due to fatigue in the tankage. This is a problem that cannot be dismissed, but it's not a show-stopper either. Propellant tanks have to be designed to withstand a certain number of pressurization cycles, and they have to be inspected thoroughly between flights. Fortunately, our ability to inspect for microcracks in these tanks is much better today than it was in the 1950's.

I share Dr. Bell's concern with the SpaceShipTwo propulsion system. I studied hybrids for a lengthy period in college. In my mind, the supposed safety benefits never seemed that great, and they never seemed to outweigh their poor performance when compared with liquid systems (or even all-solid rockets, which usually have better specific impulse than N2O-HTPB hybrids.)

Noticeably absent from Jeff Bell's data is any mention of the Douglas D-558-II Skyrocket. This research aircraft flew 313 times using three airframes. None of the aircraft were lost during the highly-successful program. The program even racked up a remarkable aviation achievement: the first Mach 2 flight in history was achieved by Scott Crossfield in the D-558-II on November 20, 1953.

Of the 313 flights, three different propulsion schemes were employed: turbojet-only, turbojet + rocket, and rocket-only with launch from a B-29 mothership. Admittedly, the turbojet-only flights should be tossed out of this safety record, because they're no different from conventional aircraft. Nonetheless, the Skyrocket demonstrated that there's nothing inherently unsafe about mounting a rocket engine with a winged airframe.

As space tourism reaches the suborbital frontier, rocket-powered aircraft will lead the way. Flying in such craft will undoubtedly carry risks, but safe practices and smart engineering can bring risks down to acceptable levels. The passengers of such spaceplanes should go into their adventures with full awareness of those risks, and a belief that their space experience will outweigh any potential hazards.

Path of the DynaSoar

Jon Goff has two good posts (here and here) about the difficulties that went into designing the Space Shuttle, the impossibility of building a fully-reusable shuttle during the 70's, and a possible way ahead.

I have long shared a similar belief. The group that sent Apollo to the moon must have felt very confident in their abilities, but they stretched too far in designing a spacecraft as long as a 737, with a design goal of flying into space every two weeks. If they had attempted building a smaller, reusable spaceplane before proceeding with the space shuttle, they'd have learned some lessons on how to do it. They'd probably go into the shuttle development with lowered expectations for the shuttle, and design the shuttle accordingly.

Specifically, I think the X-20 DynaSoar represented a painful missed opportunity. I came to this belief shortly after Columbia disintegrated, motivated by my admiration of the X-20's hot structure TPS. The metallic TPS and heat-absorbing structure of the DynaSoar would have likely been more robust than the shuttle's tile-on-aluminum construction, at the expense of a much heavier vehicle.

The X-20 DynaSoar went through many changes over its lifetime, but it appeared to have found direction prior to the December 1963 cancellation. Its original billing as a space bomber and recon platform doomed it in the eyes of defense secretary Robert McNamara, who felt that Manned Orbiting Laboratory was a more prudent use of defense dollars. In a military sense, he was justified. But X-20 was still important, in terms of establishing technologies and operational procedures for a reusable manned spaceplane. While X-20 achieved much knowledge that was applied to the shuttle, there was so much more to be learned if X-20 had proceeded with its planned 1967 launch.

I don't like playing "what if" with history, as there are too many variables to truly know how history would have proceeded had a different choice been made. I would say that the following would have been likely outcomes if DynaSoar had been built and flown:
--The shuttle would have been designed for a lower flight rate, with more realistic turnaround times
--The shuttle would not have been designed for large Air Force payloads, and may not have gotten Air Force support at all
--The shuttle would have probably been built with a hot structure and metallic TPS
--Upon seeing the difficulties of operating a reusable spaceplane, NASA may have opted to continue with an Apollo Applications Program utilizing the Saturn IB, Apollo CSM, and small space stations / mission modules instead of pursuing the shuttle

DynaSoar represented the first development spiral of a real reusable spacecraft. Subsequent spirals could have brought us farther down the path to a true RLV, through such steps as:
--Enlarging the DynaSoar glider to carry the engine and propellants for orbital insertion and de-orbit
--Replacement of the SRB's on the Titan booster with liquid-fueled, flyback boosters
--A further enlargement of the glider to include a larger crew and a useful payload
--Development of an all-reusable system with at least one flyback booster (staging around Mach 6) and a spaceplane that could fly all the way to orbit with a useful payload and return safely to earth.

I think most engineers agree that a true RLV was out of reach during the 1970's when the shuttle was designed. It's probably within reach now, but we need to work our way up to that point. Spiral development was the answer then, and it's still the answer today. The problem in the early 70's was a lack of development money. Today, we can find the money if the market sees a need for an RLV. We're caught in a classic chicken-and-egg dilemma: we need an RLV to open space for commerce, but the market doesn't currently exist to justify spending on an RLV. But a first step has to be the development of a reliable manned spaceplane that can be launched on an expendible rocket and returned to flight in a safe manner.