As the United States seeks to extend its human spaceflight exploration capability beyond low Earth orbit (LEO), there is much concern about the evolving space architecture’s sustainability. Gone are the fleeting Cold War political motives for deep-space exploration in the Apollo era. Now we are in this business for the long haul. As Jeff Greason eloquently opined at the 2011 International Space Development Conference, our de facto end game is to settle space. For that, we need an architecture that is politically and economically sustainable.
To achieve this sustainability, we should also be prepared to evolve an architecture that is responsible. At its inception, the heavy-lift Space Launch System () appears to be conceptually responsible. Its crew and payload are not subject to a potential supersonic debris stream during routine launch. In the event of serious SLS malfunction, the crew can call upon independent propulsion to be pulled free of the launch vehicle with a reasonable chance of safe return to Earth. There is hope SLS will evolve exclusively toward fail-safe propulsive systems whose thrust can be terminated at will without issuing a destruct command.
Another aspect of responsible beyond-LEO human spaceflight architecture relates to SLS performance as measured by initial lift capacity to LEO, synonymous with “IMLEO” in rocket science jargon. According to a Sept. 14 NASA release, SLS will evolve from a baseline initial lift capacity near 70 metric tons to 130 metric tons when mature. These numbers currently defy rigorous comparison because they must first be qualified by the assumed height and inclination of the LEO to which they relate. This pedigree is currently unknown to the general public and should be clarified at NASA’s earliest opportunity.
Nevertheless, there is a scenario dictating what minimal initial lift capacity should be provided by a mature SLS in support of responsible beyond-LEO human spaceflight architecture. Consider a time-critical crew rescue or resupply to be conducted somewhere between LEO and the Moon’s vicinity. We know from historic examples there will be compelling pressure to address this scenario in a timely fashion, particularly if crew self-rescue is not possible as it was after Apollo 13’s translunar abort. During Skylab and post-Columbia space shuttle program operations, a strategy known as launch on need was developed to address crew rescue. It entailed processing a potential rescue vehicle for launch virtually in parallel with the primary vehicle that might become disabled. Such processing capability has existed at the Kennedy Space Center’s Launch Complex 39 at least as early as Skylab, and it should be in place when SLS begins launching astronauts. Unfortunately, not one of NASA’s recent beyond-LEO architecture studies has considered the launch-on-need scenario.
Schedule and resources associated with a launch-on-need mission will heavily depend on the anticipated rescue/resupply scenario. For example, a strategy called contingency shuttle crew support (CSCS) evolved post-Columbia using the international space station as a safe haven for the crew of a disabled shuttle unable to safely return to Earth. With the space station as a refuge, weeks to months were available in which to prepare a rescue shuttle mission, eliminating the need for dedicated contingency shuttle processing until after a CSCS scenario was declared. In the case of Atlantis’ Hubble Space Telescope repair mission in 2009, however, no such crew refuge capability existed. Prior to Atlantis’ departure from Launch Complex 39 Pad A for STS-125, the shuttle to be processed for STS-127, Endeavour, was instead prepared as the STS-400 launch-on-need mission at Pad B.
Whenever safe haven infrastructure and cached resources are marginal, as is typically the case early in a specific human spaceflight exploration effort, launch-on-need capability is most desirable. But it is precisely at such times that a higher flight rate is also desirable. A parallel vehicle flow capability conferred by launch-on-need rationale will certainly help enable more frequent launches. By considering the launch-on-need “surge” scenario, ground infrastructure (such as cryogenic propellant generation and storage facilities) is sized with supplemental capacity and redundancy. Launch crews can similarly benefit from surge tasking to support more launch attempts and achieve greater proficiency.
Even with initial lift capacity of well over 100 metric tons from a single launch, interplanetary missions requiring multilaunch campaigns will take on launch-on-need-like surge aspects if a limited Earth departure season develops a slim schedule margin. Tight launch schedules are particularly likely for near-Earth asteroid destinations, the vast majority of which have yet to be discovered. The most compelling of these destinations may not become known and sufficiently characterized for human spaceflight until shortly before a viable mission opportunity arises.
The launch-on-need strategy in a time-critical deep-space rescue/resupply context demands sufficient translunar injection capability from a single launch. To require otherwise would not be responsible, given the delays imposed by multiple launches, rendezvous, docking, assembly and propellant transfer operations. Even if all these operations occur as planned, they introduce a delay of days or weeks compared with a single launch leading to translunar injection. Because multilaunch operations are inherently serial in nature, a single schedule delay at any point only tends to compound translunar injection delays. Furthermore, translunar injection time is fixed by the first launch in a campaign, and another opportunity will not arise for about 10 days if the first is missed.
We know from the Apollo era that translunar injection can indeed be achieved for human spaceflight with a single launch. The Apollo 8 mission even provides us with a quantitative example of the minimal initial lift capacity necessary to mount the contemplated minimal rescue/resupply in cis-lunar space. Data with which to compute as-flown Apollo 8 initial lift capacity were published in Richard W. Orloff and David M. Harland’s “Apollo: The Definitive Sourcebook” in 2006. Summing the Saturn 5 S-4B structure and propellant together with a 28.8-metric-ton command-service module and 9 metric tons of ballast replacing a lunar module results in 127.5 metric tons of initial lift capacity at a circular orbit height of 185 kilometers and inclination of 32.5 degrees. This mission carried a crew of three safely into low lunar orbit and back to Earth. By Apollo 15, Saturn 5 performance and flight profile improvements were achieving initial lift capacity of 140.4 metric tons at a circular orbit height of 167 kilometers and inclination of 29.7 degrees.
As budget cycles, Congresses and presidential administrations come and go, NASA should stay mindful of the launch-on-need scenario and the responsibility it will impart to a sustainable beyond-LEO human spaceflight architecture. We appear to have an initial plan addressing this scenario with a fully evolved SLS, but with very little performance margin to spare. In our desire to begin exploring beyond LEO with human spaceflight missions again, we should be prepared to incur delays and expense until we can do so responsibly. We owe future astronauts and “settlers” nothing less.
Daniel R. Adamo is an independent astrodynamics consultant with research interests in space mission design throughout our solar system. From 1990 until 2008, he supported 60 space shuttle missions from NASA Mission Control’s Flight Dynamics Officer Console. He welcomes feedback at email@example.com.