Well, everyone, despite a truly hell week on the part of both of the author's, it's that time once again. Last week, we reviewed the changes of policy at NASA resulting from the incoming Gore-Richards administration, in particular the elimination of the active pursuit of near-term Mars landings from NASA's goals but a renewed and tightened focus on the lunar return mission. This week, we're going to be looking at what that focus means for the mission itself.
Eyes Turned Skyward, Part 3: Post #7
Although the Richard-Davis report largely spared the Artemis Program the gutting suffered by the Ares Program, it by no means recommended continuing “business as usual” at the Artemis Program Office (now the Exploration Office). Expressing strong dissatisfaction with the pace of NASA’s decision-making, it emphasized the need to quickly begin developing hardware and mission profiles for the sortie missions Gore wanted to see, relegating base development to the future, if NASA performed favorably and budget realities allowed. Although couched in formal language, dense, technical tables, and “sand charts” of budget projections, the message was clear to everyone in NASA Headquarters, Johnson, and Marshall: Get a move on, or else.
However, to be fair to NASA, the questions it had been struggling with since the beginning of Constellation were not
easy questions for an agency aware of its reality as a secondary or even tertiary budget priority and trying to maximize the survivability of its programs in a hostile environment, nor did they have simple technical answers. Even the so-called “mode question,” a parallel to the debates of thirty years earlier that had led to the selection of lunar orbital rendezvous, contained a great deal of complexity if examined closely: How many launches to use for each mission? How to divide the necessary components of the mission between the launches? Where to bring those components together and, if necessary, to take them back apart? How many supplies to provide for each mission? Whether to take those supplies with the astronauts at each step or separate some of them out? None of these questions had an obvious best answer, and, even worse, which answers seemed better than the others was partially dependent on whether one saw the Artemis program primarily as a series of brief sorties to the Moon for scientific and prestige purposes or the beginnings of a base-building effort to parallel Freedom. Given the division within NASA between those favoring the shorter-term approach, often in centers or parts of centers closely involved with Freedom operations, and those favoring a more expansive vision of the program, it was no surprise that the agency had deadlocked on such essentially political decisions. With Gore’s support clearly behind the former faction, the impasse had already started breaking down, even while the Richard-Davies report was being prepared.
Some ground rule assumptions and requirements had already become clear even before Gore’s election. Although the Saturn Heavy was a powerful, capable rocket, it was still considerably less capable and powerful than the Saturn V, which had been only just able to carry out lunar missions itself. Combined with the evolution of safety requirements since the 1960s, an implicit desire to do more than just Apollo redux at the agency, and the unspoken assumption that no new launcher development could be funded, it was obvious that multiple launches would be required for any reasonable mission plan. This, in turn, implied that some location would be needed for bringing together the payloads launched on those multiple rockets and gathering them to form a “stack” capable of landing on the Moon and returning safely to Earth. Given the success of the lunar orbital rendezvous mode in the Apollo missions, it was generally assumed that the lander and return vehicle would be separate, with only the former landing on the Moon while the latter remained in some safe staging area nearby. Finally, a crew of four had been chosen as the default assumption for most studies, with only a few examining larger or smaller teams. With advances in automation since the 1960s, it was no longer considered problematic to allow the entire crew to descend to the surface, leaving the CSM untended. In turn, by adding an additional crew member, every astronaut would have a “buddy” for EVA or other operations, allowing a greater operational tempo than the Apollo missions.
Together, these three assumptions had their own consequences. First and foremost, two Saturn Heavy launches simply could not support a meaningful mission by four people to the lunar surface. At best, using a low lunar orbital rendezvous mode, they could spend no more than a few days on the surface, barely better than the Apollo missions. At worst, if a Lagrange point rendezvous location was selected, the crew might not be able to spend even one full day on the surface. In both cases, little more would be achieved by any lunar mission than had been done on a given Apollo mission, leading to the obvious question of why billions of dollars were being spent to recreate missions from thirty years earlier. The minimum number of launches needed for a mission was therefore three. Since the Kennedy Space Center had only two pads capable of supporting Saturn Heavy launches, at least one of those launches would need to take place a few weeks before the others. In fact, to best fit in with the center’s processing flow and minimize the amount of extraordinary effort needed to ready pads in quick succession, it would be better if it took place several months before the other two launches. In consequence, the payload launched on this first flight would need to be something that could tolerate several months in space--ruling out cryogenic liquid hydrogen or liquid oxygen, which made up the bulk of the launched weight--and which could easily be separated from other mission elements that would have to launch just before the mission itself, such as the Earth departure stage or the crew. The obvious answer was to launch the supplies needed for the desired longer missions on a separate lander, reducing the crew lander to little more than a lightweight taxi for transiting to and from the lunar surface, able to be launched on a Heavy with the crew vehicle and carry out a “two-Heavy” mission with a separately launched Earth departure stage. Since a logistics lander would be needed for a permanent base, to land supplies without the expense of a human flight and to transport large base modules and equipment, this plan gained immediate support from the pro-base contingent of NASA’s personnel. Although the pro-sortie club was more reluctant to follow, eventually they, too, conceded that it was at least acceptable, and this general plan had already started to become the default before Gore’s election.
It proved more difficult to resolve the question of where to stage from “nearby” the Moon. The Apollo missions, of course, had had their lander and return spacecraft separate and eventually rendezvous in low lunar orbit, and at first most mission plans followed suit, happy to trust the judgement of the men who had actually landed men on the Moon. As more in-depth analysis took place, however, problems in the low lunar orbit profile began to appear. Modern mission planners wanted access to the entire Moon, not just a narrow band of sites near the equator, especially in the wake of the Lunar Reconnaissance Pioneer’s apparent discovery of large deposits of water ice near the poles and the presence of a gigantic impact basin of scientific interest near the South Pole on the far side. Eliminating the communications problems was easily achieved by inserting satellites into lunar orbit to relay signals from astronauts on the far side, but the equatorial orbit used by the Apollo missions could not reach many of the more interesting sites. An increase in the delta-V budget could allow choosing an arbitrary orbit passing over any part of the Moon, but this itself led to further problems. Since the 1960s, safety standards had become more stringent as more had become known about the dangers of space, and as part of any future Moon missions it was desired by certain parts of the agency that the astronauts be able to choose to abort their mission at any time and return to Earth, a capability which became known as “anytime return”. It quickly became apparent that orbital mechanics meant that providing this capability was going to require a substantial amount of delta-V on the return vehicle, on top of the already large amount needed merely for escaping lunar orbit in the first place. Since the return vehicle was supposed to be at most a variant of the spacecraft used for crew transport to Freedom, and since these requirements were much larger than needed for the low Earth orbit maneuvers needed for that role, designers were left with the unpleasant dilemma of either accepting a mass and cost penalty for low Earth orbit missions because of a larger, more expensive service module than needed, or accepting the expense of designing and manufacturing two different service modules, one for lunar and one for Earth orbital missions.
However, while studying possible communications relay satellite locations, a Langley astrodynamicist had stumbled over an interesting observation--the issues with adding “anytime return” for low lunar orbit wouldn’t apply to a vehicle staged out of the second Earth-Moon Lagrange point, or EML-2, a region where satellites could remain hovering over the farside with relatively small stationkeeping requirements. Exploring trajectories to and around halo orbits around EML-2 for farside communications using work by Robert Farquhar in the late 1960s, Abe Lewis observed that a hyperbolic trajectory to these halo orbits consumed only slightly more delta-v than the trans-lunar injections of Apollo, while the fixed position of EML-2 relative to the moon and the much easier trans-Earth injections essentially “baked in” anytime return with much less delta-v requirement, especially on the return spacecraft. This solved in a single step the dichotomy that had been facing mission planners between the required performance required by the Earth orbital and by the lunar orbital missions. The tradeoff was that the lander would require more performance, both on the descent and on the ascent, and thus a heavier lander would be required to place payloads onto the lunar surface. However, Lewis calculated that the increases were not enough to outweigh the benefits of these EML-2 trajectories, and showed so in an impressively exhausting series of head-to-head comparisons of notional missions, comparing his conceptual designs against other NASA design reference missions for the moon. In these analyses, another benefit emerged: the large descent stage needed for the EML-2 mode was well suited to be a logistics lander, provided the necessary electronics and equipment were baked in rather than located on the ascent stage, turning a potential drawback into something of an advantage. Like Houbolt in Apollo, others were considering EML rendezvous before Lewis began his work and the influence of one man in bureaucracy as large as NASA can be hard to judge, but the EML-2 rendezvous gained much attention, and studies similar to Lewis’ side-by-side comparisons soon emerged from the main Artemis Office. Within months, EML-2 staging had begun to dominate Artemis reference missions.
Thus, the final Artemis architecture emerged. A three-launch mission would occur, with the first launch sending a logistics lander via a Saturn Heavy directly to the landing site. Several weeks later, with the cargo lander confirmed to be safely on the surface, a pair of Saturn Heavies would carry aloft the crew portion of the mission: one with a large hydrogen/oxygen departure stage, the other with the Block V Lunar Apollo and crew lander. These would meet in LEO, with the departure stage expended to put the stack into a path to EML-2. From there, the crew would descend to the surface in the lander, using supplies from the cargo lander for stays lasting up to 14 days, then ascend back to EML and return to Earth aboard their Apollo. Originally, 8 lunar flights were planned, requiring three new hardware elements: the lander, the new lunar Apollo, and the large EDS (named internally the Exploration Cryogenic Upper Stage). Each mission was to cost around $1.5 billion, with development costs and surface hardware bringing the Artemis initial sortie program to about $20 billion. Flights would begin in 1999 and continue at a pace of one per year until 2007, NASA’s bid to both smooth out budgetary requirements and allow a building of support for permanent bases. These plans were reflected in the budget recommendations Lloyd Davis brought to President Gore in late 1992 for the FY 1993 budget process. However, it has been said that no plan survives contact with the enemy, and in order to be approved, these recommendations would have to pass through the halls of the United States Congress.
Roughly speaking, Congress broke into four groups on the matter of spaceflight. One could be termed the “hawks”--largely interested in seeing the US space program continued in full force. Not coincidentally, these tended to be representatives from Florida, Alabama, and other states with large vested monetary interests in the US space program, but the memory of Vulkan Panic’s arrival after US space spending was decreased after Apollo still hung in the minds of a few other members concerned about the growing Chinese program. The second group, for a variety of reasons, saw the space budget as a massive target--either to shrink the government overall, or to be redirected to the member’s preferred programs. The third group was essentially a mix of both--worried about the United States losing its place in spaceflight (both manned and commercial) to Russian, Chinese, or European competition, but conscious of the price tag associated with the endeavor in an era focused on “reaping the peace dividend” and shrinking spending. The fourth group, and by far the largest, honestly cared only as far as the topline numbers, and was lead by whichever messages emerged from the most influential of the other three groups--particularly the third. Gore’s proposed plans, as encapsulated in the Richards-Davis Report, had therefore been calculated to appeal to this group--in his time in the Senate, Gore had plenty of experience in the way things worked, as Davis himself had in NASA dealing with the OMB. In order to re-assure the more hawkish tendencies, Davis’ advocacy of the new plan on the Hill focused on selling the budget savings of cutting Ares and of co-operation with international partners on the precursor missions, the benefits of the station crew exchange program on keeping Russian rocket engineers working for Russia and not rogue states, the potential benefits of Gore’s commercial initiatives for assuring continued US success in the commercial market even in the face of Chinese, Russian, and European competition, and the newly enhanced focus of Artemis ensuring that the money spent would produce results. In the large sense, the sales pitch was effective, as the general outline of the new direction was approved in the new Authorization bill, while Appropriations roughly followed suit. However, there was sacrifices that had to be made. To appeal to the budget cuts, the final two Artemis missions were cut to bring the program lifetime cost down to just $17 billion, shortening the initial sorties to just six flights ending in 2005. Additionally, to secure approval for Gore’s forward looking commercial development with a little precautionary protectionism, new teeth were granted to export controls of “defense technologies,” which were expanded to include launch vehicle and satellite technologies. While not actively preventing such exports, the new approvals required to export such technologies (which would include, not coincidentally, launching US satellites on foreign vehicles) were intended to discourage and otherwise limit such activities.
With the missions approved and money flowing, the contracts for the three major hardware elements could be let. Rockwell’s receipt of the “Lunar Crew Vehicle” contract for the uprated lunar Apollo was almost a formality--the mission plan’s preference for an Apollo closely related to Block IV was well known in the industry. Essentially, the final proposal would mate a Block IV Apollo CM to an SM based closely on the existing Block II Aardvark SM, allowing more room for fuel, together with a “lightweight” pressurized module to provide additional space and services--most prominently a proper toilet--during the flight to and especially from the Moon. The largest change would be overhauling the power system--for the near month of total operations expected of Artemis-model Apollos, batteries would be impractical. Instead, the Block V would introduce much smaller batteries, kept charged by solar arrays. The spectacular improvements in solar cell efficiency since the 1960s had made the conversion an “also ran” on every new block of Apollo since the 70s, and the lunar mission requirements finally pushed solar panels ahead of simply maintaining the proven and effective battery system. Given this and the intention to roll the conversion out across both lunar and Earth-orbital Apollos, the Rockwell contract (at $400 million) was slightly more expensive than might have been expected for simply “another Apollo,” but the process was both smooth and cheap compared to the contracts for the ECUS and the lander.
The lead competitors of the ECUS contract were mostly confined to companies already constructing hydrogen stages, namely McDonnell of the SIVB/C family and Northrop of the Centaur (brought in from General Dynamics when Northrop acquired them). While other companies including Lockheed and Boeing submitted bids, the experience of these firms was enough to push their proposals into the lead. Both stages were planned to use the same engine cluster--six RL-10s--and to use common bulkhead designs to minimize dry weight. However, the designs differed in the key detail of diameter. The Northrop design was set at 5.5m diameter, essentially replicating the S-IV stage of the 1960s with an improved mass fraction and higher overall fuel load. McDonnell, on the other hand, set about encapsulating the ~70 tons of propellant in a 6.6 meter tank based on the proven SIVB derivatives they had developed. In order to build a small enough LOX tank, this then required flipping the common bulkhead’s dome to nest “into” the aft LOX dome--a major revision to the common bulkhead design, requiring new structural analysis, a slightly heavier common bulkhead dome, and substantial engineering costs. Compared to this, the new tooling required for Northrop’s overgrown Centaur was judged less technically risky, and Northrop’s bid cost ended up being slightly lower. In the end, it was a deciding difference--McDonnell's contributions to Artemis would be limited to Earth orbit with their SIVC on the Saturn Heavy. Northrop’s design, which they saw as giving “wings” to the Artemis program, was named “Pegasus” after the winged horse of mythology. Northrop’s contract for the development of the stage was set at $1.2 billion, and was a major win--a chance to gain NASA funding to build their own large-stage toolings.
The lander competition was equally fierce--while the product was less commercially applicable than a large hydrogen stage, the lander was viewed as higher prestige. However, experience in lander technologies was less widespread, putting most proposals on more equal footing, with one major standout. With the experience brought in by their new Bethpage division and Starcat, Boeing had very recent history with a vertically-landing hydrogen vehicle. Moreover, the institutional memory of Grumman on the Apollo Lunar Module gave a base to build this more recent experience on. Their entry (1) was far more “conventional” than many put in by other companies, consisting of stacked ascent and descent stages. However, this created an issue of reaching the surface--the porch of the ascent stage would be more than 6 meters off the ground, requiring quite a bit more than “one small step.” Other entries were more creative in order to eliminate this issue. Several turned the lander’s launch axis horizontal. While some simply mounted the engines perpendicular to the launch axis (2), some variants on this concept used separate descent and landing engines, with the main descent performed by a larger engine mounted along the axis, then smaller engines for final descent--thus avoiding the issue of deep throttling for the main engines (3). Others used a sort of “crasher” design (4), with the descent stage doing most of the work of landing, but the ascent stage then actually landing separately, performing final descent as well as ascent, eliminating any need to climb down the descent stage to the surface and any need for equipment such as landing gear on the main descent stage. However, in spite of this, Boeing’s Grumman experience helped the technical maturity and NASA’s judgement of the risks of the design, and it was enough to win them the $5 billion of the Lunar Crew and Logistics Module (LCLM) contract.
With congressional approval secured and contracts settled, the doldrums that had gripped Artemis were largely eliminated. Most shocks to the program caused by the cuts and re-arranging of the Artemis and Ares Offices into the Exploration Office were eased by the focus on Artemis that Davis brought, and the measurable progress made in 1993. Across the country, work on Artemis was beginning to grind into gear. From being nothing but a distant possibility a few years earlier, now a return to the moon seemed to be drawing ever-nearer for American astronauts.