EELV or never?
The Air Force half of the Moorman recommendations became manifest in the form of the EELV program. A classic development and acquisition approach was envisioned. The initial phase would consist of concept development contacts awarded to multiple contractors, followed by a downselect to the two most promising designs. The next phase would see the two selected contractors further develop their concepts. Finally, one firm would be selected to build their vehicle for the Air Force, winner take all. All of this was going to be done at total budget about one half to one third of what ALS/NLS was to have cost. The ultimate objective was similar to that of ALS/NLS in that it was intended to produce a launch system that cost about half of what existing ELVs did. As things would turn out, the program’s procurement strategy would go through a series of acrobatics that would do credit to the Thunderbirds, while the cost aspects would both be bettered significantly and at the same time fall far short.
Four companies were selected for the initial phase of EELV and they all faced the same huge problem. The main driver in any launch vehicle was, quite literally, the propulsion system. That vital element largely determined the design of the vehicle. The biggest problem that every company faced was that there were very few good choices available.
Discounting the limited and aborted NLS/ALS efforts, no significant new work had been done in the US on really new booster propulsion systems for an incredible twenty years. The Space Shuttle Main Engine (SSME) was intended to be literally the end-all that did all, with the Shuttle replacing all expendable boosters. The pre-Shuttle systems used technology fresh out of the 1950s. The other element of the Shuttle propulsion system, the huge segmented solid rocket boosters, was not embraced by anyone; in fact, segmented solid motors were seen as something to avoid by most people. The available upper stage propulsion systems were little more advanced than the first stage systems, dating from the 1960s and ’70s.
One of the winning bidders, Boeing, proposed to use the newest available US first stage engine, the SSME, and recover it via parachute. Another bidder, Alliant, proposed to use the newest available US propulsion system, the solid motors developed for the Titan 4B strap-ons. The two other bidders proposed truly gutsy designs, one embracing the “evolved” concept and the other all but tossing it out.
Lockheed Martin, builder of the Titan series of boosters, and who eventually would take over Atlas production from General Dynamics, proposed building a booster that utilized a modified version of the Russian RD-170 engine. They proposed to essentially take one-half of an RD-170, called an RD-180, and introduce it in a modified version of the Atlas. The venerable Centaur upper stage of the Atlas would be modified as well, and eventually the Atlas stainless steel balloon tank would be replaced with rigid aluminum structure. A lower performance upper stage called Agena 2000 would be available as an alternative to the Centaur as well, although it would be much closer to a Delta 2 second stage in design than the old Lockheed upper stage that had not been flown for years. Some Titan 4B checkout equipment would be used as well, making the vehicle truly evolutionary in nature, and with elements from three different booster systems.
McDonnell Douglas took still another approach for a propulsion system, the most radical of all. The SSME was the most advanced US engine available, and a new engine, the RS-68, would be based on that technology. The second stage would use a variant of the Centaur’s RL-10 engine, and a lower-performance version using a modified Delta 2 second stage would be an option as well. Everything else was new; there was little or no use of existing hardware, unless you counted simply use of some older technology.
It’s fixed—and it’s back!
Having solved the requirements problem, once the Air Force got the EELV program underway, they came to realize that the problem was still there.
NASA was no longer a playe, and in fact refused to be one, initially not even committing to purchase boosters developed under the EELV program when and if they became available. Of course, with the X-33 program underway and paralleling EELV, NASA was once again working to make all expendables obsolete, so they may have thought it was pointless to think about using new ELVs anyway. In any case, the agency was committed to a “commercial” approach in booster procurement; they would buy whatever they thought best.
The other problem—even bigger than NASA—was the commercial market. The Air Force could not ignore the commercial marketplace, having recognized that sales of boosters to private firms helped spread the programs’ costs over multiple users. However, a number of mechanisms—and the lack of others—stopped commercial requirements from being included in EELV, at least as such. There was no process by which the Air Force could go to commercial industry and ask for a formal set of requirements. Even if there had been such a methodology, such requirements would have equated to the government funding commercial needs—and in terms of support to private industry that is called a subsidy. Even if the Air Force had been so magnanimous to include commercial requirements in its EELV specifications, the auditors at the Pentagon would have ripped those right back out. And if the green eyeshade guys in DoD had let that indiscretion get by, Congress surely would not have.
So the Air Force took a guess at enveloping the commercial requirements. Such a move was wise, and both noble and necessary—and they blew it.
NLS/ALS had envisioned a family of launch vehicles characterized by both produceability and a stair-step approach to performance. There would be a basic core vehicle to which could be added a single identical core for greater performance. Still another identical core could be added for the highest performance version. But EELV was more or less ALS/NLS on the cheap, and such a graduated performance envelope was thought to be a luxury. The Air Force defined two broad categories of performance requirements: 9,100 kilograms (20,000 pounds) equivalent to Low Earth Orbit (LEO) and 18,200 kilograms (40,000 pounds) equivalent to LEO. The 20K requirement was greater than anything except Titan 4 and Shuttle could handle and at the time was thought to envelope all medium launch vehicle requirements quite nicely. The 40K requirement matched what Shuttle and the big Titans could handle and was thought to be both readily achievable in terms of technology as well as essential to the most critical DoD payloads.
A traditional method of handling stairstepping performance requirements was use of solid motor strap-on stages, but by the early 1990s solid strap-ons had earned a bad reputation. Neither the Air Force nor the winning bidders for the second phase, Lockheed Martin and McDonnell Douglas, desired to use solid strap-ons.
The problem was that the commercial requirements as guesstimated by the Air Force at the start of the EELV program did not stand still. While the Air Force had been criticized on occasion for its apparent devotion to increasing the size and weight of payloads, the commercial marketplace followed much the same programmatic trajectory. Private firms found that it was more profitable to launch larger, more capable, and longer-lasting satellites than it was to keep launching smaller ones. The advent of commercially available foreign boosters—Ariane, Proton, and the Long March series—meant that the commercial satellite firms could design to business requirements rather than just American capabilities. The private satellite firms broke the 20K barrier and could be expected to build even heavier.
By the time the increased commercial requirements came to light, the Delta 4 and Atlas 5 were already pretty well defined when it came to performance, driven by the engine choice. Both companies planned to meet the 40K requirement by strapping together three first stages, but that solution was far too capable and thus too costly for commercial users. So, both companies added solid motor strap-ons to their designs, thereby rescuing the vehicles’ commercial viability.
As things turned out, commercial payloads proved to be the schedule driver for both the Delta and Atlas. Although Delta 4 won the lion’s share of military payloads, the first launch carried a commercial satellite—and used solid strap-ons. The first Atlas 5 use of strap-on motors did not occur until its third flight, but now the booster has flown more missions with strap-ons than without. Even more significantly, four of the nine EELV missions flown thus far with solid strap-ons were government missions. When the EELV companies modified their vehicles to meet commercial requirements they ended up helping government users as well.
NASA finally bought EELV boosters, coincidentally or not only after abandoning the new RLV that had been started at the same time as the Air Force program. But NASA’s lack of input to the EELV requirements process had its impact. Both the Delta and Atlas were designed to save money by flying lofted trajectories not particularly well-suited to carrying winged Shuttle replacements in particular or manned payloads in general. At best, it will take some work to make them capable of carrying manned vehicles; at worst it will take a clean sheep of paper.
Soaring toward the future, or weighed down by the past?
Based on fifty-plus years of history, we can conclude that Delta 4 and Atlas 5 will indeed be around for a long time. That in itself is neither bad nor unexpected. At least we are no longer utterly dependant on first-generation ICBM technology. But perhaps we should consider the new vehicles’ true ancestry.
Delta 4 is built around an engine derived from the Space Shuttle, a vehicle that proved to be so complex and costly that no one is even considering building anything like it ever again. Atlas 5 is designed around a engine developed to launch the Soviet copy of the Shuttle, a vehicle that would never haven been developed had the US not built the Shuttle, and which almost immediately was recognized as being something less than useless. The Soviet shuttle, Buran, flew but once, unmanned. At least the surviving US shuttles are flying.
So, Delta 4 and Atlas 5 arrived decades later than they should have, are less capable than they should be, cost more than they ought to, and are based not on new cutting-edge technology designed to conquer the universe but on old hardware developed for a failed concept and its foreign copy. And they are still, by far, the best we have—and that in its own way is rather sad.