John London's LEO on the Cheap is a unique study of the opportunities for low-cost
transportation to low Earth orbit, and why government and industry went the other way.

Go to
Foyle Publishing.

LEO on the Cheap

Methods for Achieving Drastic
Reductions in Space Launch Costs

John R. London III, Lt Col, USAF

Research Report No. AU-ARI-93-8
Air University Press
Maxwell Air Force Base, Alabama

October 1994

By permission of Air University Press

Table of Contents

Chapter 1        The Problem
Chapter 2        Existing Launch Systems
Chapter 3        Proposed Launch Systems
Chapter 4        Causes of High Launch Costs
Chapter 5        The Necessity for Complexity Myth
Chapter 6        Some Key Design Choices
Chapter 7        Cultural Changes
Chapter 8        Booster/Spacecraft Cost Relationships
Chapter 9        Minimum Cost Design Launch Vehicles
Chapter 10      Conclusions and Recommendations

Chapter 9 - Minimum Cost Design Launch Vehicles

The idea of building simple and inexpensive launch vehicles is not new. The seeds were planted in the late 1950s for minimum cost launch vehicle designs. By the latter half of the 1960s, a number of aerospace companies were busily seeking a solution to the high cost of space launch by studying minimum cost boosters. All of these efforts died when the US government decided in the early 1970s that the Space Shuttle would be the long-term solution to high launch costs. Although several minimum cost booster concepts have been proposed since 1972, none have been seriously considered by DOD or NASA. However, the high and ever-increasing cost of current launch systems, combined with the failure of new initiatives like the ALS and NLS to gain continued finding, makes the idea of a launch vehicle designed for minimum cost more relevant today than it has ever been.

To appreciate the depth to which Minimum Cost Design (MCD) methodology has been investigated and how well it has been validated, one needs an understanding of the booster studies that have been accomplished over the years (see Table 9). Therefore, this chapter will provide a survey of significant past studies and some current programs that employ aspects of minimum cost design methodologies to develop low cost launch systems.

Table 9. Minimum Cost Design Launch Vehicle Concepts

Sea Dragon
356.0 MN (80.0 Mlb)
544,000 kg (1,200,000 lb)
Aerospace Design 3 SLV
16.671 MN (3.748 Mlb)
18,145 kg (40,000 lb)*
$880/kg ($400/lb)*
Chrysler MCD Booster
21.526 MN (4.84 Mlb)
45,360 kg (100,000 lb)
$752/kg ($341/lb)
McDonnell Douglas MCD Booster
21.5 MN (4.63 Mlb)
45,360 kg (100,000 lb)
$767/kg ($348/lb)
Rockwell MCD Booster
20.1 MN (4.52 Mlb)
20,400 kg (45,000 lb)
$1,381/kg ($626/lb)
Martin Marietta MCD Booster
25.9 MN (5.823 Mlb)
20,400 kg (45,000 lb)*
$474/kg ($215/lb)*
Boeing Double Bubble Booster
12.0 MN (2.7 Mlb)
15,420 kg (34,000 lb)
$936/kg ($424/lb)
Revised Boeing MCD Booster
45,360 kg (100,000 lb)*
$1,437/kg ($652/lb)*
TRW MCD Booster
51.73 MN (11.63 Mlb)
60,000 kg (133,000 lb)
$1,235/kg ($561/lb)
30.25 MN (6.8 Mlb)
29,756 kg (65,600 lb)
$1,989/kg ($901/lb)
*Payload capacity and launch efficiency values are for an LEO polar orbit.
**Costs are in 1993 Dollars

Sea-Launched Space Booster Studies

In the late 1950s, the idea of launching rockets directly out of the ocean emerged. Ignition of the first-stage engine(s) was to take place underwater. The US Navy initiated the HYDRA program to demonstrate the feasibility of launching rockets that were partially submerged and floating vertically in sea water. The 32-meter (105-foot)-tall solid propellant HYDRA-1 was launched in March 1960, directly out of the ocean off Point Mugu, California. The Navy conducted approximately 60 launches of rocket simulators and actual rockets over the course of the project, using mostly solid propellant propulsion systems. HYDRA validated the concept of launching directly from the sea, with the rocket's initial exhaust gasses being expelled directly into water.1

During the same time that the Navy was conducting the HYDRA tests, Aerojet-General Corporation accomplished a series of tests to study the feasibility of sea-launched liquid propellant rockets.2 The Aerojet effort, called the Sea Launch Program, was dubbed "SeaBee" because it used a modified Aerobee 100 sounding rocket for its test vehicle.3 Aerojet conducted a number of demonstrations of ocean launching techniques to evaluate handling, propellant servicing, checkout, and sea launch operations. Aerojet also evaluated recovery, refurbishment, and relaunch of the test vehicle, with an eye toward future reusable launch systems.4

Aerojet successfully launched the SeaBee test vehicle on 24 October 1961 from a floating position off Point Mugu. It reached an altitude of 1.5 kilometers (5,000 feet), deployed a parachute, and was safely recovered after a water landing. Having sustained no damage, the SeaBee was refurbished and relaunched on 2 November 1961. The success of SeaBee helped substantiate the concept of sea launch and recovery for a much larger launch vehicle proposal.

Aerojet used some independent research and development funding in the early 1960s to explore various cost aspects of space launchers. Through these studies, the corporation developed a set of five design rules for low-cost launch vehicles. The low-cost booster must be big, simple, and reusable. Also, the design must not push for the absolute maximum reliability, and it must not push the state of the technological art.6 Aerojet combined data derived from the SeaBee program with the newly developed low-cost booster design rules to define a colossal launch vehicle. Called Sea Dragon, it was intended to support NASA's manned exploratory assault on Mars and interplanetary space (see Table 9).7

The Sea Dragon was to be a simple, reusable launch vehicle. Like the SeaBee, it was to use a pressure-fed propulsion system; but it was scaled to represent perhaps the largest space booster ever conceived.8 It was to have a lift-off thrust of 356 million Newtons (80 million pounds) and a lift capacity to low earth orbit of 544,000 kilograms (1,200,000 pounds).9 The Sea Dragon was to be 168 meters (550 feet) tall and to have a diameter of 23 meters (75 feet).10 Construction and transportation of such a booster was more amenable to a shipyard than an aerospace factory, and the vehicle's simple steel design with water launch and recovery made shipyard manufacturing appropriate and practical.11

Figure 7. The Sea Dragon launch vehicle concept, illustrating
the first stage recovery via water splashdown

Aerojet designed the Sea Dragon to have two stages. The first stage would use liquid oxygen and RP-1; the second stage, liquid oxygen and liquid hydrogen. Both stages would be pressure-fed, and both would use a single-engine thrust chamber. The first stage engine would be rated at 356 million Newtons (80 million pounds) of sea-level thrust--certainly the largest rocket engine ever seriously postulated.12 Aerojet settled on single-thrust chamber stages because their studies indicated it would be less expensive to develop and integrate single large engines than to develop and cluster sets of smaller engines. Also, analysis showed that even with near-exponential increases in the size of simple engines and airframes, there is only a linear increase in cost.13 The analysis results made a strong case for the economy of very large and simple boosters with large engines, and Sea Dragon was the consummate embodiment of this design philosophy.

Sea Dragon was to be constructed-and transported to the launch location (at sea) in a manner that was closer to a seagoing tanker than an airplane. The vehicle would have been built horizontally in a commercial shipyard, then staged out of a US coastal site. It was to be fueled with RP-1 in a dry dock, then towed horizontally to the launch point. Upon its arrival, propellant transport ships would have loaded the vehicle with cryogenic propellants, and technicians would have flooded a ballast device to position the booster vertically. The booster would jettison the ballast at lift-off.14

The first stage, which was to be recovered several hundred kilometers downrange, would use an inflated drag chute to decrease its water-impact velocity. The rigidity and strength of the heavy steel tankage, which was designed for the pressure-fed propulsion system, would have lent itself to surviving repeated water impacts with little damage. The second stage had an optional reusability design that would have employed retro-rockets, an ablative nosecap, and a drag-inducing device for controlled reentry to a point close to the refurbishment site.15

Cost estimates for using the Sea Dragon to place a payload in low earth orbit ranged from $59 per kilogram ($27 per pound) to $620 per kilogram ($282 per pound).16 The booster researchers were able to project these low costs because the booster had the benefit of a significant economy of size, it depended on shipyard-type (as opposed to aerospace) construction techniques, and it was reusable.

The Sea Dragon was designed prior to formal codification of the classical design-for-minimum-cost (DFMC) methodology by The Aerospace Corporation. Nevertheless, its design contained the essence of the DFMC philosophy and therefore represented the first detailed launch vehicle concept that was designed for minimum cost.

After Aerojet proposed the Sea Dragon concept, NASA's Marshall Space Flight Center contracted Space Technology Laboratories, Inc. (a subsidiary of TRW) to evaluate the proposal and re-accomplish the cost estimates. Space Technology Laboratories largely confirmed Aerojet's cost data and the soundness of the design.17 However, NASA's interest in the concept was primarily driven by the vehicle's massive lift capacity rather than its low cost. As the scope of NASA's interplanetary ambitions shrank, Sea Dragon was shelved and virtually forgotten.18

Early Air Force and NASA-Sponsored Studies

Studies that culminated in a formal design-for-minimum-cost criteria began in 1959 at the Space Technology Laboratories (later to be incorporated into TRW). By the fall of 1963, the Air Force and The Aerospace Corporation had initiated studies that applied the DFMC criteria to advanced ballistic missile concepts.19 These efforts focused on the definition of a semi-mobile ICBM.20 (Minimum cost design [MCD] ballistic missiles are still very viable weapon system concepts, particularly when considered for the delivery of conventional munitions in a limited conflict.21) Much of the data generated by these studies was applicable to space launch systems, and the Air Force and Aerospace gained important insights into the implications of applying the MCD methodology to launch vehicle designs.22

Initial MCD Booster Designs

Between 1965 and 1968, The Air Force's Space and Missile Systems Organization (SAMSO) and The Aerospace Corporation, which were collocated in El Segundo, California, began to formally apply the DFMC criteria to space launch vehicle designs. The Aerospace Corporation developed a baseline two-stage pressure-fed MCD booster concept with an 18,145-kilogram (40,000-pound) lift capacity to a polar low earth orbit (LEO). The vehicle was designed to use a hypergolic propellant combination.

The concept evolved through three major design iterations. The Design 3 space launch vehicle had a gross lift-off weight of 1,114,924 kilograms (2,458,000 pounds) and a first-stage sea-level thrust (using a single engine) of 16,671,104 Newtons (3,748,000 pounds).23 Taking a cue from the Sea Dragon concept, Aerospace designed the MCD booster's first stage to be reusable, using an inflatable drag-inducing device and ocean splashdown for recovery.24 The design projected a recurring launch cost of less than $880 per kilogram ($400 per pound) to polar LEO (see table 9).25 The moniker "Big Dumb Booster" was unofficially, and sometimes irreverently, applied to this particular vehicle configuration, although it has also been used to describe some other simple, low-cost launch vehicle designs.26

MCD Application Studies for the Titan III. In August of 1965, the Titan III System Program Office and Aerospace, with support from the Air Force Rocket Propulsion Laboratory, began a series of studies that applied the MCD criterion to a variety of potential space booster requirements. Initial studies defined a baseline reference design for an MCD launch vehicle, considered replacing the Titan IIIC core vehicle with an MCD core vehicle, evaluated replacing the Titan IIIC solid rocket motors with MCD liquid strap-ons, compared MCD liquid strap-ons with hybrid strap-ons, and considered replacing the Titan IIIB with an MCD booster of comparable performance.27

Alternative MCD Titan Core Vehicle. In conjunction with a Martin Marietta study to develop a large-diameter core vehicle that would increase the capability of the Titan III vehicle family, Aerospace studied an alternative MCD core vehicle. The Aerospace design used a lower strength alloy, had simple load paths, and required no machining of vehicle skins. Compared to the Martin Marietta design, which used a minimum weight criterion, the Aerospace MCD core vehicle weighed only 7.2 percent more but had a recurring cost decrease of more than 50 percent.25

Figure 8. The Aerospace Corporation Design 3 space launch vehicle configuration

MCD Concepts for Titan SRM Replacement. Martin Marietta also studied two different solid rocket motor (SRM) growth options to increase the capability of the Titan IIIC. Aerospace considered an alternative approach, using two different liquid propellant strap-on concepts that were designed to the MCD criteria and equaled or exceeded the capabilities of the Martin SRM proposals. Projected recurring costs for both of the MCD liquid strap-on concepts that were less than half the corresponding SRM recurring costs.29

Comparison of MCD Liquid Strap-ons and Hybrid Strap-ons. In the latter half of 1967, Aerospace engineers compared their MCD liquid strap-on design to a United Technology hybrid booster concept that had been proposed as a strap-on for the Titan IIIC. The MCD liquid design projected recurring costs that were 25 percent less than those for the hybrid design.80 (It should be noted that the United Technology hybrid motor design was significantly different from the current AMROC design.)

Titan IIIB Replacement Studies. In early 1968, Aerospace undertook a design study for an MCD booster to replace the Titan IIIB launch vehicle. The MCD design extrapolated Aerospace's Design 3 Space Launch Vehicle concept to approximate the performance of the Titan IIIB. The focus was on structure and propulsion; Titan IIIB systems such as guidance, telemetry, and power supply were retained in the MCD booster design. The Agena third stage and its adapter, which were also used on the Titan IIIB, were incorporated in the MCD design. Projected recurring costs for the MCD booster were significantly lower than Titan IIIB recurring costs. The MCD booster's stage one and stage two tankage cost was less than $160,000 per vehicle.31

The Titan IIIC launch vehicle.

A Titan IIIB lifts off from Vandenberg AFB.

Collateral Developments. All of these studies indicated that application of MCD criteria would result in significant launch cost reductions.32 During this same period, TRW was experiencing significant success with engine development activities that used a simple and inexpensive pressure-fed design based on their Lunar Module Descent Engine. Also, U.S. Steel's new HY 140 alloy steel appeared to be ideal for MCD pressurized tankage applications. These positive indicators served to add momentum to the Air Force/Aerospace MCD booster initiatives.83

Initial Industry Studies

In May 1968, SAMSO and The Aerospace Corporation conducted briefings for industry to stimulate interest in developing MCD space launch vehicles. TRW and Boeing were the most enthusiastic about the concept.34

While Aerospace continued to develop MCD booster concepts, a number of aerospace contractors (including TRW and Boeing) began to conduct minimum cost design studies and to develop candidate vehicle designs (see table 9). These contractor studies, which were funded by independent research and development dollars, were not constrained by fixed requirements.35

Chrysler MCD Booster Concept. In February 1969, Chrysler Corporation's Space Division submitted a concept for a minimum cost design booster with a pressure-fed first stage. The design called for a two-stage launch vehicle, but the MCD focus was on the first stage. Chrysler called for the second stage to be a Saturn S-IVB (the third stage of the Saturn V launch vehicle), which was manufactured by the McDonnell Douglas Astronautics Company. The first stage design had a single engine with a thrust of 21,528,320 Newtons (4,840,000 pounds). The booster was designed for a lift capacity of 45,360 kilograms (100,000 pounds) to low earth orbit. Chrysler projected the recurring cost of the first stage to be $34.1 million per vehicle (for a 20-vehicle buy, spread over five years). The cost of the stage one tankage was estimated to be $39.70 per kilogram ($18.00 per pound).36

McDonnell Douglas MCD Booster Concept. A January 1968 study by McDonnell Douglas proposed an MCD launch vehicle similar to the Chrysler design. The booster was a two-stage configuration, utilizing a pressure-fed first stage and an S-IVB as the second stage. The launch vehicle was designed to lift 45,360 kilograms (100,000 pounds) to LEO. McDonnell Douglas estimated the per-vehicle cost for the first stage to be $34.8 million, assuming a 20-vehicle, five-year procurement program. The cost of the first stage structure was projected to be $89.56 per kilogram ($40.62 per pound).37

North American Rockwell MCD Booster Concept. North American Rockwell completed a study in October 1968 for an MCD launch vehicle. After going through several design iterations, Rockwell settled on a two-stage baseline design with a 20,400-kilogram (45,000-pound) payload capacity to low earth orbit. Both stages were pressure-fed systems using a single engine. The thrust of the first stage engine was 20.1 million Newtons (4.52 million pounds). The vehicle could be off-loaded for smaller payloads and augmented with strap-ons (extra stage ones) for larger payloads. Rockwell estimated recurring costs for the entire booster to be $28.182 million per vehicle. Cost of the structure was $44.10 per kilogram ($20.00 per pound).35

Figure 9. The Chrysler MCD booster first stage (left) and
the McDonnell Douglas MCD booster first stage
Boeing MCD Booster Concept. Boeing began independent research and development studies on minimum-cost design boosters in July 1968. These efforts resulted in the development of a design called the Cost Optimized Launch Vehicle (COLV). Boeing went through three successive design iterations to arrive at the COLV III configuration. COLV III was a three-stage booster with a 15,420-kilogram (34,000-pound) lift capacity to LEO. The three stages were stacked in a tandem fashion, and a "double bubble" spherical propellant tank arrangement was used for each stage. All three stages would use TRW pressure-fed engines that were to be derived from the simple, inexpensive engines that TRW was testing at that time. The first-stage thrust was 12 million Newtons (2.7 million pounds). Total first-unit cost for the entire vehicle was $14.427 million, resulting in a launch cost per payload kilogram of $936 ($424 per pound).39

Figure 10. The two-stage Rockwell MCD booster concept (left) and
the Boeing double bubble three-stage MCD booster design

Martin Marietta MCD Booster Concept. Between November 1968 and January 1969, Martin Marietta developed a preliminary design for a two-stage pressure-fed MCD launch vehicle with a payload capacity of 20,400 kilograms (45,000 pounds) into a polar low earth orbit. The design employed a single engine for each stage, and the lift-off thrust was 25.9 million Newtons (5.823 million pounds). Martin estimated the recurring cost per complete vehicle to be $9.661 million.40

Complementary Boeing Hardware Activities. During the same period that the COLV concepts were being developed, Boeing was energetically pursuing a number of complementary MCD hardware activities. The company fabricated a complete set of tanks and structure that were sized for the TRW 1.112-million-Newton (250,000-pound)-thrust MCD pressure-fed engine, which was being tested at the Air Force's Rocket Propulsion Laboratory. Using commercial fabrication techniques through the application of ASME boiler code requirements, Boeing also conducted a variety of tank fabrication and testing activities. A complete double bubble spherical tank was constructed.41

The Boeing MCD Booster Study Contract

The Space and Missile Systems Organization planned to release a request for proposal (RFP) in early to mid-1969 for an MCD space booster design and costing study. Prior to the RFP release, SAMSO requested that The Aerospace Corporation develop a new in-house MCD booster baseline design. This new design was to take advantage of the various MCD studies that had been completed or were on-going. The SAMSO/Aerospace configuration resulting from the study was a family of three vehicles with payload lift capabilities of 11,340; 22,680; and 40,823 kilograms (25,000; 50,000; and 90,000 pounds), respectively. The basic stage one of the smallest vehicle served as the core vehicle and as strap-ons for the larger two vehicles (two and four strap-ons, respectively). The second stage of the smallest vehicle served as the third stage for the larger two vehicles.42

On 7 April 1969, SAMSO issued the RFP for the MCD design and cost contract. Titled "Minimum Cost Design Launch Vehicle Design/Costing Study," the contract was for a seven-month level-of-effort study. Four contractors (Boeing Michoud, Martin Marietta, McDonnell Douglas, and North American Rockwell) submitted proposals by the 7 May 1969 due date. SAMSO selected Boeing, and awarded a contract for $1.017 million on 25 July 1969.

Using their double bubble tandem stage approach, Boeing designers initially proposed a family of three MCD boosters; but they later determined that the parallel staging concept used in the Aerospace baseline design resulted in slightly lower costs. (This was due primarily to the commonality of design and the higher production rates that the Aerospace parallel approach allowed.)

Boeing redesigned their family of three-stage vehicles to use parallel staging, resulting in a configuration that was very similar to the Aerospace baseline design. Boeing engineers used the system/subsystem cost optimization technique (SCOT), a Boeing-developed minimum-cost design methodology in their trade studies. The design called for pressure-fed TRW engines for all stages. There was a high degree of design commonality between stages (only the third stage of the smallest vehicle was unique). The three vehicle designs had payload capacities to low earth orbit of 11,340; 22,680; and 45,360 kilograms (25,000; 50,000; and 100,000 pounds), respectively. Boeing estimated that the cost of placing a kilogram into LEO using their MCD vehicle family would range from $924 to $1,437 per kilogram ($420 to $652 per pound) (see table 9). The worst cost case was for polar launches.44

The contract period for the Boeing study ended on 25 February 1970. In an independent assessment, Aerospace calculated the launch costs to be somewhat higher than Boeing's figures. Aerospace projected the cost of placing a kilogram into LEO, using the Boeing launch system design, to be from $1,605 to $2,425 per kilogram ($728 to $1,101 per pound).46

In conjunction with the Boeing effort, SAMSO developed a comprehensive program management plan for the development of a minimum-cost design space launch vehicle. The plan indicated that significant total program cost savings could be realized by tailoring various management disciplines to the MCD design approach.47

TRW MCD Booster Concepts

TRW has proposed a number of MCD booster concepts over the years and has been a strong and consistent advocate of using simple propulsion system and vehicle designs to lower the cost of space transportation. This enthusiasm has stemmed at least partly from their remarkable success with simple and very low-cost pressure-fed rocket engine development and testing in the late 1960s. Through the Space Technology Laboratories/Aerospace studies and the Sea Dragon evaluation effort, TRW also benefited from early exposure to the concept of simple, cost-optimized boosters.

MCD Liquid Strap-on Replacement for the Titan SRMs. In October 1968, TRW proposed a low-cost, liquid strap-on design for use in place of the solid-propellant strap-ons of the Titan IIIC. The system used a single-chamber, throttleable pressure-fed engine, and had a cost of $25.95 per kilogram of structure ($11.76 per pound).45

A Proposal to NASA for a Family of MCD Boosters. NASA contracted with TRW, as a part of the national space booster study, to develop a concept for a low-cost launch vehicle family that would be capable of operating in the 1973 to 1985 period and placing payloads of 18,150 to 45,360 kilograms (40,000 to 100,000 pounds) into LEO. TRW used MCD design principles to propose nine different vehicle configurations that encompassed the specified lift requirements and provided an expanded LEO lift capability up to 113,400 kilograms (250,000 pounds). TRW's largest booster concept had a payload lift capability commensurate with the Saturn V. Several of the concepts depended on the Saturn S-IVB as a second or third stage. Each booster concept exhibited a high degree of commonality with the other vehicle designs.49

The baseline low-cost launch vehicle within the nine-vehicle family was a three-stage expendable booster with a lift capacity of 60,000 kilograms (133,000 pounds) to low earth orbit. TRW estimated the cost of placing a kilogram into LEO using the baseline vehicle to be $452 ($205 per pound), assuming recurring production costs only. After adding nonrecurring costs and launch processing and support costs, the per-kilogram price to LEO increased to $1,235 ($561 per pound). The vehicle used a single first-stage, pressure-fed engine with a thrust of 51.73 million Newtons (11.63 million pounds), although TRW included an alternative configuration for the first stage that used four pressure-fed engines with a thrust of 13.34 million Newtons (3.0 million pounds) each (see Table 9).5O

A Lost Opportunity for MCD Booster Development

There is a significant variance in the launch cost estimates of the many different MCD vehicle concepts developed by the government and the aerospace industry in the late 1960s. However, among these concepts there was universal agreement that major launch cost reductions could be achieved through the application of a minimum-cost design methodology. There is every reason to believe that similar launch cost reduction ratios are available today if we were to apply the same cost-optimized design philosophy.

During the late 1960s, the Air Force's Space and Missile Systems Organization attempted to start a new program to design and build MCD operational boosters. However, there was stiff competition for funds from new aircraft initiatives like the F-15, F-16, and B-1, as well as from the manned orbiting laboratory. These programs received Air Force budgetary priority and, beyond the Boeing study, the MCD booster program received only a small amount of funding to pursue some pressure-fed technology studies.51

In March 1969, a presidential task group was formed to determine the appropriate course the US should take in space after Apollo. The task group proposed a space shuttle as the means for future space access. Preliminary design and technology studies for a shuttle were initiated in 1969, and in March 1970, President Nixon chose to pursue a manned space station and a reusable space shuttle as the next major US goals in space. The space station was soon deferred until an operational shuttle could be fielded.52

The shuttle concept was extolled as the answer to lowering launch costs, and it would use cutting-edge technology to get there. Aerospace managers and engineers were quickly enamored of the concept, as was the public. The idea of using simple unmanned boosters with steel tanks and pressure-fed engines was not technically or operationally exciting to the aerospace community at large, and it did not seem to hold the promise of billions of government dollars for development and for thousands of aerospace jobs. Further, it did not engender within the American people or their political representatives a grand vision of the future (like the Space Shuttle did), and it was far afield of NASA's charter to advance aerospace technology. Consequently, initiatives to develop a minimum-cost launch system were quietly halted.

More Recent Minimum Cost Design Initiatives

In 1980, the Air Force contracted TRW to develop a low-cost booster configuration that would have a payload lift capability equal to the maximum capacity of the Space Shuttle. TRW took the original 1969 study that had been accomplished for NASA, which proposed a family of simple pressure-fed boosters, and updated it to be consistent with 1981 technology and cost.53 The result was an unmanned launch vehicle called the Low Cost Shuttle Surrogate Booster (LCSSB).54

Figure 11. The TRW Low Cost Shuttle Surrogate Booster, a pressure-fed
MCD launch vehicle proposed to the Air Force in 1981

The LCSSB configuration was very similar to the original baseline vehicle in the 1969 NASA study. The booster had three pressure-fed stages, with a first-stage thrust of 30.25 million Newtons (6.8 million pounds). The first stage used four engines, each with a thrust of 7.56 million Newtons (1.7 million pounds). These four engines were identical to the second-stage engine, except that the first-stage engines had a higher chamber pressure and an expansion ratio of 6:1 (for sea-level/low-altitude operations), compared with the second-stage engine expansion ratio of 31:1 (for high-altitude/vacuum operations). Keeping the designs of the first- and second-stage engines essentially the same would have kept development costs down. The booster had a payload capacity to low earth orbit of 29,756 kilograms (65,600 pounds) when launching due east from Cape Canaveral. When launching into a 90-degree polar orbit, the LCSSB had a lift capacity of 23,178 kilograms (51,100 pounds). The system had a launch cost for production vehicles of $59.2 million per launch (including all launch processing and support costs). This equated to a cost of $1,989 per kilogram ($901 per pound) to LEO, assuming an easterly launch (see table 9).55

Under Secretary of the Air Force Pete Aldridge encountered a storm of opposition from NASA and some members of Congress when he sought funding in the mid-1980s (pre-Challenger) for a small buy of Titan complementary expendable launch vehicles to augment the Shuttle fleet. It is therefore not surprising that the concept for the LCSSB, formally proposed one month after the first successful Shuttle flight, ended up going nowhere.

The SEALAR Development Effort

Truax Engineering, Inc. (TEI) has championed the original Aerojet Sea Dragon concept since the late 1960s. TEI developed a phased approach for a family of launch vehicles that led up to Sea Dragon. Starting with a small single-stage sea launch and recovery demonstrator designated the X3, TEI proposed to follow with a booster having a Shuttle-class lift capability. Called Excalibur, it was to be essentially a scaled-down version of Sea Dragon. These developmental precursors would lead ultimately to the fielding of an operational Sea Dragon launch system.56

In 1988, the Naval Research Laboratory's Naval Center for Space Technology (NCST) issued a broad area announcement for the SEALAR (Sea Launch and Recovery) concept, and TEI was the successful bidder.57 NCST called for the SEALAR program to use the design-for-minimum-cost methodology as the booster's guiding design criteria. The Navy wanted a simple, two-stage, launch system that could lift 4,500 kilograms (10,000 pounds) to low earth orbit. TEI proposed a down-sized Excalibur design, appropriately named SubCalibur, which was one-eightieth the size of the original Sea Dragon concept.55

Work moved forward over the next several years. There were a number of static tests of X3 vehicle variants, as well as drop tests from a helicopter into Monterey Bay, California.59 The X3 test articles represented near-scale demonstrators of the SubCalibur's first stage.

Progress on the SEALAR program was so encouraging that in 1990 the Senate Armed Services Committee praised the program, increased the Navy's 1991 SEALAR budget request by 900 percent, and called for a competition between SEALAR and the Air Force's Advanced Launch System (ALS) program. The Committee's report on the FY91 defense budget said that SEALAR could lower launch costs and increase operational responsiveness "for a fraction of the cost of the Air Force's advanced launch system." The report characterized the ALS development program as being "entirely unrealistic."60

Despite a promising start, the SEALAR program as originally envisioned did not come to fruition. An X3 test vehicle suffered a tank failure after repeated pressurization cycles, and the NCST decided to finish fabrication of a flight test demonstrator "in house." The vehicle was close to achieving its first flight when the Navy terminated funding in late 1991.

Another Lost Opportunity for MCD Booster Development

In August 1987 an article that appeared in Newsweek generated a renewed interest in the concept of using minimum-cost design techniques to develop low-cost launch vehicles.61 The Office of Technology Assessment (OTA) conducted a workshop in December 1987 to allow discussion of the concept among aerospace community experts.62 Although the OTA findings were generally favorable toward simple, low-cost booster designs, there was a prevailing perception that the Advanced Launch System program already embraced most of the principles of designing for minimum cost.63 The opportunity to reestablish an initiative for developing simple, cost-optimized space boosters was lost.

Current Low-Cost Booster Development Efforts

There have been several recent proposals to develop new launch systems that are based, at least to some extent, on minimum-cost design principles. There are a number of motivations behind these proposals, including the continued erosion of the US launch industry's market share; the continuing need for lower launch costs (especially among the smallsat community); the failure of government and/or industry to develop a clear path to drastically reducing launch costs; and the intuitive, as well as quantifiable, benefits of simple, low-cost designs.

The McDonnell Douglas Delta Replacement

The McDonnell Douglas Space Systems Company (MDSSC), in a cooperative effort with Allied Signal and TRW Space Systems Group, is defining a family of low-cost launch vehicles. These concepts are an outgrowth of the Advanced Launch System Phase A studies. The near-term goal is to develop a commercial substitute for the Delta launch vehicle that will provide lower launch costs and ensure MDSSC a healthy niche in the future commercial launch market. A variety of growth options, with payload capacities up to 226,800 kilograms (500,000 pounds) to LEO, have been proposed.64

The heart of the MDSSC concept revolves around three key design features. The Delta-class vehicle is configured to use an all-welded monocoque structure, simple TRW-developed engines using ablative cooling and pintle injector technology, and low-pressure, stage-mounted, turbopump assemblies using Allied Signal-developed foil bearing technology.65 MDSSC views the low-pressure turbopumps as an optimal compromise between complex high-pressure turbomachinery with ultra-lightweight tanks, and pressure-fed systems with heavier tanks. The TRW engines use liquid hydrogen and liquid oxygen for propellants, although RP-1 is still being studied for use as a first-stage fuel. An 88,960-Newton (20,000-pound)-thrust engine using TRW's low-cost engine design of the late 1960s has been extensively and successfully tested using LOX and hydrogen at NASA's Lewis Research Center. MDSSC projects 50 to 70 percent cost savings over current Delta prices. The Marshall Space Flight Center reviewed the initial proposals and concluded that the concept has promise but that some major technology questions must still be answered.66

The PacAstro Smallsat Booster

Rick Fleeter and Robert Leppo founded the PacAstro company in 1990 with the express purpose of developing and marketing a small launch vehicle that would be optimized for low cost and would meet the needs of the small satellite community. Since then, PacAstro has established a partnership with TRW for the marketing of launch services and the development and supply of the launch vehicle engines.67 PacAstro plans to keep the launch price minimized by achieving low development costs, low hardware recurring costs, and low launch operations costs. The company believes that the key to minimizing development costs is to use simple, affordable, and off-the-shelf components as much as possible, and the PacAstro booster design reflects this philosophy.68

The PacAstro vehicle is an expendable two-stage pressure-fed booster with a capability for launching a payload of 250 kilograms (550 pounds) into a 750-kilometer (466-mile) altitude polar orbit.69 The vehicle, which uses liquid oxygen and RP-1 as propellants for both stages, has a first-stage thrust of 310,000 Newtons (69,700 pounds).70 PacAstro estimates the total launch cost to be $5 million (in FY93 costs), resulting in a per kilogram cost to orbit of $20,000 ($9,090 per pound) for a 750-kilometer polar orbit.71 Although these costs are higher than those of existing large expendable launch vehicles, they are very competitive with existing small satellite launcher costs.

The Norwegian and Swedish space agencies have tentatively chosen PacAstro to supply boosters for polar launches from the Andoya Rocket Range in northern Norway, starting in 1996. The space agencies plan to launch up to eight small payloads annually using the PacAstro vehicle.72

Figure 12. PacAstro low cost launch vehicle

The Microcosm Ultra-Low-Cost Booster

Microcosm, Inc. is proposing to develop an ultra-low-cost, expendable launch system that is to be optimized for the lowest possible cost. The booster's configuration bears some resemblance to the Boeing MCD launch vehicle design that was developed for the Air Force in the late 1960s. The proposed launch vehicle would have a payload capacity to low earth orbit of 6,232 kilograms (13,740 pounds). The vehicle design clusters six nearly identical strap-ons around a central core, with the payload on top. The core vehicle uses the same design as the six strap-ons, except it includes the payload and payload fairing. The core vehicle and strap-ons all employ pressure-fed propulsion systems that use liquid oxygen and RP-1 for propellants. They also each use multiple engines, and the propellant feed systems are cross-strapped so that all of the launch vehicle's engines can use propellant from only two sets of propellant tanks at a time. All engines are burning in parallel at lift-off.73

Booster steering is accomplished through thrust magnitude control, which varies the thrust levels of appropriate engines during ascent through a network of propellant valves. This steering technique has enabled vehicle designers to avoid complex and costly thrust vector control hardware such as actuators, hydraulic systems, or liquid injection thrust vector control systems.74

Figure 13. Microcosm ultra-low cost launch vehicle concept, which uses
pressure-fed LOX/RP-1 engines and parallel staging

The staging sequence results in a four-stage vehicle. At lift-off, all engines are feeding off the propellant from two opposite strap-ons. These strap-on tanks are separated when empty, constituting the end of stage one. The process is repeated for stages two and three, until only the core vehicle (and payload), containing a full load of propellant, is left to accomplish the fourth-stage burn.75

The booster's multiple identical engines and propellant tank sets not only keep the nonrecurring development costs low, but they also create opportunities for manufacturing economies of scale through high production rates. Microcosm views low development costs as key to achieving a new launch vehicle program start in today's federal and commercial budget climate.76

On 26 April 1993, the Air Force awarded Microcosm a Phase I small business innovative research (SBIR) contract to further refine their booster concept and study its potential application for future DOD launch requirements. Microcosm is hoping to pursue a Phase II SBIR that will lead to the development of demonstration and test hardware.77


There is not today a level of enthusiasm for minimum-cost launch vehicles to match the excitement within the Air Force and the aerospace industry in the late 1960s. However, the continuing burden of high launch costs is forcing government and industry to continue to seek a low-cost launch solution.

The numerous and widely varying concepts currently being proposed to achieve lower space transportation costs can be broadly allocated to two groups of supporters. One group seeks to reduce launch costs through one or more technological leaps (the futurists). This approach is characterized by generally high-risk and expensive development programs accompanied by the promise that operational costs will be so low that the development program is justified. The other group seeks to lower costs by doing what we currently do better and more efficiently (the pragmatists). This approach is characterized by more modest technological requirements and lower-risk development programs.

Designing a very simple launch vehicle with achievement of the lowest possible life cycle cost being the dominant consideration is clearly in the latter category. It represents the design philosophy that is most different from the technological leap approach. The concept of designing a launch vehicle for minimum cost has been studied by government agencies and the aerospace industry many times over the years, and the results have consistently indicated that huge reductions in launch costs are available using this technique.

Unfortunately, the arrival of the Space Shuttle concept, which was seen in the late 1960s as the answer for reducing high launch costs, combined with aerospace industry concerns about the loss of launch vehicle production profits and combined with a general inclination and desire for high technology solutions, has prevented the MCD approach from moving off the paper stage to flight hardware. It is time to seriously explore the application of minimum-cost design techniques for developing a new low-cost launch system--a system that could facilitate a broad expansion of space exploitation activities.



1. Naval Center for Space Technology, Sea Launch and Recovery (SEALAR) System Concept to Launch Brilliant Pebbles (Washington, D.C.: Naval Research Laboratory, January 1992), 17.

2. R. C. Truax and J. D. Ryan, "Sea Launch of Rocket Vehicles" (SAE 433A, presented at the 1961 National Aeronautic and Space Engineering and Manufacturing Meeting, Los Angeles, Calif., 13 October 1961), 6-7

3. Robert C. Truax, "Sea Dragon in the Manned Mars Mission," The Journal of Practical Applications in Space, Fall 1990, 8.

4. R. A. Raffety, "Sea Launch Flight Test Program of a Liquid-Propellant Rocket," Aerojet-General Corporation, 20 November 1961, supplement to SAE paper 433A, presented 13 October 1961, 1.

5. Ibid.,4-6.

6. R. C. Truax, "Cheap Transportation for Cheap Satellites" (Paper presented at the MAA/DARPA Meeting on Lightweight Satellite Systems, Monterey, Calif., 4-6 August 1987),2.

7. William H. Ganoe, "Rockets from the Sea," Ad Astra, July/August 1990, 71.

8. Robert C. Truax, "Thousand Tons to Orbit," Astronautics, January 1963, 45.

9. H. G. Campbell, A Cost Analysis of Large Booster Systems for Planetary Exploration (Santa Monica, Calif.: RAND Corporation, August 1963), 6.

10. Truax Engineering, Inc., Sea Dragon Launch Vehicle data sheet, Saratoga, Calif., no date.

11. Truax, "Thousand Tons to Orbit," 4546.

12. Ibid., 45.

13. Truax, "Cheap Transportation for Cheap Satellites," 2, 4.

14. Study of Large Sea-Launch Space Vehicle, vol.3, summary report, contract no. NAS8-2599 (Redondo Beach, Calif. Space Technology Laboratories, Inc./Aerojet General Corporation, January 1963), 3-15.

15. Truax, "Sea Dragon in the Manned Mars Mission," 2.

16. "Project Private Enterprise-A Commercial Space Transport Program," Truax Engineering, Inc., Saratoga, Calif,, 1984, 7; Study of Large Sea-Launch Space Vehicle, 1-3.

17. Truax, "Sea Dragon in the Manned Mars Mission," 8; Study of Large Sea-Launch Space Vehicle, 1-2, 1-3, 1-4.

18. Ganoe.

19. A. Schnitt and Colonel F. W. Kniss, "Proposed Minimum Cost Space Launch Vehicle System," TOR-0158(3415-15)-1 (El Segundo, Calif.: 1 July 1968), 1-1.

20. Walter Tydon, Minimum Cost Design Launch Vehicle Design I Costing Study, vol.2, Background Studies, TOR-0059(6526-01)-2 (El Segundo, Calif.: The Aerospace Corporation, 31 July 1970), 1-2.

21. John R. London III, "The Ultimate Standoff Weapon," Airpower Journal, Summer 1993, 67.

22. Tydon

23. Ibid., 4-6

24. Schnitt and Kniss, 3-17

25. Tydon.

26. Gregg Easterbrook, "Big Dumb Rockets," Newsweek, 17 August 1987, 46.

27. Tydon, 1-3.

28. Ibid., 4-7.

29. Ibid., 4-8, 4-9.

30. Ibid., 4-10.

31. Ibid., 4-11.

32. Schnitt and Kniss, 1-1, 1-2.

33. Tydon, 1-2, 1-3.

34. R. M. AIlman, "Minimum-Cost-Desigu Space Launch Vehicle," briefing to The Aerospace Corporation Board of Trustees ad hoc committee on space Systems costs, El Segundo, Calif., 25 September 1987.

35. Tydon, 4-1, 4-13.

36. Ibid., 4-13, 4-17,

37. Ibid., 4-13, 4-18.

38. Ibid., 4-13, 4-19, 4-20, 4-21.

39. Ibid., 4-22.

40. Ibid., 4-24, 4-25.

41. Ibid., 4-23.

42. Ibid., 5-1.

43. Walter Tydon, Minimum Cost Design Launch Vehicle Design/Costing Study, vol.1, Summary, TOR-0059(6526-01)-2 (El Segundo, Calif.: The Aerospace Corporation, 31 July 1970), 11.

44. Ibid., 11-15.

45. Walter Tydon, Minimum Cost Design Launch Vehicle Design/Costing Study, vol.3, Critique of Boeing MCD Study, TOR-0059(6526-01)-2 (El Segundo, Calif.: The Aerospace Corporation, 31 July 1970), 3-2.

46. Tydon, vol.1, 17.

47. David J. Teal, "Minimum Cost Design Space Launch Vehicle Management Plan," SAMSO-TR-70-185, Space and Missile Systems Organization, Los Angeles AFS, Calif., 1970,

48. Tydon, vol.2, 4-13, 4-14.

49. TRW Systems Group, "Low Cost Launch Vehicle Study," final briefing, NASA contract no. NASw-1792, Redondo Beach, Calif., 23 June 1969, 1.6, 1.9.

50. Ibid.

51. AIlman.

52. Walter A. McDougall, The Heavens and the Earth (New York: Basic Books, Inc., 1985), 421.

53. D. E. Fritz and R. L. Sackheim, "Study of a Cost Optimized Pressure Fed Liquid Rocket Launch Vehicle" (Paper presented at the AIAA/SAE/ASME 18th Joint Propulsion Conference, Cleveland, Ohio, 21-23 June 1982), 1.

54. TRW, Inc., "Low Cost Shuttle Surrogate Booster (LCSSB)," final report (Redondo Beach, Calif., 15 May 1981), 1.

55. Ibid., 35, 46, 47, 109.

56. "Project Private Enterprise-A Commercial Space Transport Program," 10.

57. Ganoe.

58. Proposal for SEALAR Program, Truax Engineering, Inc., Saratoga, Calif., 1988, A-2.

59. "Test Successful on Scale Model SEALAR Vehicle," Sea Technology, July 1990, 66.

60. "Panel Seeks Competition of Navy, USAF Launchers," Aviation Week & Space Technology, 30 July 1990, 28.

61. Easterbrook, 46-60.

62. Office of Technology Assessment, Big Dumb Boosters - A Low Cost Space Transportation Option? (Washington, D.C.: Government Printing Office, February 1989), 1.

63. AIlman.

64. J. P. Henneberry et al., "Low-Cost Expendable Launch Vehicles" (Paper presented at the AIAA/SAE/ASME/ASEE 28th Joint Propulsion Conference and Exhibit, Nashville, Tenn., 6-8 July 1992), 1, 14.

65. Ibid., 2, 4, 14.

66. Marshall Space Flight Center, Heavy Lift Launch Vehicle Definition Office, An Assessment of the McDonnell Douglas Space Systems Company Low Cost Vehicle Family, NASA Report, Huntsville, Ma., 27 July 1992.

67. PacAstro Summary data sheet, PacAstro, Herndon, Virginia, 1992.

68. Rick Fleeter to John London, letter, 28 September 1992.

69. Ibid.

70. PacAstro, "A Low Cost Expendable Launch Vehicle for 500-Pound Class Satellites," submitted to the Strategic Defense Initiative Organization (Herndon, Va., 26 May 1992).

71. Fleeter to London, letter.

72. Peter B. de Selding, "U.S. Firm Enlisted for Small Satellite Launches," Space News, 22-28 February 1993, 8.

73. Edward L. Keith, "System Analysis and Description of an Ultra-Low Cost Ground to Low Earth Orbit Cargo Delivery System" (Paper presented at the World Space Congress, Washington, D.C., 31 August 1992), 4, 5, 9.

74. Ibid., 7.

75. Ibid., 5.

76. Ibid.

77. Bob Conger, Microcosm, Inc., telephone conversation with author, 9 June 1993.