SEARCH
Advanced Search
ARCHIVE

There are currently 222 documents in the archive.

Bibliography Archives List Library Listing

KEY DOCUMENTS

Here are some key documents from the archive to get you started:


More Documents Archives
WHAT'S NEW
29 July 2012
Added "Space Debris and Its Mitigation" to the archive.
16 July 2012
Space Future has been on something of a hiatus of late. With the concept of Space Tourism steadily increasing in acceptance, and the advances of commercial space, much of our purpose could be said to be achieved. But this industry is still nascent, and there's much to do. So...watch this space.
9 December 2010
Updated "What the Growth of a Space Tourism Industry Could Contribute to Employment, Economic Growth, Environmental Protection, Education, Culture and World Peace" to the 2009 revision.
7 December 2008
"What the Growth of a Space Tourism Industry Could Contribute to Employment, Economic Growth, Environmental Protection, Education, Culture and World Peace" is now the top entry on Space Future's Key Documents list.
30 November 2008
Added Lynx to the Vehicle Designs page.
More What's New Subscribe Updates by Email
SFJ Top Stories
More Articles
J P Penn & C A Lindley, 1997, "RLV Design Optimization for Human Presence in Space", The Aerospace Corporation.
Also downloadable from http://www.spacefuture.com/archive/rlv design optimization for human presence in space.shtml
References and Referring Papers    Printable Version 
 Bibliographic Index
The Aerospace Corporation
RLV Design Optimization for Human Presence in Space
2350 E. El Segundo Bl. El Segundo, California 90245-4691 (310) 336-0559; Fax (310) 336-0503; E-Mail penn@courier1.aero.org
Jay P. Penn and Dr. Charles A. Lindley*

The design of the Lockheed Martin RLV is currently optimized for traditional spacelift missions. For these missions the RLV design objectives such as 7-day vehicle turntimes, $1,000/lb of payload, and 99.6% reliabilities are acceptable. This design may also enable a very limited human presence in space with a correspondingly high price for a ticket.

In this paper we explore a different design approach. We attempt to optimize the design of the RLV so that it can grow to economically and reliably (safely) accommodate a much larger market for human presence in space, space tourism. Market surveys suggest that space tourism, as a full human presence market, will dictate flight rates two orders of magnitude larger than traditional spacelift. The system must also accommodate traditional spacelift for early market and revenue capture.

RLV design optimization for space tourism implies vastly improved levels of robustness (manifested primarily as design margin), improved access for maintenance, increased levels of onboard autonomy, and design for aircraft-like turnaround. Many of these operability and reliability improvements will come at the expense of payload delivered to orbit. To exacerbate the challenge, the tourism market requires a higher inclination, higher energy orbit than that used for typical satellite deployments. On a pure SSTO vehicle, this can result in a payload reduction from 45 klb to 35 klb.

In addition, with near-term technology, pure SSTOs optimized for operations result in increased fuel costs per passenger compared to an RLV optimized for traditional spacelift.

This paper will explore alternate RLV options in an attempt to yield a highly operable and robust RLV design that reduces overall operational costs. The reference RLV will be the current Lockheed Martin RLV with improved levels of robustness, and incorporating rapid turnaround technologies. An RLV similar (possibly identical) in size to the reference RLV but carrying a small fully reusable upper stage appears to provide a viable alternative to the baseline and may offer a lower operating and fuel cost per passenger.

This approach, however, leads to operational issues in the rapid/efficient recovery and relaunch of the RLV, which must now land downrange from the launch site. Means to recover and rapidly relaunch the RLV will also be explored. This involves conducting trades between multiple launch sites, self-ferry capability, and/or alternating between posigrade and retrograde orbits for space hotels.

TABLE OF CONTENTS
  1. Introduction
  2. Traditional vs. Human Presence (Tourism) Spacelift Requirements
  3. Supply vs. Demand Considerations
  4. Need for Robust Designs
  5. Lockheed Martin RLV-Based Human Presence Spacelift Options
  6. Evaluation Methodology
  7. Results and Observations
  8. Conclusions
1. INTRODUCTION

This paper addresses the development of spaceplanes that are based on the Lockheed Martin RLV (VentureStar) but optimized for the ultra-high flight rate and high reliability demands of the Human Presence mission rather than for traditional spacelift missions. Whereas only a minimal human presence in space is required to support traditional spacelift missions, studies indicate that the space tourism market will dominate the need for human presence in space. A prior paper by these authors addressed the fundamental operability, reliability, and cost drivers needed to satisfy this tourism mission [1].

This paper explores at least eight VentureStar-based configurations that are optimized for these drivers. It also addresses the ability of the tourism-optimized VentureStar-based concept to address initial market penetration in the traditional spacelift market. Figures of merit based on Aerospace RLV systems engineering were derived to evaluate each of the considered RLV based concepts against the needs and economic drivers for a viable space tourism vehicle. The results of this assessment are discussed in this paper.

2. TRADITIONAL VS. TOURISM SPACELIFT REQUIREMENTS

Traditional spacelift missions will require an infrastructure made up of launch and landing sites, vehicles, and facilities able to support up to approximately 50 flights per year. For these traditional missions, recurring launch costs of $1,000 to $2,000/lb ($2,205 to $4,410/kg) will be quite acceptable, and will present a significant improvement over today's systems. Projected VentureStar Reusable Launch Vehicle ( RLV) reliabilities of 99.6% will also provide significant improvement compared to today's systems at 95-98%.

The conventional wisdom is that once low-cost ($1,000 to $2,000/lb ($2,205 to $4,410/kg)) access to space is available, market demand will grow, chiefly by larger numbers of satellites. But the commitment and growth of such programs is a very slow process. The price is still too high to attract major traffic from such uses as low-cost passenger transport to and from space, space manufacturing, rapid, worldwide package delivery via suborbital flight, and possibly extensive levels of mass delivered via military Global Reach/Global Power missions. And the traditional approach may not provide an economically acceptable stepping-stone to other human presence missions such as manned Mars exploration, a permanent manned Moon base, or significant solar system exploration.

The driving force of a viable space tourism industry will of course be economic-the potential to create a profitable business. Market surveys [3,4] and recent analysis by the authors of this paper [1] suggest that a viable space tourism industry will require flight rates about two orders of magnitude higher than those required for conventional spacelift. Although enabling round-trip cost goals for a viable space tourism business are about $240 per pound ($529/kg), or $72,000 per passenger round-trip, goals should be about $50 per pound ($110/kg), or approximately $15,000 for a typical passenger and baggage. The lower price will probably open space tourism to the general population. Vehicle reliabilities must approach those of commercial aircraft as closely as possible. The following discusses the key requirements for a space tourism business.

Cost per Flight

The first step in determining the economic potential of space tourism was to understand the costs that the market will bear. To accomplish this, in our prior work on this topic, the authors referenced previously conducted market analysis [3]. Since no space tourism market exists, the results were considered speculative. In the survey, U.S. citizens were asked whether they had a desire to go on a space tourism adventure including perhaps a 5-day stay in an orbital space hotel. If so, the survey inquired what percentage of their annual salary they would be willing to spend on such a trip.

The survey results seemed somewhat optimistic in that they implied that everyone who intended to purchase a ticket would, in fact, do so. We more conservatively assumed that only 25% would actually do it sometime over a 20-year initial operating period. This provided demand vs. price data shown in prior work [1].

In view of these demand estimates, it appears that we must bring the launch cost down to the vicinity of $1 to $2 million per flight of a 100-passenger spaceplane. This yields a cost goal of less than $20,000 per ticket. At prices much higher than $72,000 per ticket, or $7 million/ flight, the total cash flow may not be large enough to justify the system. The corresponding projected flight rates varied between 1,500 and 9,500 flights per year.

3. SUPPLY AND DEMAND CONSIDERATIONS

The revenue stream produced from these ticket prices must be sufficient to provide for 1) operating expenses: propellants, maintenance, spares, flight crews, and insurance, 2) vehicle amortization costs, and 3) Research Development, Test and Engineering (RDT&E) costs of the vehicle fleet, facilities, and supporting infrastructure. In addition, the investment must yield a return (ROI) to the investors that fully compensates for risk and payback time.

If a spaceplane system designed for the low costs and high reliability required of space tourism becomes available, it will dominate all other space boost traffic. This is not very important to the spaceplane traffic demand; adding the present traffic of 50 to 100 launches per year would have a small impact on the traffic levels in Table 1. However, it would bring about a major launch cost savings to present space operations, and increase the ROI in space tourism.

Perhaps as important, low-cost launch offered as a result of a robust tourism industry would enable realistic consideration of many space programs presently deterred by boost cost. Programs that come to mind are: manned Mars exploration, a permanent manned Moon base, space defense, space manufacturing, space power generation, and military Space Power/Space Force missions. Many space concepts that have been discussed in the past can hope to become reality only if drastic reductions in boost cost occur.

Reliability

To have a successful space tourism industry, we must have superior passenger safety compared to past booster experience. We can accept some subsystem failures that delay a flight, or that

lead to a mission abort and a safe landing. But failures that result in fatalities and vehicle losses must be rare. It is hard to put a number on this. We cannot hope to reach the reliability of commercial aircraft, where the risk is about 1 in 2,000,000 flights, or 0.9999995 reliability against catastrophic failure. But we need better safety than the often-quoted RLV reliability of 0.999. We believe that with the proposed design philosophy, we should be able to reach a reliability of ³ 0.9999 against catastrophic failures, and that this may be sufficient to support a space tourism industry.

Table 1. Demand vs. Price
U.S. Only U.S. Only World World World
Ticket Price
$		Passengers/
Year		% of U.S.
Population/
Year		Passengers/
Year (a)		Flights/
Year (a)		Revenue/
Year
$Billion (a)
$72,000 7,500 0.015150,000 1,500 $10.8
$24,000 137,500 0.06 550,000 5,500 $13.2
$12,000 237,500 0.10 950,000 9,500 $11.4
$6,000 (b)600,000 0.24 2,400,00024,000$14.4
$2,000 (b)1,250,0000.50 5,000,00050,000$10.0
a
Excludes hotel staff (paid for out of hotel revenues).
b
Not considered realistic for first-generation spaceplane; high flight rate, low price, possible decreasing revenue stream.
Vehicle Turnaround Time

Previous analysis indicates that to support the flight rates mentioned earlier with affordable fleet sizes and infrastructures, vehicle turnaround times of between 12 and 24 hr are necessary. The recent NASA and Air Force reusable launch vehicle efforts have resulted in significant progress in proving operationally efficient technologies. As an example, whereas the Shuttle operations require 20,000 to 30,000 people to support 8 launches a year, the DC-XA operations technology effort has demonstrated a 26-hour turnaround with a crew of approximately 30 people. It is anticipated that the X-33 can demonstrate similar levels of operability and with a significantly larger vehicle.

4. NEED FOR ROBUST DESIGNS

As flight rates increase, the amortized cost per flight of development and vehicle production rapidly decreases. For a reusable system, recurring operational costs will then dominate the economics. The success of space tourism depends on the ability to develop a launch system with an operational efficiency about 200 times greater than the current Space Shuttle. To achieve this dramatic increase in operational efficiency will require a completely new approach to vehicle design.

Historically, U.S. launch vehicles have been designed to optimize the vehicle's flight performance. Sophisticated analytical tools have been developed to eliminate every ounce of excess weight and gain every possible second of specific impulse. Unfortunately, this approach leads to severe compromises in both operability and reliability. As clearly demonstrated by the Space Shuttle Main Engine ( SSME) test and operations program [5], efforts to increase thrust levels to 109% power decreased engine reliability and life by an order of magnitude compared to operating at 100% power (Figure 1).

Figure 1. SSME Historical Reliability

The Soviet rocket program has some interesting counterpoints [6]. The RD-170 rocket engine, burning LOX/kerosene, was developed for expendable applications, not reusability. But there was a philosophy in the USSR that the best proof of an engine's reliability is its ability to run for much longer periods than the design flight time. At the end of its development cycle, the RD-170 was given a long-duration test run. It was finally shut down after 14,400 seconds of continuous running, equivalent to 30 to 60 trips to Earth orbit. It was shut down not because of an engine problem, but because it had exhausted all the propellants available on the test site. At a thrust of 1,777,000 lb (806,047.2 kg) and Isp = 337 sec, it would have used up over 76 million lb (34.7 million kg) of propellants on this single run. And this engine was not even designed for reusable service.

This particular engine operates at a chamber pressure of 3,500 psi. Even so, the highest metal temperatures in the engine when running is held to only 500o C. At such temperatures, there is no thermal creep or other thermal life-limiting effect. Similarly, limiting the turbopump inlet temperatures to what the blades can endure without experiencing thermal creep significantly increases the life of these components, with only a small decrease in overall specific impulse for the turbopump diluent flow.

While these changes alone can substantially improve engine life, there are still numerous failure modes that must be addressed. This can be accomplished using a probabilistic design technology approach similar to that being developed by Rockwell and Rocketdyne for RLVs and aircraft programs, and illustrated in Table 2. This approach interrelates process, material, environmental variability, degree of testing, and test results. Using this probabilistic analytical approach, areas with higher design uncertainties have higher design margins than those with lower uncertainties. To date, this methodology has been used for structural and engine design optimization. However, it should be expanded to encompass the overall design of the vehicle and infrastructure.

Table 2. Detailed Probabilistic Analysis

Hardware/Process High Cycle
Fatigue 		Low Cycle
Fatigue 		Fract.
Mech. 		Wear Displacement Buckling/
Instability
Turbomachinery
Turbine blades × ×
Turbine disk ×××
Hydrostatic bearings ××
Inducer blades × ×
Impeller vanes × ×
Combustion devices
Injector LOX post ×× ×
Injector close-out welds ××
Interpropellant plate ×××
Fuel inlet manifold ×××
Fuel inlet splitter ××
Manifold to liner ××
Close-out ×
MCC liner × ×

While propulsion and Thermal Protection Systems (TPS) have been major sources of operational problems on the Space Shuttle Program, none of the processing time lines associated with any Shuttle system would be acceptable for space tourism. The time required just to open the payload bay doors would exceed the entire time allocated to turn around a vehicle designed for space tourism. To achieve this level of operational efficiency, all systems must be redesigned in a similar manner to the propulsion systems.

Although creative design solutions can resolve some problems, in most cases increased design margins will be a critical element in providing the engineer with the performance relief necessary to improve component life and failure tolerance, and thus improve operability and reliability. It becomes imperative that performance be realistically traded against reduction of operating cost.

Results of analysis conducted at The Aerospace Corporation as well as similar results of NASA KSC studies indicate that if a robust, highly operable RLV is to be attained, the vehicle must use structural and design margins of 50% rather than the 25% generally used in the X-33. Since the majority of the structure is not stress limited, the weight penalty of such margins is far less than the 20% difference in load between 1.25 and 1.5.

In this paper, we defined two levels of robustness for the RLV designs that were studied. The Reference RLV case is identified as a 25% design margin case. This design was assumed to employ approximately the same levels of design margin as does the X-33. It assumed a 25% margin on structure and that subsystems and propulsion systems are not derated from the maximum performance that can be obtained from these subsystems.

But the tourism-optimized RLV cases all were defined to include robust 50% design margin. In addition to the higher structural design margins employed, a 10% derating of assumed engine thrust/weight was used. Furthermore, a 20% increase in specific mass was allocated for all mechanical subsystems. These margins should give the designer the flexibility to design more robust designs, eliminate critical failure modes, allow for both low and high cycle fatigue, and provide for necessary access and handling provisions.

While these allowances probably represent a good estimate of the overall impact of optimizing a vehicle for operability/reliability, it is not adequate merely to apply this design factor to all aspects of the vehicle design. For certain design aspects, a much higher margin is required to achieve the necessary levels of reliability, particularly areas experiencing high variability in the environment or reliability of manufacture, and areas where our ability to analyze the design are limited. Other areas will require very little margin. Areas involving critical failure modes will require more generous margins that those areas with more benign failure consequences.

Ground Systems

Ground systems and supporting facilities must also be optimized for operational efficiency. While automation is often viewed as the solution to this problem, a much more challenging engineering approach is needed. Automation can reduce processing time lines, but at a cost of higher complexity and increased support system maintenance. Automating the operation of a large fluid valve requires the addition of electrical power, a pneumatic actuator, position indicators, and computer support. The more difficult task of simplification should be employed before automation is applied. Developing systems that do not require purge gases, temperature control, precise loading accuracy, clean room conditions, or computer controls can be accomplished, but it poses more of a challenge than the current practice of operation under near-laboratory conditions.

The Space Shuttle LOX propellant loading system, for instance, consists of hundreds of valves, pneumatic panels, cryogenic pumps, vaporizers, and over 500,000 lines of computer code to safely control their operation. A self-pressurizing system could be designed with no pumps or purges, and one valve to control propellant loading. This type of system would require a major development effort [5]. A comprehensive redesign of all ground systems will probably cost in excess of a billion dollars, but it could save many times that amount at very high launch rates.

Assuming a comprehensive effort to optimize both vehicle and ground systems for operational efficiency, we believe that levels of operability and reliability can be achieved that are more closely related to that of aircraft than today's rockets.

5. VENTURESTAR-BASED SPACELIFT OPTIONS

Three major variants of the VentureStar RLV configuration were evaluated, each with two to four subconfiguration options. These will be briefly described below.

Configuration 1 is defined as the baseline VentureStar RLV and is operated as an SSTO vehicle. There are four major subconfigurations. Configuration 1a is the currently planned RLV, which incorporates similar design margins to those of the X-33 vehicle. This includes a 25% structural margin, 10% proof pressure margin on tanks, and generally the maximum technologically achievable levels of engine thrust to weight and subsystem performance. It is assumed that this configuration was designed for the operability and reliability requirements of traditional spacelift markets rather than those required to support a significant human presence in space. This configuration is defined as the reference case and is the only RLV option studied that did not include the 50% margins considered to be robust.

Configuration 1b fixed the GLOW (Gross Lift Off Weight) to match that of Configuration 1a, but increased design margins to 50% to support the tourism requirements. This resulted in a loss of payload to the reference 220 nm by 35 degree inclined orbit, defined as near optimal for a space resort. There was a 7.7% increase in dry weight at the improved margins, reducing the payload from 35,533 lb to only 19,633 lb for the tourism (operability and safety) optimized design. This reduces the estimated number of paying customers per flight from 101 to 56. The reduction is partially offset by the more frequent flight rate allowed by the more operable vehicle, but certain fixed costs per flight like fuel cost per passenger actually increase in the more robust design. This configuration also would reduce the payload to a due east 100 nm circ orbit from approximately 50,000 lb to 34,100 lb.

Case 1c matched the payload to the 35,533 lb (101 passengers) of Case 1a, but the GLOW was allowed to increase to accommodate the higher margins. The GLOW increased only from 2.19 Mlb to 2.69 Mlb because the mass fraction improves with vehicle scale.

Case 1d was intended to determine whether a larger GLOW would increase revenue per flight by increasing payload while at the same time maintaining high design margins. Since SSTOs scale favorably because of improving PMF (Propellant Mass Fraction) with size, a robust large payload SSTO RLV might be attractive. The payload was set at 100,000 lb (286 passengers) with the robust design margins. Aerospace's sizing algorithms yielded a GLOW of 4.18 Mlb compared to a GLOW of 2.19 Mlb for Case 1b sized for less than one-fifth the payload.

Configuration 2 employs the RLV with a pop-up second stage and includes the robust design margins required to support the tourism mission. The reusable upper stage is powered by LOX/JP engines and carries the passengers from the deployment velocity of approximately 12,000 fps to the space resort's orbit and returns passengers to Earth. After the upper stage separates from the RLV, the RLV reenters the atmosphere and either glides to a downrange base or flies back to the launch base using installed turbofan engines and JP fuel. The landing choice is set by the mode of operation and the need to quickly process the system for the next launch. The pop-up mode of operation significantly increases useful payload but at the cost of a somewhat more complex system.

Configuration 2a lands unpowered 500 to 1000 nm downrange of the launch site. At that point a jet engine and JP fuel tank kit is mated to the RLV, and it is flown horizontally back to the launch site. With proper design for operations, this recovery operation would not preclude a one-day turnaround time. Setting the RLV GLOW to match that of the reference VentureStar configuration results in a payload of 60,000 lb or a 171-passenger vehicle.

Configuration 2b is recovered in a similar fashion to that of 2a except that to reduce turnaround delays, jet engines and the JP fuel supply are integrated into the vehicle. This allows the RLV to fly directly back to the launch site without landing downrange for engine kit installation and fueling. The additional weight of the jet engines and fuel must be accelerated to about 12,000 fps, and the RLV booster's GLOW is increased from 2.31 Mlb to 2.58 Mlb. Configuration 2b achieves the same 60,000 lb payload to the resort's orbit as in Configuration 2a.

The use of rocket propellant to fly the vehicle home was also considered, but the propellant requirement became excessive.

Configuration 2c recovers the RLV downrange as with Configuration 2a, except that to eliminate the need to return the vehicle to the launch site, the RLV is mated with a recovered upper stage and relaunched into a retrograde orbit. This returns the RLV to the launch site without the need for ferrying, but results in a substantial loss of payload. Whereas the posigrade payload is 60,000 lb, the retrograde launch working against the Earth's rotation has a payload of only 26,592 lb, or about half.

This approach requires two launch and landing sites equipped for landing, processing, and launching of the RLV system. This may not be a significant disadvantage for a space tourism industry since the traffic volume alone will most likely require more than one operating site. This retrograde-posigrade approach will also require a retrograde and a posigrade resort. This also may not be too detrimental since the very large traffic demand associated with more than one orbital resort may be required anyway.

Another variation of the pop-up mode of operations was considered and quickly discounted. In this alternative approach, the RLV and the upper stage separated at much higher velocities, and the RLV would fly forward to a downrange launch site. The downrange site would be used for another posigrade launch. It was assumed that with a sufficient staging velocity, the RLV booster could circle the globe in only two to three launches. This would also allow the RLV to serve as a rapid mail delivery service.

Unfortunately the analysis determined that with the relatively low lift-to-drag design of the VentureStar, even with a staging velocity very close to orbit, the range was very limited. With a staging velocity only 3,000 fps less than orbital, the RLV achieved a range of less than 4,000 nm. Many sites would thus be required to circle the globe, which was economically unreasonable.

Configuration 3 utilizes the VentureStar RLV as the orbiter and augments it with a simple LOX/JP booster powered by two RD-180 thrust derated engines, or the equivalent. Depending on whether the smaller or larger version of the booster stage is used, the staging velocity varies between approximately Mach 2 and Mach 3. After staging, the booster autonomously glides back to the launch base. In a matter of minutes after the landing, and while the RLV is still in orbit, the booster begins processing for its next launch. As a result, fewer booster stages will be required in the fleet than RLVs. In the proposed configuration, either stage has engine-out capability with adequate thrust for a successful abort. LOX/JP was selected as the propellant because the density and cost of the propellants makes it a good choice for a first stage, compared to an all LOX/ LH2 system.

Configuration 3a sizes the booster stage to achieve 60,000 lb of payload (171 passengers). Configuration 3b is similar to 3a, except that it is sized for a 100,000 lb payload and has a slightly higher staging velocity.

Each of the above configurations was sized to either match the reference VentureStar RLV in either (Case 1a) vehicle size (dimensions and GLOW of the RLV) or to match its payload by modestly scaling up the RLV. Case 1d was the exception, in which the RLV was significantly increased in size to enable both a robust SSTO and a very large 100,000 lb payload. This configuration may be too large in size for a practical first-generation RLV vehicle.

In sizing these vehicles, it was realized that for the Aerospace code results to match the Lockheed Martin hardware fraction numbers required assumption of a design margin of 19% rather than 25%, as used in the X-33, and 44% rather than 50%. Thus, the Aerospace code predicts a slightly heavier vehicle. Since the X-33 weight has increased substantially in design, there is reason to believe that the Aerospace numbers may be reasonable.

6. SYSTEMS EVALUATION METHODOLOGY

The Aerospace Corporation has developed a comprehensive system engineering methodology for evaluating reusable launch vehicle systems and technologies against specific missions and related supply and demand characteristics. This methodology directly relates the operability, reliability, and performance characteristics of the system to the ecomomics that can be obtained.

Although this Aerospace methodology was available to evaluate the concepts on an economic basis, it was not possible within the scope of this effort to utilize the full suite of tools for this limited analysis task. The most promising concepts identified from this first-cut analysis should be explored in a more comprehensive study using this or an equivalent comprehensive tool set.

A simplified methodology was developed to support this IR&D but was based on key drivers used in the more rigorous analysis. This simplified methodology, although not directly traceable, does utilize simplified versions of the key discriminators that would be used in a more comprehensive effort. It was realized that all of

the key discriminators relating performance, operability, and reliability can be ultimately normalized to cost.

Key MOEs

To support this effort, ten key Measurements of Effectiveness (MOEs) were selected that closely relate to the critical drivers of the economic viability of a space tourism industry. These MOEs are identified in Table 3 and are discussed below. Seven of the ten MOEs were directly quantifiable, based on the sizing results shown in Table 4. The raw values of these MOEs are shown in Table 3.

Three of the MOEs could not be directly quantified within the resources of this study. These MOEs were: Vehicle Maintenance, Facility Re-quirements, and Overall System Complexity. The values shown in the results of Table 3 for these three discriminators reflect engineering judgment of Aerospace staff based on prior studies and insight on key factors that are used in the models. The processed and weighted scores of both the seven quantitative and three qualitative MOEs are discussed in the Results section and shown in Table 3. A more comprehensive follow-on effort should quantify these discriminators in a more rigorous way.

Robustness is a key driver of maintenance, of the degree of development testing required, and also a major discriminator of launch reliability, which ultimately drives failure costs. Reliability also affects consumer confidence in the system, which affects overall market capture. Analysis conducted at Aerospace and other research centers have shown that an increase in design margin from 25% to 50% can result in profound improvements in robustness-related cost drivers. Separating the predicted stress and predicted failure stress by a combination of design margin, testing, and process controls can provide orders of magnitude less failure risk, and also minimize wearout of the system due to fatigue. This is the chief reason why commercial aircraft accept the performance penalty of a margin of safety of at least 50%.

Scoring: Concepts sized to include the following margins received a full score for robustness = 10.

  • 50% structural margin
  • 10% reduction in engine thrust/weight
  • 20% increase in subsystem weight
  • 15% weight growth allowance

Concepts sized to provide performance optimized designs with minimal levels of robustness received a score for robustness = 0

  • 25% structural margin
  • No derating of thrust/weight
  • No derating of subsystem performance
  • 15% weight growth allowance

Note that the 15% weight growth allowance is the same in both cases. This allowance is intended to cover the inevitable weight increase during the design and development process, so it is outside of our definition of robustness.

Growth Sensitivity is defined as the percent payload reduction per percent increase in dry weight of the system. Pure SSTO vehicles have the highest sensitivity. Since all of the dry weight is delivered to orbit, a one-pound growth in dry weight results in a one-pound loss of payload. Systems with low growth sensitivity have significantly lower payload penalty as dry weights increase or the system runs into design challenges. Table 4 shows that if the VentureStar baseline's (Case 1a) dry weight grows by 10%, then the payload is reduced by 5.19 ( 10 = 51.9%. The payload would be reduced from 35,533 lb (101 passengers) to 17,040 lb (49 passengers) with a resulting loss of revenue. Although not directly comparable, the X-33's weight has increased by approximately double the 10% figure used in this example. This point also illustrates the need for adequate weight allowance for design uncertainty.

By comparison, in Case 2b, where the RLV is the booster, the sensitivity is only 0.95. A 10% growth would therefore result in a more acceptable 9.5% reduction in payload. The payload would be reduced from 60,000 lb (171 passengers) to 54,300 lb (155 passengers). The sensitivity of the vehicle design has a profound effect on the development and production cost of the vehicle and on the propellant cost per passenger as payload is reduced due to weight growth. More often than not, as weight growth occurs, the designer will choose to compromise robustness and operability features to maintain promised payload weight. This was the case in the Space Shuttle and caused some severe compromises in reliability and maintainability.

Table 3. Vehicle Economic Evaluation
Table 4. Vehicle Sizing and Design-Results

Scoring: The concept with the lowest growth sensitivity received a score of 10. All other concepts received lower scores in proportion to their relative sensitivity. Thus, a concept that incurs five times the payload loss for the same percentage increase in dry weight receives a score that is divided by five, or two points.

TPS Acreage. Thermal Protection System (TPS) Acreage is a major indicator of the flight maintenance that must be performed on the vehicle. Typically, close inspection and maintenance of the TPS area is a major turn time and maintenance driver of reusable vehicles. Thus, all else being equal, systems with larger TPS areas per pound of payload will cost more per passenger carried. Note that the TPS on both the orbital RLV vehicle (Case 1) and the RLV Pop-Up stage (Case 2) endure similar thermal environments. The RLV reenters at approximately 12,000 fps and with a steep aerodynamic pullout, which causes heating similar to reentry. However, the augmentation booster of Case 3 reaches only about Mach 3 and therefore does not require any TPS. All orbiter stages, of course, were assessed the full acreage penalty for TPS.

Scoring: The concept with the smallest TPS acreage per pound of payload received a score of 10. All other concepts received lower scores in proportion to their relative TPS acreage.

TPS Robustness is a measure of the designer's ability to design a low-maintenance TPS. TPS robustness affects the level of maintenance per flight, and turn-times for the vehicle, as well as the replacement life of the TPS system, and thus the cost per flight associated with TPS. For a given vehicle cross-range requirement, two factors are considered the primary discriminators of robustness 1) TPS loading defined as reentry weight of the orbiter divided by the wing exposed area and 2) the weight sensitivity of the TPS measured by TPS weight divided by payload weight. The score is a function of the product of these discriminators. We infer that a system requiring a significant quantity of TPS per pound of payload and also enduring a stressing reentry environment will have high TPS maintenance costs.

Scoring: The concept with the smallest value of TPS loading ( TPS weight/pound of payload received a score of 10. All other concepts received lower scores in proportion to their relative TPS loading and weight sensitivity.

Payload Weight/Dry Weight is a primary measure of the cost of vehicle production per pound of payload. The other significant controlling factor is robustness, which determines the vehicle life over which the production can be amortized. This factor is taken into account in the vehicle and TPS robustness criteria. Because the technologies employed in this class of vehicles are generally the same, the payload to dry weight ratio is a good indication of the relative production cost per pound of payload of each concept. It can be argued that the production cost is higher in two-stage systems even with the same weight. However, in the two-stage cases explored in this study, all rely on a simpler, historically less costly LOX/JP system for the additional stage. This is particularly true for the JP/ LOX booster, which does not require a payload bay and stages at Mach 3 or below. It can use existing engines and does not require a TPS system. As a result, it is believed that using a constant manufacturing cost per pound for both systems may actually be slightly conservative in estimating the costs of the LOX/JP stages.

Scoring: The concept with the highest value of payload to dry weight received a score of 10. All other concepts received lower scores in proportion to their relative payload to dry weight ratio.

Propellant cost per pound of payload was determined to be a dominant driver in a well designed RLV optimized for the highest flight rates. Ultimately it is a major driver in the lowest ticket prices that can be offered. For example, the propellant cost per pound of payload (assuming $2.00/lb of LH2) for the robust SSTO RLV (Case 1b) is $31 per pound of payload, or $10,850 per passenger just for propellants. Economic analyses [1] indicate that this cost alone could preclude a viable tourism industry. By comparison, the best-performing RLV configuration (Case 3b) costs only $7 per pound of payload, or $2,450 per passenger for propellant costs.

Scoring: The concept with the lowest propellant costs per pound of payload received a score of 10. All other concepts received lower scores in proportion to their relative propellant costs.

Engine Maintenance per flight is another key cost and operability driver. Maintenance costs are strongly related to the number of engines that must be maintained, as well as the robustness of the engines themselves. Since robustness was treated earlier for the system here, we focus on the number of engines that must be maintained per pound of payload. It is believed that engine maintenance is a strong function of number of engines and a weak function of engine scale.

Most of the engines in the concepts explored were of similar scale (within a factor of two), allowing scale to be eliminated as a key factor in determining engine maintenance costs. The only difference is the upper-stage engines in Configuration 2, which are smaller in size and must be ignited during flight rather than on the ground.

Scoring: The concept with the highest value of payload to number of engines employed received a score of 10. All other concepts received lower scores in proportion to their number of engines to payload ratio.

Vehicle Maintenance per flight is a key cost and turn time driver on RLVs. This MOE is based on the maintenance required per flight per pound of payload. Thus, if two vehicles are assessed to have similar maintenance requirements but one carries twice the payload of the other, it will receive a proportionally higher score. Vehicle maintenance is a complex function of the following items: robustness, propellants employed, engine types and arrangement, integration complexity, health management, degree of testing, and recovery and abort approach.

To minimize integration complexity on staged vehicle options, crossfeed of propellants between vehicles is not used in any of the designs. Also, when a second stage was employed, it was kept relatively small in size and simple in design compared to the VentureStar RLV. Since staged designs typically doubled or tripled the payload delivered with a much lower increase in the maintenance required, they tended to fare well in the evaluation. In two-stage systems, a significant part of the launch system returns rapidly to the site for processing so that it should be completely processed by the time the orbiter stage arrives at the launch base. This has a positive effect on vehicle turn time and staff size. It is likely that the booster stage could be launched twice per day compared to only once per day for the orbiter stage. This would reduce the booster fleet size and facility requirements.

Scoring: The concept with the lowest maintenance requirements per pound of payload received a score of 10. All other concepts received lower scores in proportion to their r elative maintenance requirements.

Facility Requirements per flight per pound of payload is also a significant cost driver. These requirements are driven by 1) the individual vehicle turn times and scheduling that can be achieved, 2) the number and size of bases and facilities required to support the operations, 3) the number and complexity of operations required, and 4) the degree of autonomy and self diagnosis on board the vehicle.

Scoring: The concept with the lowest facility requirements per pound of payload received a score of 10. All other concepts received lower scores in proportion to their relative facility requirements.

Overall System Complexity is a factor that drives production costs, operations/maintenance costs, and facility costs. Although every attempt to reduce system complexity should be taken, it is important to realize that robust and well developed systems are much more critical to cost than system complexity. To highlight this point, commercial aircraft with far more complex and exotic engines and wing mechanisms operate routinely at less than $1 per pound of payload including production, maintenance, facilities, insurance, and fuel. Analysis at Aerospace has shown that robustness, design maturity, and abort options are far more important discriminators of operating costs than is vehicle complexity.

Scoring: The concept with the lowest level of complexity received a score of 10. All other concepts received lower scores in proportion to their relative level of complexity.

Weighting. The weighting column in Table 3 reflects the engineering judgment of the study participants based on prior analyses as to the relative importance of the ten MOEs in determining the viability of a concept to support a space tourism industry. A limited sensitivity analysis was conducted by varying the weighting distribution of the MOEs. Although this resulted in altering individuals scores, it had only a minimal effect on altering the relative order of ranking of the concepts. In fact only two concepts with almost identical scores using the baseline weighting criteria switched order when the weighting distribution was varied. Since the analysis was performed using an Excel spreadsheet, the weightings can be changed quickly to see the impact of individual and organizational biases on the relative importance of each MOE.

7. RESULTS AND OBSERVATIONS

Three major sets of VentureStar RLV-based configurations were studied as options for a space tourism optimized vehicle. Full engine-out, safe-abort capability was a requirement in all configurations.

Tables 3 and 4 show the evaluation results. Table 4 shows the results of the vehicle sizing, payload, and component weight comparison. Table 3 shows the separate scores of the seven MOEs quantifiable from Table 3 and the values assumed for the scores that are not quantifiable. The scores shown in the "weighted total" row of Table 3 are based on the weighting distributions shown in the "weighting" column.

Configuration Results

Configuration 1a, the reference VentureStar SSTO RLV, was estimated to lift 35,533 lb to the reference tourism orbit of 220 nm by 35 degrees inclination. This equates to 101 passengers when fully loaded, assuming 350 lb per passenger, including baggage and consumables. Although this configuration addresses the needs of traditional spacelift, it was not optimized to perform the tourism mission and thus received a lower score than the configurations specifically targeted to address them. The major weaknesses of this vehicle for the tourism requirements include the low design margins (25%) for robustness (vs 40% for the Shuttle and greater than 50% on commercial aircraft), the high sensitivity to weight growth, the large TPS acreage per pound of payload delivered, the large propellant cost per pound of payload, and the large number of engines per pound of payload.

Configuration 1b increased the robustness margin to 50% (as with all subsequent configurations) but maintained the GLOW of the configuration 1a. The payload decreased from 35,533 lb (101 passengers) to only 19,633 lb (56). The reduction of payload nearly offset the cost reductions associated with improved operability. The very modest score increase of this configuration combined with the reduction of payload that can be lifted per mission indicate that the reference RLV (Configuration 1a ) may be near optimal for a low-traffic model.

Configuration 1c maintained the robust margins but scaled up the RLV GLOW and dry weight to the original 35,533 lb payload of Configuration 1a. Because of the favorable scaling of the propellant mass fraction (PMF) with vehicle size, the payload increases much more quickly than do GLOW and dry weight. In this case the GLOW increased only from 2.19 Mlb to 2.69 Mlb while the payload nearly doubled. The reasonable payload size of this configuration (101 passengers), combined with a robust SSTO system, yielded a significantly improved score compared to the two prior SSTO cases with only a small design departure from the reference VentureStar vehicle.

Configuration 1d was sized to take advantage of the improving payload to GLOW and dry weight ratio of SSTO designs. This configuration was sized to achieve 100,000 lb of payload to the tourism orbit. The greater ratio of payload to dry weight and reduced propellants per passenger allow ticket prices about half of those for Configurations 1b and 1c. Past travel experience of the airline industry indicates that every time the ticket price is reduced by 20%, the traffic doubles.

Configurations 2a, 2b, and 2c are variations of a concept in which the VentureStar RLV is combined with a small LOX/JP reusable upper stage (orbiter). This configuration of vehicle generally outperformed the SSTO vehicle on an economic basis. This also reduces the propellant cost per pound of payload, because JP/ LOX is much less expensive than LH2/ LOX. The improved performance of this configuration allowed a robust system that carries three times the payoad of the robust VentureStar SSTO (Configuration 1b). The ability to rapidly recover the RLV allowes it to launch twice as often as the reference RLV, which also helps the economics.

Of these three configurations, 2b was the clear winner and second best overall. The slightly larger GLOW to carry turbojets and jet fuel for direct return to the launch site resulted in operability and processing time advantages compared to 2a and 2c that more than offset the small increase in GLOW. Since the volume inside the payload bay is available for the stage, the physical size of the RLV would be the same as the reference VentureStar vehicle.

Configuration 2a required a downrange landing site where the RLV would be outfitted with an engine and jet fuel kit and later flown back to the launch base. This base cost and increase in turn time resulted in a lower score than simply paying the performance penalty for direct booster flyback.

Configuration 2c also did not fare well compared to 2b. Payload loss in the retrograde launch plus the need for two launch sites, and two orbital resorts offset the turn time advantages.

Configuration 3 employs a moderately sized LOX/JP booster in parallel burn with the RLV during the initial segment of the flight. Configuration 3a and 3b are virtually identical except that the LOX/JP booster stage in configuration 3b is sized approximately 50% larger than 3a to increase the payload from 60 klb to 100 klb and also increase staging velocity to approximately Mach 3. This increase in payload only resulted in a 20% increase in dry weight of the booster and 4% of the overall launch system.

The larger configuration was also able to provide engine out capability on either the booster or orbiter stage without requiring an additional RD-180 (or equivalent) engine. The increase in paying passengers per flight from 171 to 286 for an negligible increase in system size made Configuration 3b much more economical than Configuration 3a.

Configuration 3 options are similar to those of option 2 with several exceptions. While the LOX/JP stages of options 2 and 3 are of generally similar size and dry weight, the configuration 3 booster is far simpler than is the orbital passenger carrying LOX/JP upper stage. The RLVs in Cases 2 and 3 are similar in design except that the RLV in Configuration 2 is subjected to slightly more harsh reentry thermal and structural load conditions due to the steep pull out from the pop-up deployment.

Configuration 3 also allows both stages to return to the launch site without requiring jet engines or ferrying capability. This combined with the low staging velocity of the booster stage enable it to operate without a TPS, and use existing engines helped the score. Another advantage is that where Configuration 2 limits the GLOW and thus the payload of the system, Configuration 3 is limited only by the staging velocity and/or the desire to avoid crossfeed of propellants between the two stages. This option also minimizes the modifications of the RLV compared to the pure SSTO design. It is possible that this booster (Augmenter) stage could be added to the RLV after initial, traditional spacelift operations begin.

One area in which Configuration 3 fares less well than Configuration 2 is vehicle growth growth sensitivity. Since only a small percentage of the vehicle goes to orbit in Case 2, it was less sensitive than when the more massive RLV was used as the orbiter. Of course both Configurations 2 and 3 out perform the Configuration 1 options since all of the dry weight does not have to be delivered to orbit. Even with this offset, the ability of Configuration 3b to carry more payload with a fixed size RLV, gives it an advantage over Configuration 2b in a high-traffic situation.

Figure 2 presents graphically the values assumed in Table 5. These relationships are a guide to what RDT&E and Operation Planning actions are required to make the higher traffic levels economically viable.

At traffic rates of 10 flights per year the amortization of RDT&E costs is the major cost. This amortization cost item can be held down only by minimizing development or by government cost sharing. The infrastructure for operations and the purchase of vehicles are large factors. Maintenance and insurance are smaller factors, with the insurance based on 99% reliability. At this low traffic level, the cost of propellant is so low that the fuel cost is not even a consideration. However, the choice of propellant may still have significant effects on the cost of maintenance, infrastructure (ground LH2 handling equipment), and reliability. It is the direct cost of the fuel that is negligible.

For flight rates up to 100 per year and more, RDT&E remains the dominating cost. Note, however, that we have assumed rapidly increasing RDT&E costs with increasing traffic rates. These represent the costs of intensive engineering to provide the required increases in reliability (Table 5) and reduction of maintenance hours per flight. These are not only desirable things to do; if they are not done properly, the insurance and maintenance costs will dominate the launch cost.

At flight rates approaching 1,000 per year, amortized costs still dominate the overall costs. At 5,000 to 10,000 flights per year, amortization of RDT&E continues to dominate the cost breakdown. This is partly because we increase the RDT&E estimate to $10-$12 billion with the increasing emphasis on reduced maintenance hours per flight and extreme reliability. If these engineering efforts fail, insurance and maintenance costs will dominate the economics of the system and prevent further cost reductions with increased traffic.

Also, at 1,000 flights per year, the propellant cost for the LH2/ LOX becomes significant, and at 10,000 flights it threatens to limit additional cost reduction. Remember that the propellant cost is a recurring per-flight cost, and is not significantly reduced with increased traffic. It is in this traffic range that the substitution of a TSTO vehicle using JP instead of LH2 appears to be required. The cost of LOX in either case is a small factor.

Table 5. Reference Reusable Vehicle Supply

Flight Rate/Yr 10 50 100 1,000 5,000 10,000 100,000 1,000,000
RDT&E ($B) 2 6 6.5 8 10 12 18 24
Number of Flights 30 150 300 3,000 15,000 30,000 300,000 3,000,000
RDT&E $/lb 2222.2 1333.3 722.2 88.9 22.2 13.3 2.0 0.3
Prod Cost/Unit (Avg $B) 1.5 1.2 1.1 0.8 0.7 0.6 0.4 0.3
Life (Flights) 50 150 300 2,000 4,000 6,000 9,000 12,000
Prod $/lb 1000.0 266.7 122.2 13.3 5.8 3.3 1.5 0.8
Turn Shifts 60 14 8 2 1 1 0.5 0.5
Crew Size 500 500 500 300 250 200 150 100
Maintenance Hrs/Flt 240,000 56,000 32,000 4,800 2,000 1,600 600 400
Maintenance $/lb 640.0 149.3 85.3 12.8 5.3 4.3 1.6 1.1
Cost per Failure ($B) 2 2 2 2 2 2 1.5 1
Reliability 99% 99.6% 98.8% 99.96% 99.99% 99.995% 99.999% 99.9998%
Insurance $/lb 666.7 266.7 133.3 26.7 6.7 3.3 0.5 0.1
Infrastructure Cost ($ M/yr) 300 300 350 400 400 600 1200 3,000
Flights/Yr 10 50 100 1,000 5,000 10,000 100,000 1000,000
Infrastructure $/lb 1000.0 200.0 116.7 13.3 2.7 2.0 0.4 0.1
Fuel Cost ($/lb) 3.00 3.00 3.00 2.00 0.20 0.20 0.20 0.15
Fuel Quantity/Flt (lb) 330,000 330,000 330,000 330,000 595,000 595,000 595,000 595,000
LOX Cost ($/lb) 0.045 0.045 0.045 0.045 0.045 0.045 0.045 0.045
LOX Quantity/Flt (lb) 1,970.000 1,970.000 1,970.000 1,970.000 1,489,000 1,489,000 1,489,000 1,489,000
Propellant $/lb 36.0 36.0 36.0 25.0 6.2 6.2 6.2 5.2
TOTAL $/lb 5564.8 2252.0 1215.7 180.0 48.9 32.5 12.2 7.5
Total Operating Cost/yr ($B)1.39 2.81 3.04 4.50 6.12 8.12 30.45 188.5
Figure 2. Costs at Low Traffic Levels

There has been some ongoing discussion about the future cost of hydrogen if its use as a spaceplane propellant constitutes a major portion of the market. (At present, it is a small component of the total hydrogen market.). In this study, we used a price of $3.29/lb. Prices as low as $1.50 to $2.00 have been discussed, assuming large production rates [7]. We believe a market price of approximately $1.50/lb for delivered LH2 to be the practical floor that can be expected as a result of the energy and feedstock requirements to produce and transport LH2. A large portion of the cost of liquid hydrogen is associated with the energy required to produce it. The methane feedstock, cost of production facilities, and transport and storage also contribute to the cost. In addition, the cost of energy or feedstock fuels seems unlikely to decrease in the future. However, if we assume a $2.00 cost/lb is realistic, the propellant cost per pound of payload would still be about $23/lb of payload, and would still dominate the costs for traffic of more than about 5,000 flights per year. Thus, for traffic levels projected for space tourism, unless the economics of hydrogen production can be greatly improved, the switch to hydrocarbon fuel suggested in the paper on design consideration [8] appears mandatory. If we include the possibility of a secondary use of the spaceplane as a hypersonic transport for passengers and air express packages between air traffic centers, we are considering traffic in the range of 10,000 to 100,000 flights per year, which would make fuel the dominating cost.

Most of the input numbers in Table 5 are estimates based on aeronautical and aerospace experience. They tend to be dictated by a combination of engineering judgment and the economic desire to keep any recurrent cost component from dominating the cost, by spending additional RDT&E money on components that threaten to dominate. Seen in this light, the reliability inputs under "Insurance" are approximately what is required to keep the insurance cost at about 10-20% of the total transport cost. The reliabilities must become very high to achieve the extremely low transportation costs possible at very high traffic rates. Similarly, the inputs under Production, Maintenance, and Infrastructure are graduated by a combination of technical feasibility and cost control, and are indicative of the performance required to reach the desired goals. The RDT&E costs in the top row of the table are our estimates of the increasing RDT&E costs of obtaining the assumed

performances and costs. Fortunately, these very much larger RDT&E costs can be easily amortized at the very high flight rates. Additional analysis is required to determine if the estimated increases in RDT&E with flight rate are overly conservative. Any government investment in enabling technologies such as robust engines and subsystems will offset the commercial RDT&E investments, thus increasing the financial feasibility of addressing high flight rate markets.

The traffic ranges are somewhat related to the particular mission; that is, flight rates of roughly 1,000 to 10,000 per year would be expected from space tourism, while Earth-to-Earth hypersonic transport might be expected to provide about 100,000 flights per year.

Implementation Options

Three major VentureStar RLV-based configurations were studied as options for a space tourism optimized vehicle. The first group of options included variations of the baseline RLV in which levels of vehicle robustness were increased and payload was varied. The second major option utilized the VentureStar RLV in a pop-up mode in which it was combined with a small LOX/JP reusable upper stage. The third option augmented the VentureStar RLV with a fly-back parallel burn LOX/JP booster stage. Full engine-out, safe-abort capability was a requirement in all configurations.

The best-performing configuration measured against our MOEs was Configuration 3b. The addition of a simple LOX/JP augmentation booster burning in parallel with the VentureStar RLV to approximately Mach 3 resulted in an increase in payload from 19,633 lb (56 passengers) to 100,000 lb (286 passengers) using the same size robust RLV design. This option yielded a score of 929 points compared to the reference nontourism-optimized RLV of 355 points and the robust RLV of the same size with a score of 395 points. This configuration outperformed Configuration 2b because using the LOX/JP stage as the booster resulted in a simpler LOX/JP stage, while providing more payload and more straightforward recovery. Configurations 1d, the 100,000 lb robust SSTO, and 2b, the pop-up RLV employing jet cruise back to the launch base, yielded nearly identical scores of 750 and 742 points, respectively.

Whereas the pop-up option can be a variant of the planned VentureStar RLV, the 100,000 lb robust SSTO requires a design and commensurate investment in a system of almost double the dry weight and GLOW of the currently proposed VentureStar to be developed up front. The alternative is to develop a scaled-up RLV vehicle at a later time to address space tourism and heavy lift requirements. The uncertain nature of the tourism business may prohibit such a risky approach, however. It was realized that the reference RLV may be well matched for the traditional spacelift missions based on the relatively small score increase from 355 to 395 points for robustness when the GLOW is constrained to that of the proposed RLV. However, the significant increase in score to 532 points by scaling up the VentureStar slightly from 2.19 to 2.69 Mlb to improve robustness and reduce operating costs may be well worth the small additional up-front investment.

The configurations explored in this paper open up a variety of RLV implementation options that should be considered. If the objective of the VentureStar is limited to address traditional spacelift only, the current development and implementation path may be optimal. If, however, it is envisioned that the RLV presents an opportunity to open space to the public and create significant demand elasticity, then the implementation options should be rethought.

Three major implementation options have been considered. Implementation option 1 is to initially develop the reference RLV with the non-robust margins to maximize payload and revenue for the traditional spacelift missions. Once the RLV is operating, the LOX/JP augmentor stage could be developed, which would allow payloads of over 100,000 lb to be delivered to orbit on a single launch. A robust variant of the RLV could then be added to the fleet, which would allow payloads to 100,000 lb (286 passengers) and provide the higher levels of operability and safety needed to support the tourism industry. The advantage of this option is that it limits up-front development, but it requires a re-design of the RLV to include the robust margins. This option can also be applied to the pop-up RLV development should this option be preferred to the augmented booster configuration.

Implementation option 2 includes scaling the RLV slightly to initially include the robust margins needed to later support the tourism market. The development of the augmenter booster could be delayed until the RLV was operational. This will also eventually enable a robust operable RLV that can support both the tourism industry as well as provide a very low-cost, heavy lift reusable launch option. The advantage of this option is that it requires only a single robust RLV design to address both traditional and tourism mission needs. The disadvantage of this option is that it will slightly increase the cost of the initial RLV development but will reduce its operating costs.

A third implementation option is to scale up the planned SSTO RLV initially to accommodate 100,000 lb class payloads and include robust design margins. This approach is probably not viable since it will significantly increase the initial development cost of the RLV and depends on the uncertain tourism market to justify the size and expense of the system. Although, this provides a long-term SSTO solution, it does not seem to provide a better solution during the operation phase a than do option 2b or option 3c.

A limited sensitivity analysis was also conducted on the relative weightings of the MOEs. Although individual scores varied slightly, they did not significantly modify the conclusions shown here.

It is important to note that this was a limited scope effort, and the options that appear promising should be assessed using a more comprehensive and more traceable analysis as mentioned earlier. Specific integration issues such as the interaction of the orbiter or booster stage on the aerospike plume and its resultant effect on the propulsion performance and controllability of the vehicle would also have to be explored. Similarly significant issues exist if option 2 is employed, in which high-velocity staging with significant dynamic pressures followed by safe recovery of the RLV must be endured. These issues must also be explored if options 2 or 3 are to be seriously evaluated either as an initial capability or as a capability-enhancing option. In addition the costs, schedule, and technical and mission satisfaction issues associated with each of the above implementation options should be explored in a more rigorous analysis before the RLV development.

8. CONCLUSIONS
  1. The present VentureStar configuration is reasonable if only the present traffic model is considered.
  2. The addition of either a pop-up upper stage or a booster lower stage greatly improves the operability and economics at traffic levels typical of space tourism.
  3. The economics is greatly improved by moderate GLOW increases because of the non-linear increase of payload.
  4. The most attractive alternative found so far uses a JP/ LOX booster lower stage to augment the RLV and provide a payload of 100,000 lb in a robust design.
  5. The study should be continued more rigorously, using the full suite of analytical tools now available.
ACKNOWLEDGEMENTS

The authors wish to express their appreciation to Glenn Law, for Vehicle Sizing and Mass Properties Analysis, and to John Skratt for Systems.

REFERENCES
  1. J P Penn and [an error occurred while processing this directive]C A Lindley[an error occurred while processing this directive], October 1997, "[an error occurred while processing this directive]Requirements and Approach for a Space Tourism Launch System[an error occurred while processing this directive]", 48th International Astronautical Congress, Turin, Italy
  2. [an error occurred while processing this directive]C A Lindley[an error occurred while processing this directive] and [an error occurred while processing this directive]J P Penn[an error occurred while processing this directive], 1994, "[an error occurred while processing this directive]Combined Exo/Endoatmospheric Transport Alternatives[an error occurred while processing this directive]", Space Technol., Great Britain 14(1), 99-107
  3. P Collins, [an error occurred while processing this directive]R Stockmans[an error occurred while processing this directive] and [an error occurred while processing this directive]M Maita[an error occurred while processing this directive], 1995, "Demand for Space Tourism in America and Japan, and Its Implications for Future Space Activities", Advances in Astronautical Sciences 91, 601-610
  4. [an error occurred while processing this directive]T F Rogers[an error occurred while processing this directive], 1995, "[an error occurred while processing this directive]Space Tourism: The Perspective from Japan and Some Implications for the United States[an error occurred while processing this directive]", [an error occurred while processing this directive]Journal of Practical Applications in Space[an error occurred while processing this directive] 6(2)
  5. [an error occurred while processing this directive]R A Hickman[an error occurred while processing this directive] and [an error occurred while processing this directive]J Adams[an error occurred while processing this directive], 3-5 December 1991, "[an error occurred while processing this directive]Operational Design Factors for NASP-Derived Vehicles[an error occurred while processing this directive]", Orlando, Florida, [an error occurred while processing this directive]AIAA[an error occurred while processing this directive]-01-5081, Third Int. Aerospaceplane Conf.
  6. [an error occurred while processing this directive]P[an error occurred while processing this directive] Czysz, personal communication, University of Missouri, St. Louis, Missouri, 1991
  7. [an error occurred while processing this directive]J P Penn[an error occurred while processing this directive], personal communication, Christian Lenci and Jeffrey Richards, Praxair, Inc., Tonawanda, New York, 1997
  8. [an error occurred while processing this directive]J P Penn[an error occurred while processing this directive], 1996, "[an error occurred while processing this directive]SSTO vs. TSTO Design Considerations[an error occurred while processing this directive]", in Proc. Space Technology and Applications Int. Forum, 1st Conference on Next Generation Launch Systems, Albuquerque, New Mexico Conf # 960109 [an error occurred while processing this directive]M[an error occurred while processing this directive]. S. El-Genk, ed., American Institute of Physics, New York, AIP Conf. Proc. No. 361, p. 551 [an error occurred while processing this directive] [an error occurred while processing this directive] [an error occurred while processing this directive]

    [an error occurred while processing this directive]Jay Penn[an error occurred while processing this directive] is a Senior Project Engineer and heads the Reusable Launch Office in the Space Launch Operations division of The Aerospace Corporation, where he has been employed for over 13 years. As a Systems Engineer/Project Manager specializing in reusable and low-cost launch systems and technologies, he has been actively involved in the technical and programmatic management of a wide variety of spacelift projects, including [an error occurred while processing this directive]X-33[an error occurred while processing this directive], DC-XA, DC-X, SDIO-[an error occurred while processing this directive]SSTO[an error occurred while processing this directive], [an error occurred while processing this directive]EELV[an error occurred while processing this directive], Delta II, Have Region, ALS, NLS, Spacelifter, and Shuttle, as well as commercial [an error occurred while processing this directive]RLV[an error occurred while processing this directive] ventures. He has also performed numerous launch and upper stage architecture and system engineering studies and analyses. Prior to employment at Aerospace, he worked at NASA/JSC in Shuttle Mission Operations for over 4 years. After graduation from Rutgers University College of Engineering, he worked at Pratt and Whitney Aircraft, performing turbojet engine design. He also holds multiple patents and awards, and has published numerous papers.

    Dr. Charles A. Lindley worked for 2 years in Weapons Research and Development at the Army Experimental Station during World War II, then earned a B.S. and [an error occurred while processing this directive]M[an error occurred while processing this directive].S. in aero engineering at Ohio State University in 1949. He designed experimental supersonic compressors for Thompson Aircraft Products from 1949 to 1952, and continued as turbomachinery consultant from 1952 to 1955. While a Caltech Guggenheim Fellow, he studied secondary flow in cascades, and received a Ph.D. in 1956. He was with the Marquardt Corporation from 1955 to 1963 and was Chief Research Consultant in his final position there. While at the Marquardt Corporation, he studied rotating stall in compressors, and invented the Liquid Air Cycle Engine in 1958 and numerous other air-breathing cycles for space boost and cruise missiles. In addition, he helped originate the Scramjet concept, and designed various [an error occurred while processing this directive]SSTO[an error occurred while processing this directive] vehicles, external burning ramjets, and various other exotic air-breathing engines. Dr. Lindley's years at The Aerospace Corporation (1963-91) included 7 years as Senior Staff Engineer, Vehicle Systems, working with reusable boosters, the Manned Orbiting Laboratory, and boost systems; 7 years as Director of Advanced Planning AF Space Systems; 6 years Associate Director Wind Energy Systems, 10 years Senior Intelligence Analyst, Project West Wing. He retired in 1991. From 1982 to 1992, he served as consultant and occasional lecturer on wind power and alternate energy at UC Santa Barbara. From 1991 to the present, he has concentrated on consulting in the fields of reusable boosters and propulsion systems. Dr. Lindley holds numerous patents and has published many papers.

    *Dr. Lindley has retired from The Aerospace Corporation. However, he continues to periodically provide technical assistance to the Corporation. [an error occurred while processing this directive] [an error occurred while processing this directive]