MILLENNIUM EXPRESS

First published as:
AIAA 2001-3962

37th AIAA/ASME/SAE/ASEE Joint Propulsion Conference,
8-11 July 2001, Salt Lake City, Utah

Len Cormier
Third Millennium® Aerospace, Inc. and PanAero, Inc., Washington, DC

Member, AIAA.

Abstract The Millennium Express is a two-stage-to-orbit (TSTO) space transport. This space transport is designed around existing engines and in-hand technology in order to constrain pre-operational investment costs to $200 million. We view this constraint on investment to be a basic requirement. There is no currently suitable market for a commercial space transport. Accordingly, any near-term, truly commercial investment must be small enough to allow an integrated program for development of both the space transport and one or more applications enabled by the space transport at a modest traffic level.

During the past year, we have been able to increase payload to LEO to about 2 tonnes--while holding the line on estimated pre-operational investment costs and recurring operational costs, which we estimate to be about $300,000 per flight at a flight level of about 200 flights per year. The current orbiter concept has a gross mass of about 100 tonnes and is powered by two Aerojet / Kuznetsov NK-39 LOX / kerosene rocket engines, plus one Aerojet / Lyulka D-57 LOX / hydrogen engine. Follow-on versions could use improved versions of these engines, or possibly the Pratt & Whitney RL50 or Boeing Rocketdyne MB-60.

Our currently preferred carrier stage is a modified Antonov An-22 Antheus, with four Kuznetsov NK-12, 15,000 hp turboprop engines. The turboprop engines are ideal for climbing efficiently at a dynamic pressure less than 10,000 Pa. By constraining dynamic pressure during climb and by launching the orbiter subsonically at 15-km altitude with rocket assist, the orbiter wing is designed essentially by reentry and landing mass.

Design to Cost Approach

Our basic design approach is to constrain pre-operational investment to $200,000,000 and to treat this constraint as a primary design requirement. Even payload becomes a fallout parameter-- especially when the basic business strategy is a vertical integration of vehicle and a primary application.

In general, we have designed all versions of our Millennium Express to this $200 million investment constraint. Moreover, all versions are designed to operate at a recurring cost per flight of about $300,000 or less at a traffic level of about 200 flights per year.

These cost goals are only somewhat arbitrary. The recurring cost per flight is based upon our estimates of what we think is realistic for a medium-size space transport. The constraint on pre-operational invest-ment is chosen to allow amortization at a rate that is comparable to recurring costs at a traffic level of 200 flights per year. A traffic level of 200 flights per year is about the maximum that might be achieved with an initial fleet of perhaps four operational prototypes. A traffic level of 200 flights per year also seems appropriate for an integrated business model for a specific initial application. As outlined in previous papers^{1, 2, 3, 4}, we feel that an appropriate amortization rate is about 30 percent per year during a ten-year operational period following a four-year development period; the cost methodology is detailed in the earliest of these papers⁴. Thus, 30 percent of $200 million amortized over 200 flights per year results in an amortization cost of about $300,000 per flight, which is comparable to our recurring cost per flight.

No parts of the system are expendable. We use the term "space transport" without an "expendable or reusable" qualifier. After all, there is no need to qualify a transport aircraft as a "reusable air transport." Using the term expendable or reusable for launch vehicles is an absurdity that should have become obsolescent three decades ago. This absurdity, in turn, has led to serious misconceptions with respect to the practicality of space transports. Two of the most important of these misconceptions are: a) the perceived market for a space transport; and b) the perceived technology requirements for a space transport.

The first misconception is that the perceived market for a space transport is an extension of the market that has developed for expendable launch vehicles (ELVs). The market that has developed for ELVs is completely inappropriate for a space transport. Accordingly, the existing ELV market should not be allowed to impact the design of a space transport--or the perceived need for a space transport. Unfortunately, this means that there is no established market for a space transport. Accordingly, a commercial approach to developing a space transport must likely allow for a vertical integration with the initial, primary application. This, in turn, leads to a basic constraint on the amount of money that can be invested in a near-term space transport--and, indirectly on the size of a near-term space transport. It probably also constrains the design to in-hand technology and existing engines. This essentially rules out consideration of an SSTO as a first-generation commercial space transport.

Non-commercial approaches have not been successful in the past --and are not likely to be successful in the future, because of the lack of business discipline in subsidized approaches.

The second misconception is that a space transport is going to require a lot of advanced technology. This misconception stems from an unwarranted belief that only an SSTO can provide the operational benefits required to justify reusability. Even many approaches to a two-stage space transport tend to rather exotic first stages with staging at hypersonic velocities. We believe that the real requirements for a near-term space transport are: a) low pre-operational investment costs; b) low recurring costs; and c) relatively simple, practical operations. We further believe that these requirements can be met by "assisted SSTO" concepts--particularly TSTO concepts that use a relatively simple subsonic carrier capable of staging at relatively high altitudes. The primary functions of such a carrier stage are to relieve orbiter design requirements and to make operations simple and practical, while not adding greatly to overall development and operational costs.

Current Vehicle Description

Figure 1 depicts the orbiter stage of the current version of our "Millennium Express." This rendering depicts the orbiter stage as it might look just prior to approach and landing at the original launch site.
The Millennium Express is a two-stage-to-orbit space transport that stages subsonically at about mach 0.5 at an altitude of 15-km altitude and at a high flight path angle. Dynamic pressure during climb does not exceed 10,000 Pa. This--combined with the 2100-Pa dynamic pressure at staging--allows us to avoid large aerodynamic forces that would otherwise design the orbiter wing. The preferred concept for the carrier stage uses a large straight wing and large turboprop engines to allow efficient climb while constraining aerodynamic forces on the orbiter wing during the climb to altitude. Rocket assist above about 10-km altitude enables climb at a 25 flight path angle to reach 15 km and relatively thin air prior to staging.

Figure 2. Two-View of Passenger Version of Orbiter.
The orbiter stage follows an essentially ballistic trajec-tory after separation at a high altitude and high flight path angle. Thus, the orbiter wing is used only for reentry, approach and landing. This is one of the key design tricks for the Millennium Express. The savings in orbiter wing mass is about 6600 kg--or about 55 percent of burnout mass. Our performance and cost goals are dependent upon good system design and some design tricks -- not on exotic technology^{1, 2, 3,4}.

Figure 2 shows a two-view of the 16-passenger version of the orbiter. There are also provisions for a pilot and copilot / flight attendant.. The orbiter stage is powered by one Lyulka D-57 LOX / hydrogen and two Kuznetsov NK-39 LOX / kerosene rocket engines. These engines are offered by GenCorp / Aerojet of Sacramento, California in accordance with a long-standing marketing agreement with the Russian engine companies. These engines were first manufactured in the mid-1970s, and have been moth-balled since that time. Nonetheless, specific impulse of the D-57 is 4472 m/s and thrust is about 370,000 newtons⁵; specific impulse of the NK-39 engines with the fully expanded nozzle is 3452 m/s and thrust is in the 400,000 newton class⁶. We envisage this engine combination for the first four orbiters. Later orbiters may use updated versions of these engines, or, possibly, a Pratt & Whitney RL50 or a Boeing Rocketdyne MB-60 in place of the D-57.

The flight deck and cabin areas comprise the upper portion of the nose section. The cabin floor ramps upward to make full utilization of the available space. The aft portion of the cabin is elevated sufficiently to allow a small elevator aft and a lower deck with two "partial-capability" toilets. The spherical helium tank forms the nose of the vehicle. The nose gear and forward kerosene tank are located below the flight deck; the remaining kerosene tank is below the forward portion of the cabin. The large liquid hydrogen tank for the low density hydrogen is located just behind the cabin; the liquid oxygen (LOX) tank is aft of the hydrogen tank and forward of the engine compartment. There is very little shift in the center of gravity during rocket burn. Note, however, that the liquid oxygen that comprises about 70 percent of the gross mass can absorb most of the engine thrust in a short distance.

Figure 3 shows a two-view of the cargo version of the orbiter. There is a high degree of commonality between the passenger and cargo versions. However, whereas the passenger version is designed to reenter and land with a full payload, the cargo version is designed to reenter and land empty for most cargo missions. This necessitates moving the wing aft about 2 meters for proper balance. At gross mass, the cargo version is too stable. However, the orbiter flies a basically ballistic trajectory during exit. Accordingly, the overly stable condition at separation should not be a problem. The wing position can be changed on the ground with especially adapted fittings that carry the main bending loads through the liquid oxygen tank. The other main difference from the passenger version is that the kerosene tank is shallower and extends under the full length of the cargo bay.

The cargo door accommodates cylindrical payloads up to 3 meters in diameter and 3.6 meters in length. A conical shaped payload could be somewhat wider at one end. For structural reasons, the door is a single piece, not a clamshell arrangement as on the Space Shuttle. As with the passenger version, kerosene occupies the lower portion of the nose section that includes the payload bay. However, the cargo deck is significantly lower than the deck of the passenger cabin, and the kerosene -- plus the nose gear -- occupies the entire lower portion of the nose section. The cargo version deck is indented to accommodate a 3-meter diameter preassembled payload. A special adaptor provides a level deck for other purposes.

Cargo payload to a 110-km, 60-degree orbit is an estimated 1800 kg. The cargo bay is suitable for a 3.6-meter long by 3-meter diameter satellite. This is the primary orbit for a potential application: a large number of self-propelled communications satellite for deployment to 800-km altitude.
Our secondary orbit is a 450-km, 15-degree orbit for assembly of geosynchronous satellites with large antennas. Payload to this orbit is about 2400 kg. Cargo flights also carry a pilot and one copilot / crew member.
Note that in the above figures, that the ventral fins have been translated forward from their aft reentry position. In their forward position, these ventral fins also serve as landing skids in conjunction with a steerable nose wheel. The X-15 used this arrangement. The aft position of these ventrals is necessary during reentry to avoid severe heating problems associated with ventral intersections with a lower surface. The ventrals are rotated upward for ground clearance prior to takeoff of the carrier stage-- or during landing with the orbiter stage still attached in the event of mission abort.

Near the end of burn, we shall use only the D-57 engine to limit acceleration and to maximize specific impulse. Aerojet plans to improve the throttling capability of all of the NK-39 engines--as they have with the similar, but larger AJ-26 engine they are providing to Kistler Aerospace for the K-1 RLV. However, our three-engine approach does not require much throttling to limit acceleration forces for missions with people on board to 3-g's.

Figure 3. Two-View of Cargo Version of Orbiter.

Carrier Stage. Our currently preferred concept for the carrier stage is modification of the Antonov An-22 Antheus large turboprop transport. This aircraft was developed 36 years ago for transporting military cargo. Aeroflot used a later version as a commercial transport. The An-22 is powered by four Kuznetsov 15,000 HP NK-12 turboprop engines and counter-rotating propellers. In case you're wondering--yes, you're right, that's the same Kuznetsov that built the NK-39 rocket engines that we plan to use in the orbiter.
Figure 4 is a three-view of the modified An-22 with the Millennium Express orbiter attached. Not much of the orbiter shows in these views; a bottom view would show more of the orbiter. Three major modifications are necessary to make the An-22 suitable for our concept for the carrier stage of the Millennium Express, as shown in figure 4.
First, we shall have to move the main gear outboard into pods suspended on pylons. A beefed-up version of the long landing gear used on the Tupolev Tu-114 or Tu-95 is an alternative to the underslung pods. This allows us to carry a winged orbiter near the center of gravity and under the An-22 wing without a conflict between the orbiter wing and An-22 landing gear.
We would then want to cut a cavity in the belly of the An-22 fuselage to carry the orbiter in semi-submerged fashion. We shall also have to beef up some of the fuselage structure that is otherwise weakened for the cutout; this should be possible, since the An-22 fuselage diameter is nearly 2 meters larger than the orbiter fuselage. We shall also have to find or to develop appropriate hard points for suspension of the orbiter and for extending the orbiter below the An-22 fuselage prior to separation. The wing carry-through structure is undisturbed.
The third major modification is to install a rocket assist system for subsonic climb from 10 km to 15 km at perhaps a 25-degree flight path angle. Some of the NK-39 engines already come with truncated nozzles; a pair of these engines may be suitable. Alternatives include pressure-fed rocket engines or hybrid rocket engines with nitrous oxide oxidizer. At the present time, we also show the orbiter dorsal vertical protruding through a slot in the upper fuselage of the An-22 behind the wing carry-through. If this proves to be impractical, then we should either make the orbital vertical somewhat stubbier, or we would fold it partly.
The An-22 is very well suited for our purposes. We want simple, straightforward operations. We also want to be able to climb efficiently to altitude at relatively low speed to preclude having the orbiter wing encoun-ter large aerodynamic forces. The Kuznetsov NK-12 turboprop engine is much larger than any other turbo-prop engine in the world. It is also used on the Tupolev Tu-95 "Bear" bomber and other aircraft, such as the Tu-114 transport derivative of the Tu-95. The An-22's large, straight, high-aspect-ratio win --along with the turboprop engines--is quite appropriate for medium- speed climb to relatively high altitude. This is highly beneficial for constraining dynamic pressure during subsonic climb and for avoiding altitude-compensation problems for the orbiter rocket engines .
The An-22 has a normal gross mass at takeoff of 250 tonnes. Although our proposed orbiter mass is about 100 tonnes, we would be using the An-22 with a light fuel load. Accordingly, we hope to be able to manage the increase from normal maximum payload of 80 tonnes. The An-22's wing span is 64.4 m; length is 57.8 m.
We have considered other potential modifications of the An-22. One of the approaches involving the least modification might be to mount the orbiter on top of the An-22; the existing twin-tail could avoid the fatigue problems associated with a single tail behind a large, top-mounted payload. However, separation would be more complex.

We have considered other potential modifications of the An-22. One of the approaches involving the least modification might be to mount the orbiter on top of the An-22; the existing twin-tail could avoid the fatigue problems associated with a single tail behind a large, top-mounted payload. However, separation would be more complex.
Perhaps the most suitable approach would involve building around the wing, engines and nacelles that are the most appropriate parts of the An-22 of interest to our project. However, this would involve a new, but small, flight deck that would also house a long nose gear. This flight deck might also serve as a partial fairing for the orbiter that would be slung under the center portion of the wing, just behind the new flight deck. For stability, we envisage twin booms and a high horizontal tail between two vertical fins. Beefed-up versions of the very long Tu-114 landing gear might be appropriate for both the main landing gear and the nose gear.
The An-22 can be legally flown into U.S. airspace. Once converted to a launch vehicle, our plan is to confine the converted An-22 to the restricted area of the launch site--where it would come under launch vehicle rules, rather than FAA aircraft rules.
There are, of course, aircraft options for the carrier other than the Antonov An-22. These include: a) the Antonov An-225, which has recently been returned to flight status; or b) some new airframe. Modification of American or European transports does not appear to be as promising because of engine and tail arrangements and other considerations. Moreover, modification of certified aircraft as launch vehicles can become quite complicated, if one intends to fly the modified aircraft outside of a restricted launch area. See our web site for configuration updates.⁷

Performance and Cost Estimates
Performance Estimates. Our trajectory calculations for low heating during exit indicate a required v_ideal of about 9450 m/s--toward which the carrier stage contri-butes about 800 m/s. With reference to the group mass estimate shown in table 1, we estimate that the orbiter delivers about 4058 m/s while all three orbiter engines are burning. The NK-39s are throttled to limit acceleration to 3 g's. The D-57 burning alone delivers the remaining 4592 m/s. This should yield a burnout mass of 12,092 kg, including the v_ideal required for transfer of 2400-kg payload to a 450 km orbit, for circularization, and for deorbit from a 15 orbit.
Cost Estimates. Table 2 presents our budget for constraining pre-operational investment to $200 million. Table 3 presents estimated recurring operational costs for the Millennium Express.
Table 2. Pre-Operational Budget
(millions of 2001 dollars)
Four orbiter operational prototypes    60
Three carrier / boosters       60
Engineering          25
Ground Testing                7
Training            3
Flight Testing             10
Base Preparations             10
Miscellaneous and Contingencies       25
TOTAL Pre-Operational Budget       200

Table 3. Estsimated Recurring Operational Costs.
(2001 dollars)
FLYING OPERATIONS 132,000
Kerosene 10,000
Jet A Fuel    20,000
Liquid Hydrogen    30,000
Liquid Oxygen    12,000
Booster Propellants       10,000
Other Fluids    5,000
Pilots & Ground Crew    20,000
Liability Bonding        5,000
Loss Reserve          20,000
MAINTENANCE       72,000
Rocket Engine Materials 40,000
Nk-12 Materials            2,000
Other Materials    10,000
Maintenance Personnel       20,000
EQUIP'T REPLAC'T          45,000
Carriers / Boosters 10,000
Orbiters    25,000
Other          10,000
BASE OPS COSTS/FLT 20,000
ABORT REFLIGHTS       2,500

TOTAL OPERATING COSTS / FLIGHT 271,500

Table 1. Millennium Express Orbiter
Group Mass Estimate (kg)
(1 D-57 + 2 NK-39 engines)--Jan 2001
STRUCTURE     6,120
Body Group 3,620
Cockpit/Payload Bay 925
Kerosene Tank    320
LH2 Tank    868
LOX Tank 1387
Engine Section    120
Wing Group                      910
Tail Group          350
Landing Gear       235
Thermal Protection Sys (TPS) 450 Contingency (10%)       555
PROPULSION 2,666
1 D-57 (143 exp)       839
2 NK-39 (114 exp.)    1392
Propulsion System    280
Attitude Control System      90
Misc.    25
Contingency 40
OTHER Equipment/Provisions 286
Batteries and Power Cond. 75
Flight Control System 75
Avionics 40
Environmental Control System 40
Miscellaneous 30
Contingency     26
EMPTY MASS 9,072
USEFUL LOAD 90,942
Pilot    100
Payload    2400
Trapped Propellants, etc.    350
Attitude Control Propellants    120
Helium    50
Usable Rocket Propellants    87,922
LH2 5,787
LOX 69,353
Kerosene 12,782
GROSS MASS 100,014
MASS AT END OF BOOST 12,092

Figure 5 compares launch-cost estimates for various types of launch vehicles. On the scale depicted on the Y-axis -- millions of kg per year -- expendable launch vehicles (ELVs) barely show. The world currently spends more than $5 billion per year on current ELVs. However, this only pays for launching about 350,000 kg of payload per year, or about a third of million kg per year. A single- stage-to-orbit (SSTO) space transport would improve launch economics greatly, if the market for such a vehicle would appear overnight. However, an SSTO might take a $6 billion investment before it could become operational. If this were a private investment spread over a four-year period and if expected return on investment (ROI) were perhaps 26 percent per year (three-year doubling time), then revenue operations would have to produce about $1.8 billion per year⁴--assuming a ten percent interest rate during the ten-year operational period. This would be before any money is spent for recurring operational costs. SSTO recurring operational costs might only be about $5 million per flight. And these would drop even more at higher traffic levels in accordance with a "learning curve." However, at traffic levels of perhaps 25 flights per year, the ROI part of the price would have to be about $72 million per flight. Note that the total tonnage at 25 flights per year would be 25 percent greater than 200 flights per year with the Millennium Express.

The main problem for an SSTO is that it will take time to transition from an ELV-based market measured in terms of hundreds of tonnes per year of high-cost payloads to a space transport-based market measured in terms of tens of thousands of tonnes per year of low-cost payloads. The SSTO is not likely to be economically attractive during this transition period.
Our Millennium Express two-stage-to-orbit (TSTO) space transport addresses this transition problem by minimizing pre-operational investment. We assume $200 million for purposes of the above chart. We also expect to hold per flight recurring operational costs to about $300,000 per flight, since our orbiter is quite small and does not rely on any exotic technology. With the economics promised by the Millennium Express, we expect to make economic sense with a relatively small amount of initial revenue, i.e. as little as $100 or $200 million per year. This enables consideration of a vertical integration strategy. One potential initial application is low-cost, high-quality telecommunica-tions for remote and low-density areas.. However, we also expect to be able to offer attractively priced space tourism services -- once the regulatory hurdles become manageable. There may be other initial applications that become economically attractive with the low-cost space access enabled by the Millennium Express.
Our payload is also much smaller than the payload for the SSTO, of course. However, our recurring cost per kg of payload to low Earth orbit (LEO) promises to be only about 40 percent that for the SSTO. The SSTO is attractive to potential customers desiring large payloads. An SSTO is also operationally attractive. However, we believe our approach to a TSTO space transport also provides for relatively simple operations. Once the transition is made to an space transport-based market, there should be room for a variety of space transports. We believe that our Millennium Express is the key to making the transition happen.

Conclusions
1. The largest barrier to the development of a near-term space transport like our Millennium Express is the lack of a suitable established market for such a vehicle. We believe that the answer to this problem is to construct an affordable integrated project that results in both the availability of the space transport and the initial "proof-of-concept" of one or more applications enabled by the space transport. We feel that this, in turn, requires that pre-operational investment in the space transport be constrained to about $200 million in order to make the integrated approach a reasonable business proposition. We have sought to show that this constraint is adequate--with appropriate management in a commercial environment--to bring a space transport like the Millennium Express to operational status. The Millennium Express would then be available to conduct initial orbital operations to demonstrate the "proof-of-concept" of an appropriate application.
2. Low-cost access to space is possible in the near-term without any dependence on exotic new technology. We have sought to prove this point by postulating a specific design approach and by subjecting this design to analysis. The specific design approach uses a modified version of an airframe that has been flying for 36 years, powered by engines that have been flying for 46 years for the carrier stage. The orbiter stage of our two-stage space transport is powered by rocket engines that have been moth-balled for 25 years. Rather than exotic technology, the technical feasibility of our concept appears to be dependent upon careful system design and highly effective "design tricks."
References
1. Cormier, Len, "Propulsion for the Millennium Express" AIAA-2000-3839, 36^th AIAA / ASME / SAE / ASEE Joint Propulsion Conference and Exhibit, 17-19 July 2000, Huntsville, Alabama.
2. Cormier, Len, "Small TSTO RLVs: Market-Building Stepping Stones to SSTO RLVs," AIAA-99-2620, 35^th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit, 20-24 June 1999, Los Angeles, California.
2. Cormier, Len, "X Van Economics," AIAA-98-3954, 34^th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit, 13-15 July 1998, Cleveland, Ohio.
3. Cormier, Len, "Bantam Boosters: The Key to Small RLVs?," AIAA-97-3124, 33^rdAIAA/ASME/SAE/ ASEE Joint Propulsion Conference and Exhibit, 6-9 July 1997, Washington State Convention and Trade Center, Seattle, Washington.
4. Cormier, Len, "The Economics and Technical Benefits of the Assist-Stage Concept for Space Launch," AIAA-96-2773, 32^nd AIAA/ASME/SAE/ ASEE Joint Propulsion Conference and Exhibit, 1-3 July 1996, Walt Disney World Dolphin, Lake Buena Vista, Florida.
5. Private communication from Aerojet.
6. http://www.astronautix.com
7. http://www.tour2space.com

MILLENNIUM EXPRESS

First published as: AIAA 2001-3962

37th AIAA/ASME/SAE/ASEE Joint Propulsion Conference, 8-11 July 2001, Salt Lake City, Utah

First published as:
AIAA 2001-3962

37th AIAA/ASME/SAE/ASEE Joint Propulsion Conference,
8-11 July 2001, Salt Lake City, Utah