This paper describes the vision for Space Exploration that would return humans to the moon by 2020. Creating architecture for returning humans to the moon requires the comprehension of the physics of spaceflight, knowledge of the hardware that can realize the physics, and an understanding of how these many parts interact and interconnect. The NASA team concluded early in its study that the direct–direct mode would be possible only if a single launch vehicle approaching twice the lift capacity of the Saturn V were available. The three mission modes were compared as higher levels of technology were engaged. The key was to find a workable architecture that involved the least amount of mass. The direct return mission that involved no operations in lunar orbit seems to be the least operationally complex, but it tends to be the least efficient because it moves the largest mass-including the Earth-entry heat shield- the entire velocity change of lunar landing and ascent.


The Apollo project ended a long time ago. About half of all Americans were not born when Gene Cernan closed the hatch on lunar module Challenger and the astronauts of Apollo 17 left the moon. And fewer than one in six were adults when, 45 years ago this month, John F. Kennedy challenged the nation to land a man on the moon.

So when President George W. Bush visited NASA headquarters in January 2004 to announce a new Vision for Space Exploration that would return humans to the moon by 2020 in preparation for future exploration of Mars, it had a familiar ring, but one more from the history books than from personal experience.

It was also addressed to a more mature space agency in a more mature space age. More than three decades and many billions of miles had been accumulated in space since our fleeting first steps on the lunar soil. NASA has launched robotic spacecraft bound for all corners of our solar system, and operates a human spaceflight program that included the largest-ever space station assembled in Earth orbit and the world’s only reusable spacecraft. But a space agency that was a product of a Cold War-era race to explore space had been slowly transitioning into the space operations business. Now it had been given a new set of challenges that would again exercise its exploration muscles.

First, though, some tasks had to be finished. NASA would set its sights on completing the International Space Station to honor its international commitments. Along the way, it would retire the Space Shuttle fleet by 2010 and build and fly its replacement, the new Crew Exploration Vehicle, by 2014.

John F. Connolly is an engineer at the NASA Johnson Space Center in Houston, the pre-project manager of the new lunar lander program, and a registered engineer in Texas.

In March 2005, NASA Administrator Mike Griffin challenged a handpicked group of engineers and managers to design a mission architecture that would accomplish the next phases of the vision—returning humans to the moon and beginning the push to Mars. “Architecture” is NASA-speak for the combination of spacecraft, launch vehicles, orbital mechanics, and operations that assemble to accomplish a space mission.


Many of the team members who took on this task are children of the Apollo generation, the kids who gazed wide-eyed into our televisions as grainy images of Armstrong, Aldrin, and the rest loped across the lunar surface. We—and I count myself as a child of Apollo—have the Space Shuttle and space station to count as our engineering triumphs, but what NASA accomplished in the late 1960s and early 1970s, was pure engineering art.

There have been many advances in aeronautics in the past 30 years, and the NASA architecture team members studied the volumes of reports published since Apollo, detailing the many innovative ways to conduct the human return to the moon. The team also became students of history, studying the decisions that led to the Apollo mission architecture and consulting with Apollo-era astronauts, managers, and flight directors. Drawing on that accumulated wisdom would be essential.

Creating an architecture for returning humans to the moon requires the comprehension of the physics of spaceflight, a knowledge of the hardware that can realize the physics, and an understanding of how these many parts interact and interconnect.

The physics is both straightforward and inflexible. “Rocket science” is the art of managing velocity changes that are dictated by physics. Leaving Earth orbit on a three-day translunar traverse requires a velocity increase of 3,100 meters per second; capturing into a preferred lunar orbit requires a velocity decrease of 1,100 m/s; and descent to the lunar surface requires a further decrease of 1,900 m/s. Returning to Earth requires the same basic velocity changes again—in reverse order and sign.

Then the creative component of mission design enters the fray. With the help of the many past lunar mission studies, the NASA team was able to reduce the mission design problem to two fundamental questions: Do you perform any dockings or undockings in Earth orbit? And do you do any dockings or undockings in lunar orbit?

The answers—yes or no—to these two questions can be thought of as a two-by-two matrix describing four fundamentally different mission architecures. A mission that required Earth orbit rendezvous as well as rendezvous in lunar orbit is termed “EOR-LOR.” A mission that injected directly to the moon (bypassing Earth orbital operations) and returned directly from the surface of the moon (bypassing lunar-orbital operations) was termed “direct-direct.”

One can also have EOR-direct return and direct-LOR architectures. The EOR-direct return mission was the mode favored by Wernher von Braun early in the Apollo program, while direct-LOR (or simply LOR) was the mode eventually chosen for Apollo missions.

The NASA team concluded early in its study that the direct-direct mode would be possible only if a single launch vehicle approaching twice the lift capacity of the Saturn V were available. Because no launchers of this size were considered feasible, direct-direct was eliminated as a mission mode. The three remaining mission modes—LOR, EOR-LOR, and EOR-direct return—were then studied in detail to compare their performance, cost, and risks.

Each architecture has a different degree of complexity. A lunar orbit rendezvous mission is characterized by no operations in Earth orbit, and the lander and Earth-moon crew transit elements first meet in lunar orbit. After the crew transfers from its transit vehicle, they descend to the surface in the lander, leaving the Crew Exploration Vehicle in lunar orbit to await their return. Following the surface mission, the crew ascends from the surface to rendezvous with the CEV in lunar orbit, jettisons the lander, and returns to Earth.

In a dual-rendezvous (EOR-LOR) scenario, the CEV and lander first dock in low Earth orbit and transit together to lunar orbit where the mission proceeds in much the same way as the LOR scenario. The difference between the two is the assignment of propulsive maneuvers to different flight elements. In the LOR case, the upper stage of the launch vehicle, or Earth Departure Stage (EDS), performs the velocity change necessary to depart Earth as well as that necessary to capture into lunar orbit. For a dual rendezvous mission, the EDS performs only the Earth departure burn, and capture into lunar orbit is assigned to the lunar lander.

Then there’s the EOR-direct return architecture, where a combination of exploration vehicle and lander is assembled in Earth orbit. This vehicle then travels to the moon, lands on the lunar surface, then returns to Earth—all without any operations in lunar orbit. Operationally, this is the simplest, as it requires no transfer of crew from one vehicle to another.

Space mission analysis often begins by calculating the mass of the various mission elements. Spaceflight is a business where you literally buy things by the pound, so mass is a good first-order estimate of cost. For lunar landers, space capsules, launch vehicles, or rovers, less mass is almost always better. Once the spacecraft mass is estimated, fuel mass can be determined based upon the velocity changes required.


“Since early in my term, our efforts in space have been under review.... [W]e have examined where we are strong and where we are not, where we may succeed and where we may not. Now it is time to take longer strides—time for a great new American enterprise—time for this nation to take a clearly leading role in space achievement, which in many ways may hold the key to our future on Earth ....

First, I believe that this nation should commit itself to achieving the goal, before this decade is out, of landing a man on the moon and returning him safely to the Earth. No single space project in this period will be more impressive to mankind, or more important for the long-range exploration of space; and none will be so difficult or expensive to accomplish. We propose to accelerate the development of the appropriate lunar spacecraft. We propose to develop alternate liquid and solid fuel boosters, much larger than any now being developed, until certain which is superior. We propose additional funds for other engine development and for unmanned explorations—explorations which are particularly important for one purpose which this nation will never overlook; the survival of the man who first makes this daring flight. But in a very real sense, it will not be one man going to the moon—if we make this judgment affirmatively, it will be an entire nation. For all of us must work to put him there ....

I believe we should go to the moon. But... it is a heavy burden, and there is no sense in agreeing or desiring that the United States take an affirmative position in outer space, unless we are prepared to do the work and bear the burdens to make it successful. If we are not, we should decide today and this year.”


“Inspired by all that has come before, and guided by clear objectives, today we set a new course for America’s space program. We will give NASA a new focus and vision for future exploration. We will build new ships to carry man forward into the universe, to gain a new foothold on the moon, and to prepare for new journeys to worlds beyond our own ....

Our ... goal is to return to the moon by 2020, as the launching point for missions beyond. Beginning no later than 2008, we will send a series of robotic missions to the lunar surface to research and prepare for future human exploration. ... [W]e will undertake extended human missions to the moon as early as 2015, with the goal of living and working there for increasingly extended periods ....

Returning to the moon is an important step for our space program. Establishing an extended human presence on the moon could vastly reduce the costs of further space exploration, making possible ever more ambitious missions. Lifting heavy spacecraft and fuel out of the Earth's gravity is expensive. Spacecraft assembled and provisioned on the moon could escape its far lower gravity using far less energy, and thus, far less cost. Also, the moon is home to abundant resources. Its soil contains raw materials that might be harvested and processed into rocket fuel or breathable air. We can use our time on the moon to develop and test new approaches and technologies and systems that will allow us to function in other, more challenging environments. The moon is a logical step toward further progress and achievement ....

We may discover resources on the moon or Mars that will boggle the imagination, that will test our limits to dream. And the fascination generated by further exploration will inspire our young people to study math, and science, and engineering, and create a new generation of innovators and pioneers.”

The analysis gives engineers two tools to work with in order to reduce mass: Technologies can be changed—for example, using composite materials rather than heavier metal structure—and flight elements can be reordered, so that the minimum amount of mass is moved through the largest velocity changes.

The NASA team approached the problem by first varying propulsion technology. Specifically, different propellant combinations have different specific impulses, ISp, the rocket equivalent of gas mileage. In the case of rocket propulsion, however, this term lies in the exponent of the all-hallowed rocket equation, so a doubling of Isp yields a large downward change in the mass of propellant.

The team compared poorer-performing hypergolic propellants (chemical combinations that spontaneously combust when mixed) with higher Isp cryogenic propellants. Seeking to increase performance even further, they examined various combinations of liquid oxygen and hydrogen, methane, and ethanol. In the end, high ISp propulsion won out wherever large velocity changes were needed, and slightly poorer-performing propulsion was recommended whenever long-duration storage would be complicated by super-cold liquid hydrogen.

Although propulsion technology often has the biggest effect on spacecraft mass, the NASA team investigated other subsystems where mass could be shaved. One example was the nearly two tons of polyethylene radiation shielding that was part of the original CEV design.

That hydrogen and carbon-dense shielding system was added to protect astronauts against a worst-case solar flare. But moving that large parasitic mass through the compounded multiplication of velocity changes exacted a heavy toll on the mass of the vehicle. A combination of risk analysis and structural redesign resulted in both lower radiation shielding requirements and a new, composite-rich structure that would provide the same high level of shielding without the need for additional shield mass.

The three mission modes were compared as higher levels of technology were engaged. The key was to find a workable architecture that involved the least amount of mass. Technology and clever design could not make every mission mode competitive, however. The direct return mission, which involved no operations in lunar orbit, seems to be the least operationally complex, but it tends to be the least efficient because it moves the largest mass—including the Earth-entry heat shield—the entire velocity change of lunar landing and ascent. In this case, operational simplicity comes at the cost of vehicle complexity and additional mass. In spite of its elegance, that option had to be abandoned.

The final step in the selection process was to compare the relative cost and risks of the two remaining lunar mission architectures. Both the LOR Apollo-style mission mode and EOR-LOR dual rendezvous require approximately the same number of new developments, launches, and mission operations, so the life-cycle cost of the two were approximately equal. The team then looked at risk.

Risk was measured both as the probability of loss of mission, and the probability of loss of crew. On these measures, the dual-rendezvous mission was superior.


The resulting architecture chosen by NASA to return to the moon is a dual-rendezvous mission that uses a heavy-lift launch vehicle to lift a lunar lander and Earth departure stage into Earth orbit. That launch is followed by a small launch vehicle to lift the CEV and crew into Earth orbit, where the exploration vehicle and EDS with lander rendezvous. The EDS then performs a trans-lunar insertion burn, and is expended. The remaining CEV-lander stack performs a four-day cruise to the moon, and upon arrival is inserted into lunar orbit by the descent stage of the lunar lander. The crew transfers from the CEV to the lander and descends to the lunar surface, leaving the CEV in lunar orbit. The crew can spend up to a week anywhere on the surface, performing science investigations, resource utilization experiments, and technology demonstrations that begin the preparation for a Mars mission.

The crew returns to lunar orbit in the ascent stage of the lander, redocks with the CEV, and expends the ascent stage. The service module of the CEV then performs a trans-Earth injection burn that puts the CEV and crew in a four-day Earthbound trajectory that ends in a landing on the West Coast. Splashdowns and the naval fleet required to support them have been outmoded.

Those NASA folks working on this new mission architecture—me and my fellow children of the Apollo generation—sought to build upon, but distinguish our design from the familiar icons of NASA’s past successes. We grew up with Star Trek, Star Wars, and model rockets, and while Apollo was our parents’ space program, the next lunar mission moon will be ours.

We naturally thought that our moon mission ought to be in something like the Millennium Falcon. But a funny thing happened on the way back to the moon: Physics and technology intervened. Physics, or at least our current understanding of it, dictates the velocity changes needed to travel through space, and puts Sir Issac Newton in the designer’s chair.

Technology has improved some since Apollo, so our missions will be more capable, but we still haven’t developed hyperdrive. With physics the same and technology improved only incrementally, it’s no surprise that the solution looks a lot like Apollo. It isn’t due to lack of imagination on the part of this generation of engineers, but rather to the fact that the Apollo engineers understood the problems as well as we do today.

Although they were inventing a new field of science and engineering using only slide rules, our parents—the guys in the skinny black ties—got it exactly right. It would be nice for our kids to one day say the same about us.