Flybys, Launch Windows, and Selfies with the Earth and Moon:

The Artemis I Flight from the Perspective of a Member of the Trajectory-Design Team

By Robert E. Harpold

Artemis I was the first full test flight of the new Orion spacecraft, which was designed to return people to the Moon. If certain aspects of this flight had failed, we would have had to repeat it before the spacecraft could be certified to carry people. Our main objective was to test the heatshield at speeds it would experience from a lunar return, but we also wanted to test every system aboard the spacecraft and recover the vehicle. If you followed the news about the flight, you know it achieved every major objective and several objectives that weren’t proposed until mid-flight.

Artemis I, and the Artemis program in general, was possible because of many teams consisting of many individuals who dedicated themselves to the mission. Each team focused on its own subsystem, and the individuals within those teams further focused on detailed aspects of that subsystem. It took thousands of people, spread across the country and in other countries, to make Artemis I a success.

I have little knowledge about other teams because I was siloed within my own specialty within my own team, but hopefully describing our work will give you an idea of the work and enthusiasm of every team who supported Artemis I. I also hope it gives you at least a hint of what it was like to be part of this project and a sense of pride at what humans can accomplish when we push ourselves to explore.

Figure 1. Artemis I after the Return Powered Flyby burn. Photograph courtesy of NASA.

Trajectory Design

Our team designed the trajectories, the path the spacecraft would take from the Earth to the Moon and back again. In order to prove the Orion spacecraft could carry humans, the trajectory had to satisfy several mission objectives, including testing the heat shield at lunar-return velocities, demonstrating that Orion’s propulsion, navigation, communications, and other systems could operate in the space environment, and returning the vehicle to a designated landing site.

Jacob Williams designed the original trajectory over a decade ago. Orion’s destination was a Distant Retrograde Orbit (DRO) around the Moon, an interesting orbit where the spacecraft is actually in Earth orbit, but, from the perspective of an observer on the Moon, it would look like the spacecraft was orbiting the Moon backward.

In the nominal mission plan, five major burns would be performed (Figure 2). Ninety minutes after launch, the Interim Cryogenic Propulsion Stage (ICPS) would perform the Trans-Lunar Injection (TLI) burn to send Orion to the Moon. During the flyby of the Moon, Orion would perform the Outbound Powered Flyby (OPF) burn which, with the assistance of the Moon’s gravity, would send Orion on a path toward the DRO. It would then perform the Distant Retrograde orbit Insertion (DRI) burn to enter the DRO. Six days later, it would perform the Distant Retrograde orbit Departure (DRD) burn to head back to the Moon, and then perform the Return Powered Flyby (RPF) burn to use the Moon’s gravity to return to Earth.

Tim Dawn updated the original trajectory to account for additional mission constraints. One of those constraints was making certain Orion’s landing occurred during daylight, which would help recovery operations. To achieve this objective, the mission duration could be between 26 – 28 days or 38 – 42 days. Max Widner wrote software that would select the mission duration that achieved all mission objectives and minimized fuel usage.

But, since the Earth-Moon geometry changes, the trajectory would change due to the launch date, and differences in launch azimuth would cause changes throughout the two-hour launch window. Some trajectories would violate the constraint for length of time in eclipse (when either the Earth or Moon are blocking the sunlight on the spacecraft), which would deplete the battery-power reserves and thermally stress the spacecraft. If those eclipses weren’t eliminated or reduced below the limit, Orion couldn’t fly those trajectories, which would result in the loss of available launch dates. Sarah Smallwood developed an eclipse-mitigation algorithm to recover those launch dates by adding correction burns or inclining the DRO.

Before flight, we also needed to know the available contingency options. One set of possibilities was the spacecraft missing a major burn, delaying the burn, or the engine only performing part of the burn. Brian Killeen developed software to determine a correction burn that would return the spacecraft to the nominal trajectory in those situations.

We also needed to protect for the possibility of the Space Launch System (SLS) rocket or the ICPS engine not performing their burns for the full durations, which would result in Orion being placed on the wrong trajectory. Colin Brown and Tim developed several types of alternate missions, some flying by the Moon and others remaining in Earth orbit, that would still achieve some or all of the main mission objectives.

We also needed to be ready to bring the spacecraft home early in case of mission-ending spacecraft failures. I developed software to find abort trajectories for each phase of the mission for different types of failures. For each combination of launch date and time, we generated hundreds of abort trajectories. Within each month, we would typically have between 10 and 15 available days to launch, and we would generate abort trajectories for five epochs each day, so we generated thousands of abort trajectories for each month. This amount of data became unwieldy, so Colin wrote a database to store them and provide efficient access to them.

Figure 2. Artemis I Trajectory. TLI = TransLunar Injection, OTC = Outbound Trajectory Correction, OPF = Outbound Powered Flyby, DRO = Distant Retrograde Orbit, DRI = Distant Retrograde orbit Insertion, DRD = Distant Retrograde orbit Departure, RPF = Return Powered Flyby, RTC = Return Trajectory Correction, EI = Entry Interface. Figure courtesy of Batcha, Amelia L., et al. "Artemis I Trajectory Design and Optimization." 2020 AAS/AIAA Astrodynamics Specialist Conference. No. AAS 20-649. 2020. Diagram courtesy of NASA.

In addition, Jeff Gutkowski was our team lead, Randy Eckman wrote trajectory-design tools, Amelia Batcha was our main analyst studying the feasibility of each trajectory, Badejo Adebonojo ran the missed-burn cases, Elizabeth Williams wrote software to plot different data for each trajectory, Josh Geiser joined our team shortly before launch, and Matt Horstman and Jacob generated off-nominal trajectories in the office after launch. Most of us worked on future Artemis missions at the same time, but I am only listing the tasks we performed specifically for Artemis I.

Our team had fun with the tool names. There was Damocles, an ominous name that raised some eyebrows. That name inspired Elizabeth to call her plotting tool Plotacles. We also had a trajectory vending machine, RoboCopPy, and RATGATOR. Similarly, one of the teams in Mission Control named all their tools after Batman characters.

Every month until we launched, our team cranked out all the necessary trajectories (nominal and off-nominal) and presented them to management and our flight controller counterparts. This work was in addition to our development of new contingency options we kept realizing we needed, which continued up to the day we launched.

Because of our trajectory work, the flight controllers in Mission Control needed our team to be on console in the Mission Evaluation Room.

MER and TARGO

Most people have heard of Mission Control in Houston. It’s the room where the Flight Director and the flight controllers, each in charge of their own specialty, help operate the spacecraft and troubleshoot any anomalies during flight. Each flight controller is supported by a counterpart in a back room (called the Multi-Purpose Support Room or MPSR, pronounced ‘mip-ser’) with whom they are in constant contact. Where a flight controller is in charge of integrating their system with the entire team, the people in the MPSR focus on work related to their specific subsystem.

There is also a back backroom called the MER (Mission Evaluation Room, pronounced ‘mer’) for the people who designed, built, and/or tested the subsystems. They monitor their systems and help the flight controllers troubleshoot anomalies. For an Apollo 13 analogy, there was a time where NASA engineers had to figure out how to fit the command module’s square carbon-dioxide-filter canister into the round slot for the lunar module’s canister. That problem is an extreme example of the kind of work the MER would be called upon to do. There were twenty consoles in the MER (Figure 3), and at any given time, there were about thirty people in the room.

Our team’s console in the MER was called TARGO (Trajectory Analysis, RetarGeting, and Optimization). We primarily supported the Flight Dynamics Officer (FDO, pronounced like the dog name), the Mission Control console in charge of the spacecraft trajectory. The FDOs have been treated with high respect since the early days of human spaceflight, and, from my experience interacting with them, they deserve every bit of that respect. During flight, we provided them with updated nominal trajectories and off-nominal options for each phase of the mission, as well as performing additional analysis when requested.

For another Apollo 13 analogy, in the movie, there is a scene where Gene Kranz points to a chalkboard and draws the two options available to them: turn around right away or loop around the Moon and take longer to return. If that flight happened today, TARGOs would have been the ones to have told the team about those options before the flight.

Brian and Randy led our training. They assigned each of us to prepare lessons on our specialties and even gave us a written test, just like in our school days (it was an open-book, take-home test). The fun part of training, though, was the simulations.

In the simulations, our testers would create scenarios with different systems failing so we could train to fix the problems or find a way to perform the mission despite the problems. The simulations gave us valuable experience in interacting with our flight controller counterparts and other MER teammates and learning things like speaking on the appropriate loops and not talking when the Flight Director is talking. We even had three simulations that lasted 36 hours each which let us practice passing files between shifts and having team members pick up a problem from where one shift left off and then transferring that work to the next shift. It reminded me of what I’d read about training during the Apollo era.

Figure 3. Ops Suite 3, the room where the MER team did its work. Photograph courtesy of NASA.

The Flight

After so many delays, it felt surrealistic to watch Artemis I launch on November 16. I woke up my wife, Julie, and 13-month-old daughter, Isabelle, to watch the SLS rocket take off at 00:47:44 Central Time. Tim and Brian were on console for our team, Randy and Sarah were getting ready to take over from them two hours after launch, and Colin and I were going to take the third shift several hours later. I should have been asleep, but I didn’t want to miss the historic launch, and I also wanted to find out what kind of day I was going to have. The SLS performed perfectly, and the TLI was perfect, so I was able to go to sleep with a positive feeling.

We supported the flight 24/7, with day, swing, and night shifts. Our team rotated through shifts, so no one was on any one shift the whole time (although Max had a lot of night shifts). Typically, two TARGOs were on console at a time, a prime and a second to help with the workload. We also rotated being prime and second. It was a rough schedule, particularly the first two-thirds of the mission where we not only supported on console, but spent time off-console cranking out off-nominal options for each mission phase based on the most up-to-date trajectory information.

So it was exhausting and a lot of work, but also fun and exciting. I loved getting to be in such a great seat to see this historic mission taking place and to be among so many hardworking and passionate experts who had dedicated so much of themselves to sending humans beyond low Earth orbit again. Even when I was sleep-deprived and waking up at odd hours, I was happy being in a place and time where I could do what I had dreamed about doing since I was nine years old. I’m sure many of the people there felt the same.

Figure 4. Amelia Batcha, left, and Tim Dawn, center, in the process of handing over to Robert Harpold, right, when Orion was at its farthest distance from Earth. Photograph courtesy of NASA.

Throughout most of the mission, we talked over off-nominal options with the FDOs, did troubleshooting with the trajectories they were building, and performed additional analysis, like finding out if we could change our landing location after we had already left the Moon or finding our delay capability if we missed the Return Powered Flyby. Other teams in the MER asked us questions related to their own subsystems, some of which required analysis on our part. We also updated our console log so other TARGOs would know what had happened on previous shifts, and we had an hour at the beginning and end of each shift to perform the handover from the current team to the oncoming team (Figure 4).

My personal highlight was being the prime TARGO on the shift where we performed the Outbound Powered Flyby. This burn took place near perilune, the closest approach to the Moon on the outbound leg of the trajectory, and would send us on a path so we could insert ourselves into the DRO. It was an exciting moment, because it was the point where we were returning a human-rated spacecraft to the Moon after 50 years.

The displays in the MER showed live video from the spacecraft, so when Max and I arrived for our shift that night, the Moon was starting to fill a substantial portion of the screen. Sarah, from whom we were taking over, gestured to the screen and said, “Welcome to the Moon.”

Throughout the shift, the Moon kept getting larger and larger on the screen. Amelia joined us later to help with some of the work. The FDO and his backroom counterpart were setting up the spacecraft commands to perform the burn. Since the burn would be performed behind the Moon, we would lose contact for twenty minutes and wouldn’t know if the burn was successful until the spacecraft reappeared from the other side. If the burn failed, we would regain contact with the spacecraft two minutes earlier. Both countdowns were displayed on the MER screens.

Up to the moment Orion went behind the Moon, everyone was checking their systems. The MER Manager announced we would have a go/no-go poll in a few minutes. It took me a second to realize what that meant, and I said it out loud to Max and Amelia: “Oh, wow, I get to be the one who says, ‘Go!’”

When it came time for the poll, the MER Manager called, “TARGO?” I said, “Go!” maybe a little more forcefully than necessary. A few seconds later, the MER Manager repeated the question, and I realized I had spoken on the wrong loop. Even after all the training. I switched to the correct loop and said it again, with most of the enthusiasm of the first time.

On the screens, we could see Earth as a small blue ball. As the spacecraft got closer to passing behind the Moon, Earth moved closer and closer to the horizon. The final image before we lost contact showed the marble-sized Earth just above the Moon’s horizon.

Once the spacecraft was out of contact, there was nothing we could do. People stood and started talking, and the atmosphere became festive. We’d gotten to the Moon! And Orion had taken some cool selfies with the Earth and Moon! It was also a good time for restroom breaks. The MER Manager quieted us down two minutes before the first countdown, the one we didn’t want to see, where the spacecraft would arrive early if it missed the burn.

That time came and went, and everyone felt relieved. Then, right on time, on the screens, we saw the Earthrise. Since the side of the Moon closest to Orion was unlit at that point, the view looked like Earth floating by itself in darkness. It was proof that OPF had been performed correctly, and that Orion was headed to the correct orbit. Everyone in the room clapped.

We remained busy for several days, performing three other major burns. After the RPF, Orion was locked on course toward its landing site on Earth, so there wasn’t much the TARGOs could do. We kept supporting up to the shift before splashdown, doing occasional analysis and answering questions from other MER people.

I watched the splashdown from home. It was great seeing everything happen just as planned: the command and service modules separating, the spacecraft entering the atmosphere, the parachutes deploying, and the spacecraft landing in the water. After so many years, we had sent a human-rated spacecraft to the Moon, and it had performed extraordinarily well.

Next time, we’re sending people there.