Landing as a System

Five major subsystems defined the landing. First, the LM descent engine (officially known as the ''descent propulsion system'' or DPS) stuck out the bottom of the spacecraft in the characteristic rocket bell shape. Like the enormous Saturn engines, the DPS was on gimbals, and could be angled fore and aft and side to side—useful for ''trimming'' the thrust, to be sure it went through the LM's exact center of gravity, which changed as it consumed fuel. The engine was unusual in another respect as well: any engine could get a vehicle through a spot on the lunar surface;but the LM engine had to get the vehicle to that point and make it stop gently, even as the LM's mass changed from second to second as fuel was burned.13 This requirement meant the descent engine had to be ''throttleable:'' its thrust level must be controlled in real-time, by the computer.

For a control engineer, such flexibility seemed ideal, but, as always, there was a catch. The descent engine could only throttle from zero to about 55 percent;above that value, it had to run at 100 percent of its rated thrust, otherwise the fiery exhaust would erode the bell. This limitation added complexity to the control equations and, as a departure from the ideal trajectory, would cost additional fuel (380 lbs., to be exact).

The second major subsystem, the landing radar, bounced radio waves off the lunar surface to detect both the altitude and velocity of the LM. The radar projected four beams. Three of them detected the Doppler shift in the reflections and compared them to determine the spacecraft's lateral velocities. A fourth beam measured the travel time for the reflection to return, which indicated altitude. The antennas of the radar could be switched into two positions: one angled outward so it would point straight down when the LM was horizontal, and another pointed downward so it would aim correctly when the LM was vertical. Ryan Aeronautical of San Diego, the same company that had built Lindbergh's plane, The Spirit of St. Louis, built this critical device.

Problems beset the landing radar from the beginning. Only on Apollo 10 did a realistic flight test provide sufficient confidence in the landing radar, and on Apollo 11 the crew still looked down by eye to verify its readings. The device would nearly cause an abort on Apollo 14.

Communications formed the third major subsystem for landing. The LM had two types of antennas to communicate back to earth. The parabolic ''high gain'' antenna transmitted high-quality voice and data, but it needed to point directly at the earth to work. During landing, the computer could figure out the proper direction and point the antenna automatically, or the lunar module pilot could take over and ''slew'' the antenna with a knob. Two smaller, omnidirectional antennas had no such pointing requirements other than being on the side of the vehicle facing the earth, but they generated lower data rates and voice quality. Getting the communications properly lined up and free of noise proved a burden on the crew at a critical time. The LM also had a rendezvous radar for pointing at the CSM to provide range and closing rate data when the two craft were reuniting, but the device was not used during landing. Its antenna was mounted on servos so the computer could control its pointing and keeping it focused on the CSM during the maneuver. The rendezvous radar stayed on during the landing in case it was needed for an emergency abort;its computer connection would cause a startling problem on Apollo 11.

Integrating the system in real-time was the computer, the fourth critical element in the landing. The LGC (for LM guidance computer identical to the AGC in the CM) had to talk to all of the different components, each made by different companies with different specifications. It ran the digital autopilot, directed the small thrusters that controlled attitude (the RCS, for reaction control system), commanded the descent engine, controlled indicators on the pilots' panels, took data from the DSKY keyboard, and read the commander's control stick inputs. The software also measured time intervals, counted pulses, and performed a variety of other tasks. And, of course, it gathered data from the inertial platform. Together, the computer and the inertial unit comprised the primary navigation and guidance system, or PNGS, pronounced ''pings.'' A separate, backup guidance system, the ''AGS'' for ''abort guidance system,'' could come into play to get the LM back to a rendezvous if the PNGS failed.

Software had to account for the vagaries in behavior of all of these systems. The descent engine, for example, ''ablated'' or wore out as it burned, changing its thrust characteristics. The software thus had to incorporate the changes into its equations (a problem in this correction nearly caused an unstable throttle on the early landings).

The final components of the landing system, of course, were the humans—at once operators and cargo. A strange terminology supported their sense of control. The ''commander,'' in the left position, actually flew the LM. The other operator, in the right position, was technically a co-pilot, but he was titled ''lunar module pilot'' (LMP) even though he never touched the hand controls. He served as a systems engineer and operated the DSKY. During descent, the commander verified the computer programs and landing phases, redesignated the landing sites, ''flew'' the final few hundred feet, and stopped the engine on lunar contact. The LMP kept his eyes on the numbers, verifying the agreement between the two navigation systems, AGS and PNGS, punched keys to issue commands to the computer, and, in the final moments, called out altitude, velocity, and fuel quantities so the commander could keep his eyes focused outside the window. Not only did the operators interface with the computer through a variety of displays and controls, they also interfaced with each other in a social relationship. Furthermore, they communicated with the ground controllers, who cleared them for the next steps and aided in managing the systems.

When Bennett and his group began thinking about how the astronauts in the LM should approach the landing site, they naturally modeled the procedure on aircraft approaches. To the pilots, the moon seemed like an unfamiliar airport, so they wanted to fly around it and get a visual read before going in for a landing (wasting precious fuel in the process).14 Inevitably they ran into the same issue that had proved so contentious with the launches: would the landing be fully automatic or manual? More realistically, how much of the landing would be automated? How would the human and machine trade control?

Automatic landing offered a variety of options. Wernher von Braun's original vision for lunar landing had no human involvement, ''accomplished entirely by the automatic pilot running on a guidance tape,'' as he put it, although it did allow the ''captain'' to avoid obstacles.15 Grumman engineers had been studying a fully automatic LM mission, under a NASA contract. Selecting the landing site could be accomplished by telescope while the crew was still in the CSM, or by a real-time TV image transmitted from the surface to the LM.16

Bennett and his group, working together with the astronauts on ''pilot-in-the-loop'' simulators, developed the techniques for lunar landing. Mostly, the pilots' roles were limited to systems monitoring and abort situations, except in the final moments before touchdown. In certain aborts, especially with a failure in the PNGS, the crew could become overwhelmingly busy. To Bennett it was a given that the pilots would land under manual (semiautomatic) control, even though the computer was capable of landing automatically. ''None of the crew wanted to land in an automatic system,'' Bennett recalled.

Bennett once suggested to Chris Kraft a fully unmanned, automated lunar landing as a test. He recalled that Kraft would not allow it, because if an initial, automatic landing failed, then Kraft felt Congress would insist on a successful demonstration before attempting it with astronauts aboard.17 Bennett wasn't trying to undermine the idea of the lunar landing, noting ''I was just being the purist engineer,'' trying to ensure a safe landing. With that in mind, Bennett got into the lunar mission simulator one day and instructed it to land in a fully automatic mode. He remembers the astronauts' reaction: ''That's not flying,'' they told him with contempt.18

While the LM and CSM were still docked, the astronauts would enter the LM, power it up, and run through a series of checkouts. No electrical connection existed between the guidance computer in the LM and that in the CSM, so the crew had to transfer data, time synchronization, and orbital parameters by calling the numbers out from the AGC in the command module, punching them into the LGC by hand and calling ''mark'' to synchronize them. They would similarly initialize, calibrate, and align the PNGS. Then the two craft would separate. The LM commander would rotate the craft around so the command module pilot, now alone in the CSM, could visually inspect the integrity of the LM. If all seemed well, the LM gradually backed away and headed for a lower orbit, about ten miles above the moon.

In June 1966 NASA convened a symposium in Houston to work through the operational plans for a full lunar landing mission. MSC engineer Owen Maynard, the symposium's organizer, was one of a handful of core NASA engineers who had emigrated from the Avro Canada company after their fighter jet program was canceled. Maynard designated a series of different mission types, from the ''A'' missions of unmanned tests to ''G,'' the first lunar landing;''H,'' further basic landings;and ''J,'' enhanced landings with heavier payloads and longer stays. Maynard also laid out a strategy of nine ''plateaus'' for each lunar mission, relatively safe positions along the flight allowing careful assessments and decisions before proceeding (or aborting).19

Strange as it seems, for astronauts in a little metal balloon located a quarter million miles from earth, lunar orbit was a relatively safe position, and therefore one of May-nard's plateaus. From here everything could be put on hold, safety checked, and replanned. If a problem arose, ground control could send the LM around the moon for another orbit or two while they troubleshot the problem. The astronauts could return home, if necessary, with relative safety. The CSM could even swoop down and rescue the LM from its low orbit.

Beginning the landing sequence was a major commit point. It started at 50,000 feet, when the LM's descent engine fired to slow the craft, which would cause it to descend toward the lunar surface, leaving the plateau of lunar orbit. Once the engine fired the LM either had to land, hit the moon (in about ten minutes), or execute a dangerous abort. The clock was ticking.

Telescopes Mastery

Telescopes Mastery

Through this ebook, you are going to learn what you will need to know all about the telescopes that can provide a fun and rewarding hobby for you and your family!

Get My Free Ebook

Post a comment