A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
From: Frank Reed
Date: 2010 Oct 23, 20:13 -0700
In a recent message, Gary Lapook noticed a reference to a sextant supposedly on John Glenn's first orbital flight in the video about Gago Coutinho. I suspect that there was some confusion in the original quotation, maybe mixing two separate comments made by Glenn.
John Glenn, as I understand it, made a couple of public appearances in Portugal in the mid-1960s. These were public relations events. Glenn was famous as the first American to orbit the Earth, and he was also a politician (even before becoming one professionally). He probably said something about his "first orbital flight" and he may have said something about "sextants in orbit" and being "inspired by Coutinho's contribution to aerial navigation" which was then slightly misquoted to imply that there was a sextant on his first orbital flight which was directly related to Coutinho's sextant. But there was no sextant on any of the Mercury flights. There was no need for one, and the capsule (or "spacecraft" depending on who was spinning things) had no spare room for one. The Mercury capsule had simple thermal "horizon finders" which were used to orient it correctly for retro fire. These had no significant angular measurement capability though they could find the "vertical" perpendicular to the Earth's surface with reasonable accuracy. On later Mercury flights, there was also a "navigation reticle" which could be used to make crude measurements of angles, but this was little more than some lines and arcs on a transparency. Angles could be eyeballed to a few tenths of a degree with some care.
So what about other sextants in space? A follow-up to Gary's post suggested that the sextant referred to by Glenn must be some super sextant, the epitome of manual navigation. The facts are a long way from here...
The Mercury spacecraft had no orbital maneuvering capability except the ability to change its orientation and, of course, fire its retro-rockets to get out of orbit. For the astronauts, it was not quite "flying". This changed with Gemini which was a second-generation spacecraft with significant maneuverability but too early for significant automation. It had a very primitive computer which could calculate burn times but it did little more and it did not fire the rockets directly. The astronauts literally flew the Gemini spacecraft, and naturally they liked that very much. On Gemini missions a few experiments were conducted to see if astronauts could navigate autonomously using sextants. These experiments were mostly failures, or, in the parlance of the NASA bureaucracy, at best "satisfactory" but they did not lead to future sextant use with any navigational significance. Paul Hirose wrote up an account of the Gemini 10 experiments just a few months ago: http://fer3.com/x.aspx/13322.
There was one exception to the rule during the Gemini program. Usually their sextant experiments were of no real value, but on Gemini 12, Buzz Aldrin, known to the other astronauts as "Doctor Rendezvous" for his expertise and obsession with the science of rendezvous, was able to save the mission using a sextant when the spacecraft radar failed. They were supposed to rendezvous and dock with an Agena target vehicle, but without range and rate data from the radar, this appeared to be impossible. Aldrin knew better and measured the angular position of the Agena relative to bright stars to solve the problem in a different way. It should probably be noted here that this was not so much a celestial navigation problem as a piloting problem or even a "firing solution" problem --like using a traditional sextant to measure the angular position and velocity of another vessel so that we can meet up with it when we need to.
The principle difficulty with celestial navigation (position-finding) in orbit around the Earth is essentially the same as using a marine sextant in an airplane: the horizon is not visible. At aircraft altitudes, the horizon is lost in the haze and extinction from many miles of thick atmosphere. In low Earth orbit, the problem is similar though beccause of the greater range to the horizon, the angular width of uncertainty is smaller. Here the bigger problem is that there are frequently false horizons created by layers in the high atmosphere. For spacecraft that were well above low orbit, it was believed that this problem could be solved by setting some arbitrary height, e.g. 20 miles, above the Earth's surface as the altitude of the "apparent horizon". Unfortunately, different astronauts saw this apparent horizon in different places by some minutes of arc. Note that this is NOT "irradiation". It's just individual personal judgement on where to "call" that apparent horizon. In the later Apollo missions, this was left as an adjustable parameter to take into account the actual experience of the astronauts in flight.
Incidentally, there are brief reports that Soviet cosmonauts also experimented with sextants, in particular on the flight of Voskhod 2. Interestingly, the sextant on that mission was indeed used for navigation --after the astronauts landed in six feet of snow in the middle of Siberia hundreds of miles from the planned landing point. They were planning to get two Sun altitudes and then radio their position to the rescue teams but the helicopters arrived before they got their second LOP.
So now we get to Apollo. The Apollo spacecraft had a built-in sextant, clearly labeled as such in numerous diagrams. So therefore, it was flown and navigated by celestial navigation, right? Well, no, not really. The Apollo guidance and navigation system was the very first element of the Apollo design that was fixed and formalized by contract (with MIT). The contract was awarded just a couple of months after President Kennedy convinced Congress to authorize the massive Apollo moon program. The sextant was in there right from the start and it was quite simply a legacy of this very early contract date. In fact, MIT had previously designed a speculative "Mars probe" which would have flown to Mars (unmanned, of course) using inertial navigation and an automated sextant where it was intended to take a single photograph and then return that to Earth in a small Mercury-like capsule. The Apollo guidance contract and the initial design was settled in 1961. By 1969 when Apollo 11 made the first Moon landing, the sextant and many other aspects of the original Apollo design had been relegated to a backup role. And note, too, that this design choice was made before any of those tests with sextants during Gemini. Almost none of the Gemini experience contributed to the design of Apollo, and yes, the astronauts did see the irony in this.
The actual navigation of the Apollo spacecraft was fundamentally done by what we might call "backwards GPS". The ground tracking stations sent signals, carefully timed by atomic clocks, to the spacecraft which reflected those signals back with simple transponders. The extremely accurate round-trip travel time and the Doppler shift of the signal determined the range and rate to an accuracy comparable to modern GPS positioning. Indeed, GPS simply reverses the system, placing the atomic clocks in the satellites and determining accurate positions on the ground. But like a car in a long tunnel unable to receive GPS signals, so the Apollo spacecraft was not trackable by the ground when it was on the far side of the Moon. In addition, there was always the possibility that the ground might not be able to transmit their highly accurate position information back up to the spacecraft if there was a communications failure. And finally those ground observations could not determine the spacecraft's orientation which was critical for maneuvering. So Apollo also employed a highly accurate inertial navigation system to supplement ground tracking. This system used the sextant for orientation checks, and that's the primary reason the sextant made it through numerous Apollo design changes in the mid-1960s.
The inertial navigation system on the Apollo spacecraft was tightly integrated with the guidance computer itself. They were effectively one machine. The exact orientation of the spacecraft was critical to the computer-controlled rocket firings so that orientation had to be as accurate as possible. But gyros drift, and the INS could not keep its orientation accurately enough for more than about a day. That's when the astronauts got to shoot the stars. The procedure was known as "P52" (program 52) and it was as close as the astronauts came to manual navigation for the most part. In P52, the astronaut would re-align the inertial platform by requesting a particular star out of the standard list of 37. The computer would then electronically point the sextant at the expected position of that star. The astronaut would then look through the sextant (which had a high magnification for a sextant: 28x) and he would adjust the pointing by a few hundredths or a few tenths of a degree until the star was right on the crosshairs. He would press a button. The computer would read the offset in angle electronically to a precision of a thousandth of a degree, and the pointing error would be displayed. The astronaut would then accept or reject the adjustment and repeat the procedure with a second star. That was it. No manual reading of scales. No paper calculations or table look-ups. No adjustments for refraction or "irradiation" or anything else. You'll note that this "P52" process was really using the sextant as an astro-compass: the angles were relative to the orientation of the ship, not angles between celestial bodies.
The Apollo "sextant" was a true sextant in optical terms. It could measure angles between celestial objects by double reflection. But in many respects it was more like an electronic theodolite tied to the long axis of the spacecraft. Angles were measured relative to the spacecraft axes, and this made it ideal for the orientation check of the inertial platform. By the way, the "platform" in inertial navigation is sometimes an abstraction, but in the Apollo guidance system it was an actual block of metal. The INS and the sextant were mounted together on a rigid beryllium frame. Thus the INS plus the sextant plus the guidance computer constituted a single machine. Clearly this is very different from the mariner's sextant which was not mounted to anything and also very different from those experiments with handheld sextants during the Gemini missions. It's also worth mentioning again that the sextant served an important role it was not originally designed for. It was a computer-driven telescope. An astronaut could command the computer to point the sextant at a particular place on the Moon's surface and from orbit the Apollo CM pilot could try to locate the LM and the other two astronauts down on the lunar surface. This role, as a telescope, turned out to be every bit as critical as its role in navigation.
The Apollo spacecraft also had an instrument called a "telescope" mounted right next to the sextant. This was actually just a unit power viewer with a computer-readable orientation, useful for gross alignment of the sextant. The astronauts enjoyed joking about the fact that only a government project could pay good money for a telescope with no magnification. This telescope was used occasionally in experiments and at least once in actual practice. In 1968 on Apollo 8, returning from their historic lunar orbit mission, Jim Lovell accidentally entered a key sequence that told the computer it was back on the ground in pre-launch mode potentially wiping out the inertial platform's alignment. Just to be safe they ran "P51" which re-aligned the platform starting from scratch with gross alignment provided by the so-called telescope. This was a case where the astronaut had to identify the bright stars visually based on his knowledge of the constellations. It was a rare moment in manned space flight.
During the Apollo 8 mission, Jim Lovell took some of the only observations in the history of space flight that led to accurate determination of the position of a manned spacecraft --true celestial navigation in space. These were observations of the relative angles between celestial objects and the limb of the Moon or Earth or known landmarks on those bodies. He found the spacecraft's position this way over a hundred times. Like the "P52" orientation, there was a particular code for the program that determined the spacecraft's position (really its state vector in the orbital calculation). This was program "P23". If you browse the flight journals (the transcriptions of the audio traffic between mission control and the various Apollo spacecraft), you can easily find these various sextant activities by searching for those codes, P52 and P23. Like the orientation task, this navigation task was mostly automated. There were no hand calculations nor even options for hand calculations. The computer was not optional. Lovell's positions resulting from his observations were very good and confirmed the ground-derived positions. And because they confirmed ground, they were always thrown out. That was the way of it. Lovell's work confirmed that this could be done, but it did not change the fact that it was not required. The INS orientation checks --the P52 observations-- could not be dispensed with. There was no other way to constrain the slow drift in the orientation of the gyros of the inertial platform. But the position fixes were simply extraneous. They proved that the system worked, and astronauts on later missions still practiced the methods a few times on each mission, just in case communication with ground tracking was lost, but it was purely a backup.
When emergency finally struck during Apollo 13, the visual navigation checks waiting as a backup, turned out to be worthless. The oxygen tank rupture on the way to the Moon created a huge cloud of tiny ice crystals which stayed with Apollo 13 for many hours, like the "unfortunate dog" in Verne's "From the Earth to the Moon" written just over a century earlier. Those ice crystals, illuminated by the Sun, made it nearly impossible to see the stars and check the alignment of the inertial platform. Even when the stars became visible and the astronauts (including Jim Lovell from Apollo 8, now as commander) asked whether they should do a "P52", mission control told them not to bother. The ground tracking was more than adequate, and the expected drift in the INS was smaller than other uncertainties due to the damage from the explosion. So the sextant, there for backup in case of emergency, was irrelevant to the one true emergency of the Apollo missions.
After the Apollo moon missions came Skylab. While Mercury was derided as "spam in a can", Skylab was "spam in a really big can". The astronauts were tasked to prove their value in space by doing things that, it was already known, could be done better by machines, like astronomical observations as well as experiments that served no lasting purpose. And here is where we meet the T002 experiment. As Apollo could learn few lessons from Gemini, having been planned before the Gemini missions even began, Skylab could learn few lessons from Apollo since it was already well into planning while the Moon landings were underway. Skylab, therefore, picked up where Gemini left off, and the astronauts were given the unenviable task of using a handheld sextant yet again. This was "make work" experimentation. It led nowhere, and it is a classic case of the "space station" problem. Institutionally, space agencies find space stations irresistible: a permanent presence in space means a permanent role for the space agency. But alas, space stations are, to co-opt an old joke about boats, "big holes in space surrounded by aluminum into which you pour money". We spend vast sums of money to build them, so we damn well better find something to do with them. That's the bottomless pit into which those experiments with sextants aboard Skylab fell. The sextant in question, the Kollsman space sextant, was an accurate and compact handheld instrument, but it was really not much better than a Troughton sextant from 200 years ago. And it literally was the same sextant used on Gemini 12. It was as if Apollo had never happened.
The Kollsman space sextant as used on Skylab had a magnification of 8x, and it was used primarily to measure angles between pairs of stars and between stars and the limb of the Moon. In other words, they were shooting lunars. They're even recorded like traditional lunars: "Lunar Limb - Nunki - Far" and "Lunar Limb - Fomalhaut - Near". But these lunars were not intended to find GMT. Rather these observations, if processed for navigation, would have led to "cones of position" and a position fix (as I have described in years past, you can use lunars at a known GMT to get a position fix right here on Earth without any need for a visible horizon --space navigation on the surface of the Earth). Reported accuracies were around 10 seconds of arc at the 1 s.d. level. As I've noted many times, I get roughly 0.25' of arc for lunars or in other words 15 seconds of arc at the 1 s.d. level using relatively ordinary, well-calibrated marine sextants. So that's what your money buys you: 10 seconds of arc instead of 15 seconds of arc.
To give some idea of how ridiculously basic these experiments were, consider this description from the official report:
"The lunar limb— In this kind of sighting the angular subtense of the lunar disk diameter was measured by placing in tangency one limb of the moon viewed through the FLOS and the opposite lunar limb in the SLOS. Since the angular subtense of a planetary disk is a function of spacecraft range from the planet, this too is a potentially usable navigation measurement. It should be noted, however, that changes in the moon's angular subtense as measured from a vehicle in low earth orbit are extremely small This was FO3,six sets of sightings were required with 10 to 15 marks per set, GMT correlated."
Got that? An observation that has been performed since the late 18th century, bringing two limbs of the Moon together, described as if it's grand science.
While the basic principles of celestial navigation can continue to be applied to space navigation and have been on at least one unmanned deep space probe, the mechanics will probably never be the same. A digital camera beats a sextant in space any day. The space sextant was obsolete before it was ever practical.
NavList message boards and member settings: www.fer3.com/NavList
Members may optionally receive posts by email.
To cancel email delivery, send a message to NoMail[at]fer3.com