The Apollo program is a constant reminder that we just don’t need so much to get the job done. Sure it’s easier with today’s tools, but hard work can do it too. [Bill Hammack] elaborates on one such piece of engineering: The Alignment Optical Telescope.
The telescope was used to find the position of the Lunar Module in space so that its guidance computer could do the calculations needed to bring the module home. It does this using techniques that we’ve been using for centuries on land and still use today in space; although now it’s done with computer vision. It was used to align the craft to the stars. NASA used stars as the fixed reference points for the coordinate system used to locate objects in space. But how was this accomplished with great precision?
The alignment optical telescope did this by measuring two unknowns needed by the guidance computer. The astronaut would find the first value by pointing the telescope in the general area necessary to establish a reading, then rotate the first reticle (a horizontal line) on the telescope until it touched the correct star. A ring assembly was then adjusted, moving an Archimedes spiral etched onto the viewfinder. When the spiral touches the star you can read the second value, established by how far the ring has been rotated.
If you’ve ever seen the Lunar Module in person, your first impression might be to giggle a bit at how crude it is. The truth is that much of that crudeness was hard fought to achieve. They needed the simplest, lightest, and most reliable assembly the world had ever constructed. As [Bill Hammack] states at the end of the video, breaking the complicated tool usually used into two simple dials is an amazing engineering achievement.
The preview image looks like someone explained all this with a good cup’o’coffee ;D
Check out this one, if you’re curious about very precise alignment scopes: https://einstein.stanford.edu/TECH/technology1.html#telescope
Yeah, I’ve seen it; I guess that such crudeness would have been much more tolerable back in the 1960’s than it would be today.
Just finished W. David Wood’s book “How Apollo Flew to the Moon” on the back of a comment made on these pages, highly recommended reading if this sort of thing floats your boat.
This looks to be an ultra simplified manual version of the SR-71, B-52, and Polaris automated sextant. This still requires precise timing to get a useful fix, but a very crude fix it is compared even to a paper or plastic sextant with a vernier scale, this looks to be more like a clinometer made from a weighted line and a protractor. What I was not clear on was if they radioed the readings to Houston or the computer could do this itself. I suppose with the crudeness of the reading and possible radar topo mapping from orbit they might not worry about altitude but where is the horizon reference, wouldn’t gravity be a bad reference with the weird peanut shaped core of the moon vs earth where a spirit level is so close to true handheld instruments can’t reliably detect gravitational distortion vs other noise?
I would still LOVE to build an arduino and stepper version of the fully automated star tracker, but farther I get from doing real math the less likely I will ever sit down and reverse engineer regular sextant calculations.
FWIW air travel was synonymous with celestial navigation at this point in history and every pilot and skipper could do it in their sleep though they had a seat in the cockpit for a dedicated navigator. The old jet airliners built until inertial and then GPS took over even had a little poke through periscope hole for the air-sextant to poke through, I knew one navigator who had a special vacuum cleaner hose which used this poke through port to clean up his station(difference in pressure between pressurized cockpit and FL30.
The AGU did the math I’m sure Houston got the information too but radio contact was not required for navigation.
The only reason we need chronometers on earth to get a fix is for longitude since it has no physical represenation. You can find latitude usefully accurately with no working clock. People sailed by latitude for centuries.
In space you can see multiple reference stars so you don’t need time to determine where you are. Likewise, you don’t need a horizon. If you can see Polaris you’re looking ‘up’, Alpha Curcis ‘down’,The Moon, Mars, Sirius, &c give you a side reference. Once the LM is on the moon, you’ve conveniently got a horizon. It’s worth mentioning that the AOT was on a swivel so you could lock it in 6 different views 60* apart to get a sighting on multiple reference points.
As for accuracy according to the NASA page below
“A reticle control enables manual rotation of the reticle for use in lunar surface alignments. A counter on the left side of the unit, provides angular readout of the reticle rotation. The counter reads in degrees to within ±0.02″ or ±72 seconds. The maximum reading is 359.88 degrees, then the Counter returns to 0 degrees . Interpolation is possible to within ±0.01 degrees.”
Further reading on the AOT
http://www.hq.nasa.gov/office/pao/History/alsj/aot.htm
https://heasarc.gsfc.nasa.gov/W3Browse/all/acrs.html
I think the point about timing is that in space, as on an airplane, you’re constantly rotating somehow and taking two different readings some time apart will produce inaccuracies.
Sure, but rotation is irrelevant to getting a fix on 3 of 50 ‘fixed’ bodies. I’m not understanding your point.
While you’re getting a fix the craft should be nearly rotationless since any rotation would be noted & interfere with trying to get a fix. Apart from then, rotation doesn’t particularly affect your course.
Ideally taking measurements some time apart gives your velocity, I don’t see where time affects the accuracy of your measurement.
At any rate they had the RTC of the AGU, or their own wristwatches (likely Omega Speedmasters, maybe a Breitling).
While there are fun sextant tricks like lunar distance(angle between Moon and a star) used before accurate chronometers became available or affordable to all mariners (I think the USNO uses Jovian moons) to get accurate time I no of no method to navigate without a time reference(except as you said latitude) double that when you are not doing spherical plane with altitude modifications for aircraft but in 3D space where the bubble level becomes useless without a gravity reference for that sphere. I have navigated using several stars both in the air and at sea but I always referenced from a GPS, quartz watch/clock, or clicking a stopwatch when I got a fix and waited until I was below deck to do the math and get the time on shortwave. I wonder how accurate the bubble level is on the moon with the lumpy gravity. I would love to see the formula that the LEM computer used for this navigation. Truth is I would love to try to rig an approximation of the whole thing on embedded hardware; radar and all(or sonar if I have to save $) with our modern cheap components.
@Dave
I’m fairly certain Apollo didn’t use spherical co-ordinates but instead used inertial space which is time-independent and Cartesian.
“[speaking about gyros]–with respect to inertial space, which, for navigation purposes, is identical with celestial space. ”
pg 32, paragraph 2, line 9
http://web.mit.edu/digitalapollo/Documents/Chapter7/gnccoursenotes.pdf pg17 is where the relevant parts start
Further reading about REFSMMAT reveals that Apollo used 2 separate coordinate systems for each half of the mission, each based around the Earth or Moon.
The link below indicates they used both spherical & quasi inertial coordinate systems.
http://www.ibiblio.org/mscorbit/mscforum/index.php?action=dlattach;topic=2854.0;attach=1368
And here’s a walkthrough of how they navigated.
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20090016292.pdf
Hot damn, those are some juicy links, work’s gonna suffer til I am done!
Wow, way better than I expected, still surprised it was easier to build or use, or more reliable than a standard air sextant.
Anyone, who missed it and is interested in this sort of thing, check out the more mechanical version of this, HaD posted about a mechanical targeting computer a while back.
That would be this article [0], on electromechanical artillery computers used by the U.S. Navy during WWII.
[0] http://hackaday.com/2013/03/13/retrotechtacular-mechanical-targeting-computers/
That Crude lander is well outside the capabilities of most PHD wielding Engineers today. A lot of it was innovations that we still do not see today as engineers refuse to get off their chairs anymore. Entire Teams worked tirelessly to create it with teamwork ideals that are lost outside of NASA.
Personally, I think every engineer needs to be forced at gunpoint to work on the items they design. and if anything is too difficult to repair, they are then beaten with the object until they understand that everything they learned in college is wrong.
I disagree, and you sound like someone who doesn’t work as an engineer. Today’s engineers are at least as good, if not better. It’s not a matter of “lack of teamwork” and too much “chair-sitting”, it’s a matter of money. Adjusted for inflation, the cost of the Apollo program would be close to $200 billion US dollars. Give engineers/scientists an interesting problem and a blank cheque for a budget and you will see the interesting stuff coming out. Just look at the LHC and tell me there are no good engineers today.
I agree (blank check engineering –> innovation) but I also see a lot of pet projects screwing up the works like some of the NASA manned Mars proposals. You need the blank check, a clear goal, a short deadline and a lot of potential shame (public visibility) then maybe innovation comes out to play.
Excellent piece!