Apollo 11 vs. mascons

In the course of some other work, I ran across an interesting historical
tidbit, one that many of the usual references get wrong.

We all know that Apollo 11 landed about 7km from its target, mostly
because it overshot. The last little bit of that was because Armstrong
had to go a bit farther downrange to find a decent landing spot, but the
real problem was that the descent path overshot the target by several
kilometers. There were indications earlier in descent, from both landmark
timing and radio tracking, that the LM was slightly downrange of where it
should be at the time, but the error wasn't corrected during approach
because the crew was distracted by the computer alarms at the crucial
time. The question is, where did the navigation error come from?

The answer you'll find in a lot of references, including normally
authoritative ones like Murray&Cox, is that the orbit was perturbed in
unexpected ways by poorly-mapped mascons. This is (mostly) wrong.

When in doubt, go back to primary sources... The Apollo 11 Mission Report
(found in Apogee's Apollo 11 volume 3) discusses the navigation problem in
a bit more detail (Apogee pages 80-81). There were *two* different
errors, a small one in latitude (crossrange) (about 1.4km) and a much
larger one in longitude (downrange) (about 6.7km).

The minor latitude error is ascribed to deficiencies in the lunar gravity
model, and some related issues. They also caused a longitude error that
was deemed too small to worry about.

For the major longitude error, on the other hand, the finger points at a
large number of small perturbations in the last few orbits before descent:
unbalanced LM RCS use for thruster tests, undocking, and stationkeeping;
operation of the LM sublimator for cooling; and "possibly tunnel and cabin
venting".

Preventive measures on Apollo 12 (which came down pretty much exactly on
target) featured some last-minute radio tracking and a final position
update(*) to take out late errors of all sorts. And they did include some
gravity-model improvements. But also, there were a number of procedural
changes to reduce pre-descent LM thruster firings and minimize their
effect on the orbit (e.g., the CSM did the undocking and stationkeeping
maneuvers), and generally minimize late perturbations.

(* Done by lying to the computer and telling it the desired landing site
had moved, because that required rather less keyboard work than correcting
its estimate of LM position and velocity, and so was easier to fit into
the very busy approach schedule. )

So, bottom line, the usual histories are wrong: there *was* a gravity
issue but it was relatively minor, while the big overshoot was caused by
neglect of accidental last-minute perturbations, from thruster activity in
particular.
--
spsystems.net is temporarily off the air; | Henry Spencer
mail to henry at zoo.utoronto.ca instead. |