Mike’s Tweets

Finding Snoopy

As you may have seen on the web in the last few weeks, the search is on to locate Snoopy, or the ascent stage of the Apollo 10 Lunar Module:

Discovery.Com
Universe Today
Skymania
CollectSpace

I am helping Nick Howes and his students in the UK (and elsewhere) to try and locate Snoopy, the ascent stage of the Apollo 10 Lunar Module that was sent into a solar orbit back in 1969.

Before I go on, Nick is with the Faulkes Telescope Project which is just about one of the coolest things I’ve ever ever been associated with. They have telescopes that students can remotely control over the internet, for free! As a kid that grew up in Seattle, I can appreciate the utility of that! On their “about” page they say:
———-
The Faulkes Telescope Project is an education partner of Las Cumbres Observatory Global Telescope Network (LCOGTN).

“Our aim is to provide free access to robotic telescopes and a fully supported education programme to encourage teachers and students to engage in research-based science education. Access to our resources and those of our partners is provided at no charge to teachers and students.

Robotic Telescopes

LCOGTN operates a network of research class robotic telescopes. Currently there are two telescopes, one in Hawaii and the other in Australia. These telescopes are available to teachers for them to use as part of their curricular or extra-curricular activities and are fully supported by a range of educational materials and a team of educators and professional astronomers.
———-
Take the liberty to google Nick and you’ll find pictures of him hanging out with Sir Patrick Moore!

If you don’t know Sir Patrick Moore, you may google him ( which I already did for you here)

[Next someone is going to show me a picture of Nick hanging out with Brian May, then I’ll be really jealous!]

Here’s my first communication with Nick describing the work I (and a few others) had done on this project. I figured I’d post it here, and in the future I’ll just post things here as the default. I cleaned it up a bit, but this was originally written as an email, so there’s a lot of “stream of consciousness” going on. Please excuse any lousy grammar you may encounter.

___________
I have some stories to tell with some preliminary data. I’ve got 3 different state vectors at this point that all pretty much agree.

I’m working with Chuck Deiterich and Emil Schiesser. I’ve also had several discussions (and will have several more) with my good friend John Carrico (Astrogator_John).

Chuck is an ex-JSC Apollo “Retro”. Emil worked for the Mission Planning And Development (MPAD) group, and was a deep space navigation expert.

I’ve known Chuck since I met him when I was working on the Apollo 13 trajectory. He is a great guy, and really enjoys working on this stuff with me. Once he referred to me as a “Freelance FIDO” which I considered to be the ultimate compliment.

I told Chuck about this project and he went to work trying to find data for me. He didn’t work as Retro for the entire Apollo 10 mission (he worked Apollo 11, 13 and most of the rest, plus Skylab, ASTP, Shuttle, etc..) but he was there for lunar operations hanging around in the control room (that’s where I would have been). He worked the launch of Apollo 10 (Retro’s were responsible for Aborts) but didn’t work the LM/CSM ops at the Moon. However, he was there when they separated the Ascent stage of Snoopy from the CSM Charlie Brown. Chuck says that prior to launch, his guys had suggested to the Astronauts that they ought to have a set of maneuvers and a procedures in place for the separation of Snoopy and Charlie Brown, i.e. a set of burns to back Charlie Brown away from Snoopy after separation. The Astronauts didn’t think they needed that, as they they (the pilots), were perfectly capable of backing the CSM away without any pre-planned maneuvers.

When they were going to fire off Snoopy, they pre-pointed it in the right direction with the CSM, and then separated. They commanded the burn from the ground, but did not command the attitude.

When they went to separate from Snoopy, the air pressure did not get cleared out of the tunnel between the CSM and LM (as it was supposed to), so when they detached the spacecraft the air pressure immediately pushed Snoopy away and right into the sun where the Astronauts couldn’t see it! Then they had to figure out how to back away from a vehicle they could not see. It was a dangerous situation that could have been avoided with proper planning. [Side note. My twitter friend @Blackprojects provided me with a transcript of the exchange between Houston and the command module Charlie Brown during separation. Those are here and here .]

So, while I was recreating the departure conditions and geometry for the mission, Chuck and I were talking about how we’d tell our vectors were right and what sort of basic geometry we’d expect to see. For one, Chuck figured the burn would take place between the Earth and the Moon, probably very close to the Earth-Moon line. Since the Lunar orbit was retrograde, this is the side that would be where Snoopy would be going in the same direction as the Moon moves around the Earth, and thus would be the best place to get an escape trajectory. You want to leave the Earth-Moon system as tangential to the Moon as possible (Hohmann transfer).

If you were looking out at the Moon from the Earth, the Sun ought to then be to the right, because that’s when they’d launch to get the lighting the way they wanted. They wanted to have the sun behind them when they landed, with good shadows, so the terminator was usually to the east of the Earth-Moon line.

So, when I started to get my vectors all set up, I figured I should be able to see that geometry.

I have 3 different ways of producing my outgoing Snoopy trajectory.

1. A spherical post-burn state vector from Emil.
2. A spherical post-burn state vector from the document @BlackProjects tweeted us.
3. A combination of a pre-burn state vector Emil gave me, with a delta-V vector added to it that Emil gave me.

When I was creating number 3, Chuck and I were talking about the whole “sun behind the LM” event, and it occurred to me that I ought to be able to recreate that in my software
( Satellite Tool Kit ).

Since I have a pre-burn orbit state of the LM minutes before the burn, all I need to do is propagate backwards to the time of separation, and take a look at what we’d see if the LM was pointing along the delta-v vector (which it would be, since the CSM had to pre-align Snoopy to that orientation) and we were looking down the docking port. That’s exactly where the CSM would have been.

So, I did that, and saw this:

Snoopy Solar Release Geometry

Or you can see the actual footage (and a whole lot more Apollo 10 video) here:
Snoopy Release Video
Sun right in your face. I asked Chuck about this. I asked him why they didn’t know about it. He figured they must have seen it in the simulator and not figured it was a big deal. So Chuck fires off an email to Fred Haise. I’m sure you’ve heard of him, the Apollo 13 Astronaut (Bill Paxton in the Movie).

———————————-
Chuck: “Hi Fred, I am sure that you spent many hours in the Apollo CMS and LMS. I know that these simulators had star fields including the sun and moon. Can you remember the brightness of the simulated sun as compared to the real sun? For example, would the real sun be blinding but not the simulated sun?”
Fred: “Chuck, Best I can recall the simulator was not as bright as the real sun nor a medical risk to look at it. ”
———————————

I thought that was pretty funny.

I had another exchange with Emil that was also pretty awesome. We were discussing coordinate frames, and I was trying to figure out what frame the data he sent me was in. He told me:

—————————-
Mike,

During Apollo 10 the CSM lunar orbit the inclination values relative to moon fixed equator varied between 178.8 and 179.2 deg. The incliniation didn’t change much but the ascending node did due to the gravitational field. The spread in the numbers also reflects state determination uncertainty. The solutions for the CSM ascending node relative to moon fixed system varied from 180 to 183 deg so the LM value should be near this, I would think .

The inertial system used during Apollo was mean equinox of the nearest Besselian year, except that near the end of the program the system was frozen to Besselian year 1972 so we could shut down the weaving facility that manufactured the rope memories.

Emil
——————————-

The weaving facility that manufactured the rope memories! Now that is some wild stuff!

Here is the geometry of the system at separation:

Snoopy Escape Geometry with Sun and Earth Lines

Sun direction is yellow, Earth direction is blue. Purple arc is Snoopy departure trajectory, this is the hybrid constructed solution from the pre-burn state and the delta-v vector.

Here we are 1 hour later, and now I’m showing all 3 solutions

Snoopy Escape Geometry: Pole View

Here you can see the separation of the 3 solutions as we leave the Moon.

Snoopy Escape Geometry: 3 different Solutions

Now, where do these trajectories go?

Let’s see where we would expect. Looking at the directions in Cis-lunar space we get this:

Snoopy Escape Conditions

Earth-Sun line is the yellow line. Earth orbits the sun moving to the right. Snoopy is leaving the system mostly radial away from the sun, with a retrograde component. So the radial component is going to give us some extra eccentricity WRT the sun, and our retrograde component is going to take off some semi-major axis so we’ll speed up and Snoopy will have a shorter orbit period than the Earth does.

We leave the Earth with an orbit period of approximately 342 days. The way I figure, that gives us about a 14.7 year synodic period. So let’s look at the orbit for about 14 years and see where it went.

If we look at it in a Sun Inertial frame, we can see what’s going on:

Orbits In Heliocentric Space

If I get a bit tricky, I can display the same trajectories in a Sun-Earth Rotating frame, and see how things line up:

Snoopy Orbits in Sun-Earth Rotating Display

The yellow line is the Earth-Sun line, and Earth is all the way to the right of it, with the sun in the middle. Notice there are multiple lines here, from two different trajectories.

What this does tell us, is that Snoopy started off close to the Earth, and then spent the next ~15 years not very close to the Earth at all, as it moved around the sun. This shows one complete synodic period, and you can see that the different Snoopy trajectories got back near the Earth-Sun line in 1984.

This isn’t a pure synodic period problem though, because the Earth’s gravity speeds up and slows down Snoopy. This isn’t 2-body motion, so if I plot out the 2 orbital periods over time, I get this:

Snoopy RV Solution Orbit Period History

Snoopy ES Solution Orbit Period History

This tells me that the orbits are being perturbed a lot by their Earth encounters, and that their Earth encounters (and thus the perturbations) are drastically different. These encounters are spreading the different solutions apart.

By 1996 the we can see how far the 2 solutions have split apart. The Snoopy_RV (from the Report Vector) solution is near the Earth-Sun line, but the Snoopy_ES (from Emil Schiesser) solution is still far away.

Snoopy RV solution after 2 Revs (1996)

Snoopy ES solution after 2 Revs (1996)

By 2007 they’ve split even further. Snoopy_RV is on the Earth-Sun line, but Snoopy_ES is not there yet.

Snoopy RV Solution after 3 Revs (2007)

Snoopy ES Solution after 3 Revs (2007)

Here though is where it gets a bit tough. If you look at where these 2 different solutions would be on Oct. 1, 2011:


[caption id=”attachment_260″ align=”alignnone” width=”704″ caption=”Snoopy ES Solution Oct 1, 2011

“][/caption]

You can see that the spread between these 2 is away from the Earth and moving further away.

Lets look at these orbits in an inertial view:

Snoopy RV solution, Oct 1, 2011 Heliocentric View

Snoopy ES solution, Oct 1, 2011 Heliocentric View

Note the radius from the Earth in both cases. The best case has us 0.88 AU from the Earth!

So where are we? I’ve got 3 different solutions (I only showed 2 here, because these both involve 50+ year trajectories which tasks my system a bit – holding all that ephemeris in memory).

Somehow I need to figure out what “spread” of solutions is appropriate, and then use that to bound where we look. I’m going to chat a bit with John and Chuck to see where to go. I have Monte Carlo tools at my disposal (we used them on IBEX and are using them again now on LADEE) and I’ll likely get those out and generate a spread of vectors around one of my current 3 and see where the spread goes 42 years later. By varying both the magnitude and pointing of the Moon escape maneuver back in 1969, we should be able to see what the possible spread is in space.

My preliminary take is that this is going to be very hard. If the 2 states we have here represent the spread, we are going to be looking for something that is both far away and receding.

I’ll let Nick and his folks solve that. They are the gurus in that part. They know how to find asteroids and comets and have the people-power (and the mojo: Sir Patrick Moore!)

In the mean time, I’ll get back to my team and our high-tech modelling techniques.


  • Phil Karn

    The idea of finding Snoopy’s ascent stage in solar orbit ocurred to me independently some time ago, but I figured it would be so difficult that I didn’t even try to model it. I figured the tiny size of the target, plus the uncertainty in the departure trajectory and the chaotic nature of future interactions with the earth would make it extremely difficult to pin down. You seem to have confirmed this.

    Just to make sure, can you explain each of your state vector sources? I assume that at least one came from Snoopy itself as it propagated its state vector during the burn (perhaps separate ones from AGS and PGNS) and another came from MSFN tracking. Is that right?

  • Phil Karn

    I should still say that you’ve done some excellent work, even if the data you’re working from is just not good enough to give a definitive solution.

    So here’s an idea. How about looking for one of the early S-IVBs in lunar orbit? Prior to Apollo 13, when they were all sent crashing into the moon, they were sent into solar orbit just like Snoopy. By my count there should be five of them in solar orbit, from Apollos 8-12 inclusive. (Although Apollo 9 was an earth-orbital mission, its S-IVB was fired out of earth orbit as a test of the restart capability.) You probably know about Apollo 12’s S-IVB coming back in the 2002-2003 timeframe and being temporarily recaptured, so we know we can detect one were it to come close enough.

    Alhough the S-IVB is a bigger and brighter target than Snoopy, it is probably also lighter for its size and therefore more subject to the effects of solar radiation pressure. Have you accounted for that in your modeling of Snoopy’s trajectory?

  • Phil,

    I’ll break out the state vectors a bit more:

    1. A spherical post-burn state vector from Emil.
    Emil had an Apollo 10 post-mission report with a state vector in it. It turned out this vector also available in other sources I had. It was a moon-centered spherical state vector from after Snoopy “burn to depletion” burn. I don’t know what process produced that vector.

    2. A spherical post-burn state vector from the document @BlackProjects tweeted us.
    This is from MSC-00126, the Apollo 10 Mission Report. It has a slightly different state vector, also right after the depletion burn.

    3. A combination of a pre-burn state vector Emil gave me, with a delta-V vector added to it that Emil gave me.
    This is the state vector right before the depletion burn, and then I added the delta-v vector that Emil also gave me.

    All 3 of these were state vectors that I assume came from on-board, in that they were the state measured on board the spacecraft during the burn. I’m not sure why there are 2 different vectors, as the sources aren’t given. I do not believe that there is a vector from MSFN tracking. As far as I understand from Chuck and Emil, they didn’t track Snoopy after the burn was complete. The reports do not give the actual source of the data.

    I’d say that not only is the data not good enough, I don’t think any data can be good enough to give a definitive solution. In the quick monte-carlos that I did (I’ll have to post some of the graphs of these, they don’t appear to be up here) errors of less than 1 m/sec in delta-v create a huge swatch of possible outcomes 45 years later. The vehicle has made multiple encounters of the Earth-Moon system and has been perturbed in a slightly different way each time. I thought I had posted that stuff here but I don’t see it. Anyway, the trick is that any state vector with any sort of reasonable covariance is going to give a huge swath of possible outcomes.

    I am familiar with the Apollo 12 stage. I did a bit of work trying to reconstruct that trajectory back in 2003. They are bigger, and possible easier to see. They are less cool though, just because Snoopy is the only surviving Ascent module that flew.

    SRP is tough. All you can really do is estimate some average reflectivity over the known surface area and assume you get some sort of tumbling that averages everything out. I have a guess at the reflectivity and I’ve got the size down fairly well. But, you can turn the SRP on or off, and it doesn’t affect the total spread you get in a monte carlo. I think the best thing we can do is to come up with a spread and try to start scanning the areas in space that correspond to it. In the end, I expect this to look a lot like the asteroid surveys that are being done these days.