SNAPShot Shenanigans

From: George Olshevsky (
Date: Thu Mar 30 2006 - 19:05:23 EST

  • Next message: Stephan Szyman: "iridium system"

    Monday (April 3, 2006) is the 41st anniversary of the launch of
    SNAPShot, still, as far as I know, the only working nuclear
    powerplant, designated SNAP-10A (SNAP: System for Nuclear Auxiliary
    Power), ever orbited by the USA. (Russia, on the other hand, has
    orbited a number of nuclear-reactor-powered spacecraft in its RORSAT
    and other series. We're not talking about mere RTGs here, which the
    USA uses on certain interplanetary missions, but full-fledged
    reactors). In going through the launches for 1965, I noted a
    moderately lengthy series of "debris" associated with this launch and
    I became curious as to what those objects might be. SNAPShot is not
    listed among the known orbital breakups and I presume it remains
    largely intact. Googling up "SNAP-10A" uncovered a number of Internet
    references, among which were several conflicting accounts of the
    mission. Perhaps the best account is a short one published in Russian
    that provides some good clues as to the nature of the objects orbited
    from this launch. I hope I've reconstructed the mission fairly
    accurately; I'd appreciate having errors and misconceptions corrected
    by more knowledgeable list members.
    According to the Summer 1965 TRW Space Log, SNAPShot was launched
    aboard an Atlas Agena D on April 3, 1965 from Vandenberg AFB. The
    payload was a conical structure about 1.5 meters in diameter at the
    base and about 3.7 meters tall, permanently attached to the front of
    the Agena D second stage. The reactor was a mechanism with a mass of
    about 115 kg located at the apex of the cone; the entire assembly had
    a mass of about 440 kg. The reactor comprised 37 fuel-moderator
    elements with fissionable U-235 fuel and zirconium hydride moderator.
    A cutaway diagram in TRW Space Log indicates the cone was hollow, with
    internal plumbing running just beneath the surface. The function of
    the reactor was to generate heat that would be converted to
    electricity via thermoelectric (T/E) elements. These elements were
    situated at various subsurface locations inside the cone below the
    reactor, and the reactor's heat was carried to them via molten
    sodium-potassium (NaK) alloy, pumped hot from the reactor at the
    cone's apex to a manifold at the cone's base. There the cooler but
    still molten alloy was pooled and pumped back to the reactor. The pump
    was located at the cone's apex alongside the reactor. Although it must
    have had a temperature of several hundred degrees C, the circulating
    alloy cooled the reactor through the conical radiator while
    transmitting its heat to the T/E elements. The electricity thus
    generated powered the various reactor components, such as the coolant
    circulation pump and the motors that controlled the moderator
    elements, and the spacecraft's radio transmitters, and it was also fed
    to a small 1 kg experimental cesium-ion engine to test the feasibility
    of this propulsion method. The ion engine is said to have been
    attached to the "side" of the Agena D, but I think it was mounted on
    the Agena's aft rack, which is where such secondary payloads and
    structures were usually positioned on other Agena-based spacecraft. At
    startup, I presume the essential mechanisms were powered by batteries
    or a fuel cell until the reactor went critical; there were no solar
    cells. A lithium hydride shield was located close beneath the reactor
    to prevent radiation damage to the spacecraft's electronics hardware
    from subatomic particles radiated from the reactor, and all the
    SNAP-10A components were radiation-hardened. Most of the electronics
    were located in a compartment at the base of the cone, as far as
    possible from the reactor.
    Also attached to the aft rack of the Agena D was a secondary geodetic
    payload, SECOR 4 (SECOR: Sequential Collation of Range, also called
    EGRS 4: Electronic Geodetic Ranging System). This was a rectangular
    box about 23x28x38 cm covered with solar cells and bristling with nine
    antennas to be ejected from the main spacecraft into a separate orbit.
    For a while Space Log listed two secondary payloads for this launch,
    but I have found no hard evidence of a second secondary payload, and
    later issues of Space Log delisted it. It has been suggested to me
    that there was one and it failed to release, but what with the SECOR
    and the ion engine, the Agena's aft rack seems to have been pretty
    full. Perhaps some reports counted the ion engine as that second
    secondary payload?
    At launch, the SNAPShot spacecraft was dormant. The command to go
    critical was issued on the second orbit, and it took six hours and 25
    minutes for the reactor to start up, and another two hours and 15
    minutes to achieve full power (500 watts). The earliest orbit I have
    for the SNAPShot, which became object 1965-027A (SatCat #1314), has
    period 111.52 minutes, inclination 90.03 degrees, and apogee/perigee
    of 1312x1280 km. The SECOR 4 was likely ejected well before the
    reactor went critical, becoming object 1965-027B (#1315).
    Unfortunately, it failed to turn on and thus became just another piece
    of space debris rather than an active satellite. I do not know whether
    anyone ever figured out how and why it malfunctioned. The earliest
    orbit I have for it has period 111.42 minutes, inclination 90.20
    degrees, and apogee/perigee 1315x1271 km, a bit lower but not
    significantly different from that of the main payload, the inclination
    suggesting it was released posterolaterally from the Agena aft rack.
    In starting up the reactor, it was necessary to liquefy the NaK alloy.
    Most of the startup time was probably consumed in melting it. I guess
    this was done by the reactor itself as it heated up during its "going
    critical" phase, when it did not need to be cooled. I don't imagine
    the satellite was launched loaded with already-molten alloy(!). The
    Russian account mentions that there was an insulating cover or shield
    around the reactor to trap its heat until the alloy melted and began
    to circulate. The shield was jettisoned after the reactor went
    critical and became self-sustaining, so this must be object 1965-027C
    (SatCat #1316), which was observed and catalogued at about the same
    time as the spacecraft and the SECOR subsatellite were. Its earliest
    orbit has period 111.51, inclination 90.21 degrees, apogee/perigee
    1311x1280 km, almost exactly the same as those of the main payload and
    suggesting it was released mainly laterally rather than forward or
    Once the coolant liquefied, the reactor ran continuously for 43 days
    to May 16, 1965. I guess the conical radiator, or certain parts of it,
    had an operating temperature of hundreds of degrees C, because the
    coolant would have had to remain liquid at all times; it would not do
    to have the stuff solidify in the conduits! Initially the reactor put
    out 600 watts, but as the experiment progressed this decreased to
    about 530 watts. The total energy output during the duration of the
    experiment is said to have been more than 500,000 watt-hours, which is
    in the right ballpark (43 days at 24 hours per day is 1032 hours,
    times 530 to 600 watts). The ion engine was turned on when the reactor
    achieved full power, but noise from the ion plume scrambled data
    transmission and the device had to be turned off. That part of the
    mission didn't work out too well. It had been hoped the reactor would
    run for at least 90 days, but a faulty voltage regulator triggered
    premature shutdown. According to the Russian report, shutdown entailed
    jettisoning two key reactor components without which the reactor could
    not function. These were called "reflectors," and I think their
    purpose was to reflect neutrons emitted during the chain reaction back
    into the reactor core. Without the reflectors, the neutrons would
    dissipate into space and the reaction would quickly cease. Indeed, two
    more objects were catalogued in late May and early June 1965, a couple
    of weeks after shutdown, as 1965-027D (#1389) and 1965-027E (#1399),
    and I presume these are the reflectors. Space Track flags the latter
    with an asterisk, indicating uncertain identification, but their
    earliest orbits (D: 111.50, 90.19, 1323x1267; E: 111.51, 90.19,
    1313x1278) are so similar to those of the main payload and the other
    objects in the launch that there can be little doubt that E is
    correctly identified. TRW Space Log listed E as a classified secondary
    payload in several issues but, as noted above, eventually delisted it.
    The primary SNAPShot mission is considered a great success, although
    the US didn't follow it up with more such space reactors.
    Some sources (e.g., Wikipedia) say that reactor shutdown was
    accomplished by reactor core jettison "into a higher orbit," but there
    is no object in such an orbit attributed to this launch. The Wikipedia
    account is probably based on how orbiting Russian reactors (as in the
    RORSATs) are deactivated, but this did not take place with SNAPShot.
    The story doesn't quite end here. As reported in the 13th edition of
    the History of On-orbit Satellite Fragmentations, SNAPShot suffered a
    series of seven "anomalous fragmentation events" beginning in November
    1979. These resulted in 51 more objects in the 1965-027 series,
    catalogued as 1965-027F through 1965-027BH. These were catalogued
    intermittently but fairly continuously from late November 1979 through
    late October 2002. More may yet turn up. What might these things be?
    One suggestion has been that these are all pieces broken off the
    radiator cone. But such chips would be flat and light, with a high
    surface area to volume ratio (to efficiently radiate heat) that would
    make them susceptible to "solar sailing." The orbits of 1965-027F
    through BH are remarkably stable. Some 27 years following the first
    fragmentation event their orbits have lost only a few kilometers in
    major axis. This tells me that they're dense objects, likely solid
    metal, not light and flaky. Some Russian RORSATs have leaked NaK
    coolant that solidified in the cold vacuum of space into blobs of hard
    metal, but leakage while the coolant was molten doesn't seem to have
    occurred during the 43 days the SNAP-10A reactor was operating.
    Rather,  the coolant must have rapidly solidified inside the SNAPShot
    conduits once the reactor turned off. Broken chunks of solid coolant
    and plumbing could have bounced around inside the radiator cone soon
    after the reactor cooled down. Then, perhaps because of a
    micrometeoroid strike or years of thermal stress in the space
    environment, a crack opened up in the side of the radiator. The crack
    gradually widened and allowed the debris within to float gently out
    and away from the main payload into nearby orbits. Most of the debris
    objects are quite small, with radar cross sections of less than 0.1
    sqm, but two are substantial, with cross sections of about 0.2 and 2.0
    sqm (these are 1965-027X [#18268] and AG [#18670], respectively).
    These happen to be the only two objects in the entire 1965-027 series
    listed as decayed, but neither one's TLEs show signs of significant
    orbit diminution around their decay dates, and I consider those dates
    administrative, not real. So as far as I know, all 56 objects from
    this launch remain in orbit. It is peculiar that the two largest
    fragments were administratively decayed while the many tinier
    fragments are still being tracked. In view of their relatively large
    radar cross section, I imagine the two big fragments might be pieces
    of the radiator that broke away where the radiator cracked open. They
    didn't break free right away but probably dangled from the satellite
    for a few years, since despite their size they were catalogued quite a
    bit later than the earliest fragments of the series. Repeated thermal
    stressing from time to time may have broken off more coolant chunks
    inside the radiator, perhaps with pieces of stainless steel conduit
    still attached. (Note the distinction between coolant >chunks< and
    coolant >blobs<: the former freeze inside the spacecraft's plumbing
    before breaking free, whereas the latter leak out while molten and
    freeze outside the spacecraft.) I can picture the satellite slowly
    tumbling and releasing a piece or two now and then from the gaping
    crack in the cone..
    I don't know whether this scenario explains how 56 objects came to be
    orbiting from this launch, but it fits the facts as I know them and is
    not obviously implausible. If anyone has a better idea or firmer
    knowledge of the nature of the secondary objects from this launch, I'd
    sure like to hear it.
    Subscribe/Unsubscribe info, Frequently Asked Questions, SeeSat-L archive:

    This archive was generated by hypermail 2b29 : Thu Mar 30 2006 - 19:08:48 EST