Superbird A issues

From: Matson, Robert (
Date: Wed Dec 11 2002 - 20:32:48 EST

  • Next message: Kevin Fetter: "more geo sats :)"

    Hi All,
    Well, it took me about 4 hours last night, but I finally reduced
    all the Superbird A observational data from February 25 to
    December 10th, 2002.  Observers were:  Frikkie van Zyl, Ron Lee,
    Ed Cannon, Don Gardner, Leo Barhorst, Bjorn Gimle, Mike McCants
    and Tom Wagner.  I have not computed an updated spin axis yet --
    will do so later tonite.
    One comment about the non-"Main Event" flashes that have been
    observed by Don Gardner and perhaps others -- this behavior
    has been seen in years past, and when the surface normals are
    plotted it is quite clear that it cannot be a simple reflection
    from a different surface on Superbird A.  (The surface normals
    do not lie on a curve that is concentric with that of the main
    flashes.)  What this means is that the extra flashes result
    from a double reflection -- probably a combination of a solar
    wing and the body of the satellite.
    And this is a good segue into a related topic -- the source of
    the main flashes themselves.  We (or at least I) have assumed
    all along that these bright glints could only be coming from
    the very large solar wings on Superbird A.  And due to the
    the fact that the flash brightness is more or less symmetric
    about the phase shift, and that solar panels have very different
    optical properties on their front and back sides, the two solar
    wings on Superbird must be oriented in nearly anti-parallel
    directions.  (From an operations standpoint, this seemed like
    a reasonable emergency orientation for the solar wings:  it
    ensures that one wing is always sunlit when the satellite is
    not in eclipse.)  Due to a slight misalignment between the
    two wings, the flashes for one wing start and end a minute
    or two before the other.  This neatly explains what is observed
    from the ground:  flashes starting with a 22.6-second period,
    transitioning to an 11.3-second period (alternating flashes
    from each wing), and ending with a 22.6-second period.  It
    also explains why the satellite is spinning up -- differential
    torque from the arrays due to the unsymmetrical arrangement.
    There's just one problem:  this isn't what's happening!
    A few months ago while attending a space mission analysis
    class, I had an opportunity to talk with some other attendees
    involved in comsat operations.  I asked them if it was standard
    practice to point solar wings in opposite directions when
    pointing and control for a satellite is temporarily lost.  They
    said this is NEVER done.  Solar wings are the primary source
    of torque on a satellite, and thus symmetry must be maintained.
    Due to the very different reflective properties of the front and
    back sides of solar wings, they must point in the same direction
    to maintain a torque balance.
    So what does this tell us about the flashes we see from
    Superbird A?  Simply that they cannot be coming from the
    solar wings.  Specular glints can only be produced by the
    front side of the arrays.  So what we've been seeing all
    along are actually glints from the satellite body!
    At first this might not seem possible for a satellite at
    a range of some 38000 km or more.  But when you do the math,
    it's not at all unreasonable.  Compare to Iridium for
    instance.  Typical range for a bright Iridium flare is from
    900 to 1000 km.  Call it 1000 to simplify things, and assume
    a visual magnitude of -8.
    Now if you kept everything the same, and just moved an
    Iridium satellite to a range of 38000 km, the brightness would
    fall by a factor of 38 squared (1444) which is 7.9 visual
    magnitudes.  That drops your glint magnitude to roughly 0.
    The flashes from Superbird can be as bright as about +2,
    though usually +3 is more typical.  +2 is a factor of 6.3
    dimmer, so you see it really takes very little surface area
    on a specular surface to produce a +2 or +3 glint from GEO.
    As it turns out, the sides of the FS-1300 Bus built by Loral
    Space Systems have a decent specular component based on the
    images I've seen and are sufficiently large to produce glints
    of +2.  These are the two sides from which the solar wings
    extend.  The other sides do not look favorable for producing
    All that said, the mystery is:  when *DO* the solar wing
    glints occur?  I suppose it's possible that the wings are
    oriented with their surface normals close to parallel with
    the spin axis (i.e. very small cone angle).  And this cone
    angle may produce reflections that only rarely intersect
    the earth (if at all).
    For curiosity's sake, I decided to take a look at the impact
    of changing the coning angle on the predicted flash time
    for central Texas.  Using my old axis and precession rate
    (probably good to within 10 minutes) here's what my software
    computes for December 12th UT:
    89.95 deg (main glint):  03:50 UT
    89 deg:  03:57
    88 deg:  04:05
    85 deg:  04:29
    80 deg:  05:08
    75 deg:  05:48
    70 deg:  06:27
    60 deg:  07:46
    50 deg:  09:06
    40 deg:  10:27
    30 deg:  11:51
    The good news right now for North America is that a wide
    range of coning angles can be investigated during our long
    winter nights.  I should point out that this only addresses
    one half of the coning solution -- cone angles can be greater
    than 90 degrees (to cover the anti-pole case).  These of course
    will occur earlier than the 03:50 time indicated above.  The
    east coast is in a better position to evaluate the early
    solutions due to their earlier sunset.
    By having different observers sign up to keep "glint watch"
    for times corresponding to different coning angles, it wouldn't
    take a handful of observers too many nights to cover all the
    angles.  Looks clear here tonite, so I may take a little time
    to cross off a range of angles.
    Unsubscribe from SeeSat-L by sending a message with 'unsubscribe'
    in the SUBJECT to

    This archive was generated by hypermail 2b29 : Wed Dec 11 2002 - 20:54:25 EST