Fake Banner
    The Risks Of Shielding Electronics
    By Project Calliope | June 21st 2011 07:44 PM | 1 comment | Print | E-mail | Track Comments
    About Project

    Alex "Sandy" Antunes is the mastermind behind 'Project Calliope', a pico-satellite funded by Science 2.0 and being launched in 2011 by a mad scientist...

    View Project's Profile

    Calliope, like any Low Earth Orbit satellite (LEO), is going up to, well, LEO.  Space weather-- radiation and energetic particles emitted from an active Sun-- can damage satellites.  This region of space is partially protected from the worst effects of space weather by the Earth's ionosphere, but it is an active and threatening place.

    If space wasn't active, there wouldn't be any point in sending up Calliope to measure it.  However, we'd prefer to keep the physical damage to the electronics to a minimum.  The primary source of damage due to solar activity is due to highly energetic electrons, protons and ions emitted by the Sun.

    These particles can penetrate past the satellite's skin and the surface of the electronics and dump their energetic charge into the electronics itself.  This can cause glitches-- Single Event Upsets, where the electronics briefly get a wrong signal value.  It can also degrade or erode the solar panels and other sensitive bits-- though this is less of a factor for our short-lifetime (6 weeks, nominally) mission.

    If Calliope was a person, the radiation could damage DNA and similar.  I'll cover that in a future Daytime Astronomer, when I recap today's Space Weather Enterprise Consortium.  For now, though, we can just shield Calliope from particles and be safe, correct?

    In a French word, 'non'.  Shielding can help protect electronics, but due to the actual mechanism that causes damage, shielding can also increase the risk of damage.  Unlike armoring a medieval knight, where you want to add as much armor as they can carry, for satellites you need to understand how the potentially damaging particles interact with the spacecraft.

    First, it is only specific particles, of a specific energy, that will react with the electronics.  Particles of low energy will be shielding by the body of the spacecraft or even the paint and thin silicon layer coating the electronics.  Like these incredibly detailed schematic show:

    Particles (electron or proton or ion): --->

    the electronic component:  |     |

    their interaction (particle stopped):

    -->|     |

    Particles of very high energy will barely be slowed by the small amount of material and will typically pass straight through the satellite and electronics, reacting negligibly:

    ----|-----|--->

    The damaging particles are, in a Goldilocks fashion, the particles at a 'just right' energy.  They have enough energy to pass through the blocking materials to get into the electronics, but not so much energy as to zip out.  Instead, they deposit all their 'impact' into the electronics, causing the SEU or damage:

    ---|->  |

    To protect from these, first, you have to know what the range (or spectrum) of particles are in space.  Here are two sample spectra, of very different shapes.  The energy of the particles range from some low 'E0' up to higher values, ending at 'E3'.  One spectrum has more low energy particles than high energy, the other is the opposite (more high energy particles than low energy ones).

    2 sample particle spectra

    Let's add a detail-- say all particles of energy E0 or lower are blocked by the satellite as it is, and do no damage.  And all particles of energy E2 or higher go straight through and cause little or no damage.  The only particles that will damage our satellite are those of energy 'E1'.

    If the particles appear as per Spectrum 1, then, we get around 20 of the damaging E1 particles.  If Spectrum 2 is the case, we get around 30 of the damaging E1 particles.  The damage the satellite receives obviously depends on which spectrum the Sun is emitting.

    Now let's add shielding.  Shielding attenuates, or slows down, all particles.  If particles are of low energy, they get slowed down so much they are completely blocked.  If particles have high energy, they lose some energy passing through the shielding, but still continue on-- at a lower energy.

    So now we add a block of shielding.  Let's say it's enough to slow down each particle by 1 of our labeled energy levels.  An E3 particle gets slowed down to become an E2 particle.  E2s turn into E1s.  E1s turn into E0s.  E0s get blocked completely by the shielding.

    What then happens to our satellite if we shield it?  Suddenly, with shielding, everything at E1 or lower is blocked but the E2 particles get 'downshifted' to E1 and cause damage (and the E3 get slowed to E2 and therefore still pass through without causing any trouble).

    Under this, Spectrum 1 has 30 formerly-E2 particles that get slowed by the shielding, 'become' E1, and thus damage the electronics.  With Spectrum 2, there are only 20 formerly-E2 particles that turn into damaging E1 particles.

    As a little table, then, we find that, without shielding, the satellite takes more damage if the Sun emits Spectrum 2, but with shielding, Spectrum 1 will do more damage.  As a table:

       bare, no shielding 
     with shielding 
    Sun emitting
    Spectrum 1
     20 damaging
    particles
     30 damaging
    particles
     Sun emitting
    Spectrum 2
     30 damaging
    particles
     20 damaging
    particles

    The choice of whether to add shielding or not therefore depends not just on the amount of shielding, but on understanding what the specific energies and amounts of particles the Sun is likely to emit.  Adding more shielding can, in some cases, actually increase the damage and risk to your electronics.

    For a larger space mission, you can obtain the predicted space environment and also create a 3D model of the 'attenuation curves' for the materials and thicknesses of your satellite in order to create an estimate on particle damage.  To be very accurate, you should run multiple 'Monte Carlo' simulations of particles hitting the satellite, since real particles have a 'probability of interaction' rather than the simple hit/miss model I give here.  That is a subject for a future... no, wait, that was my Master's thesis way too long ago (later published in Experimental astronomy, vol. 4, no. 2.)

    Fortunately (?) for Calliope, we are not just limited in weight, but aren't really worried about short-term transient damage like SEUs.  Instead, we'll just take our lumps and hope that the bulk of the data we download-- itself just a fraction of the entire data captured-- will suffice.

    Alex

    Launching Project Calliope, sponsored by Science 2.0, in 2011
    News every Tuesday at The Satellite Diaries, every Friday at the Daytime Astronomer

    Comments

    MikeCrow
    I've worked with both Space hardware, and latch-up testing in cmos electronics.

    If I remember correctly our box had ~1/2" aluminum walls, I don't know how much shielding the rest of it had though. Nor do I know what it's lifetime was expected to be.

    At another job we tested a lot of different parts for latch-up sensitivity. With cmos if you exceed the rail voltage(+ or -) of an imput or output, you inject current into a parasitic scr, which you could turn on, and shunt the supply to ground.
    Depending on it's gain, it would either shut off after the event, or stay latched until the power was reset. That said, if your supply will automatically reset after an over current event, your sustem should recover from those types of SEU events.

    Good luck with Calliope!
    Never is a long time.