Space radiation

Radiation is the environmental factor that most sharply separates space electronics from terrestrial electronics. It sets part selection, drives architecture, and is the single hardest constraint on long-duration human missions beyond low Earth orbit.

Three sources

Trapped radiation: the Van Allen belts

Earth’s magnetic field traps charged particles, which spiral along field lines, mirror near the poles and drift in longitude. The result is two persistent populations.

Inner belt

Roughly 1000 to 6000 km. Dominated by protons up to several hundred MeV. Stable over years.

Slot region

Near 2 Earth radii. Normally depleted, but fills during large geomagnetic storms and can take months to clear.

Outer belt

Roughly 13 000 to 60 000 km. Dominated by electrons up to about 10 MeV. Highly dynamic, varying by orders of magnitude within hours.

The geomagnetic dipole is offset from Earth’s centre by about 500 km and tilted by about 11°. Where the field is consequently weakest, over the South Atlantic, the inner belt dips to altitudes as low as 200 km. This South Atlantic Anomaly is the reason a 400 km orbit sees any significant trapped proton flux at all, and it is why science instruments on low-altitude spacecraft are routinely commanded into a safe configuration on each SAA passage.

Solar particle events

A large solar eruption can raise the >10 MeV proton flux by five or six orders of magnitude for a period of hours to days. Total fluence in a severe event reaches on the order of $10^{10}$ protons/cm² above 10 MeV.

These events are the dominant acute risk to crews outside the magnetosphere and the dominant single-event upset driver for spacecraft during the event. Historically significant cases include August 1972, which fell between the Apollo 16 and 17 surface missions, and October 1989. The Carrington event of 1859 remains the reference extreme.

Because onset is fast and prediction is weak, the operational response is a storm shelter for crews and an autonomous safe mode for spacecraft, not avoidance.

Galactic cosmic rays

GCR flux is low but the particles are extraordinarily energetic and include fully stripped heavy nuclei up to iron. These HZE particles are only about 1% of the flux by number, but they deposit energy so densely along their track that they dominate both the single-event upset rate for high-threshold devices and the biological risk for long missions.

GCR intensity is anti-correlated with solar activity: a strong solar wind excludes them more effectively, so GCR dose peaks at solar minimum. That produces the uncomfortable planning conclusion that there is no good time to fly a long deep-space mission, only a choice between higher GCR dose at solar minimum and higher SPE risk at solar maximum.

Critically, GCR cannot be shielded away with any practical mass. This is the constraint that governs Mars habitats.

Units

Radiation quantities are a common source of error because four distinct things are all loosely called “dose”.

For electronics, dose is quoted in rad(Si) or Gy(Si), because the deposited energy depends on the target material. A dose figure without the material in parentheses is incomplete.

Effects on electronics

Total ionising dose

Charge trapped in gate oxides accumulates over the mission, shifting threshold voltages, increasing leakage current and eventually preventing the device from switching. TID is cumulative and its onset is gradual, so it manifests as a lifetime limit rather than an event.

Mission dose depends strongly on orbit and shielding. Behind 2.5 mm of aluminium, order-of-magnitude annual figures are:

These are planning numbers only. A real budget requires a model run against the actual trajectory, using AE9/AP9 or SPENVIS with a sector-shielding analysis of the as-built structure. Dose falls steeply with shielding at low thickness and then flattens, so the first millimetre of aluminium buys far more than the fifth.

Displacement damage

Non-ionising energy loss knocks atoms out of the crystal lattice, creating defect sites that act as recombination centres. This degrades any device that depends on minority carrier lifetime: solar cells lose end-of-life power, optocouplers lose current transfer ratio, CCDs and CMOS imagers develop hot pixels and worsening charge transfer efficiency.

Solar array sizing must account for it explicitly. A GEO array is typically sized for 15 to 25% power loss over 15 years, most of it from displacement damage by trapped protons. See Power systems.

Single-event effects

A single ionising particle deposits enough charge along its track to disturb a circuit node. Unlike TID, these are stochastic events, and their rate scales with flux rather than accumulating.

The design response is layered:

Latch-up protection is the one that must be in hardware. By the time software notices, the device has already drawn destructive current, so the supply itself has to detect the overcurrent and remove power within milliseconds. This is a standard argument for using latch-up-immune silicon-on-insulator processes in critical paths.

Effects on crews

The comparison that matters is the third row against the last. A round-trip Mars cruise alone consumes most of a career limit before any surface time is counted. This is not a shielding problem that more aluminium solves, for a reason worth understanding.

Why more shielding can be worse. A high-energy heavy ion striking a high-atomic-number nucleus fragments, producing a shower of secondary particles including neutrons. For GCR, adding aluminium initially reduces dose, reaches a shallow minimum, and then increases dose as secondary production dominates. Hydrogen-rich materials avoid this, because hydrogen has no nucleus to fragment and is the most efficient moderator per unit mass:

Polyethylene

Roughly twice as effective as aluminium per gram per square centimetre against GCR

Water

Comparable to polyethylene, and dual-use as a consumable and as storm-shelter mass

Regolith

The only option that is not launch-mass-limited, which is why lunar and Mars habitat concepts bury or cover the pressure vessel

Active magnetic shielding

Repeatedly studied; the required field strength and superconductor mass have never closed for GCR

A practical architecture therefore separates the two threats. Solar particle events are survivable with a small, heavily shielded storm shelter occupied for hours to days, because SPE protons are comparatively low energy. GCR is not avoidable and is managed by minimising mission duration, which turns a radiation problem into a propulsion problem. See Nuclear propulsion.

Designing to the environment

Two rules of thumb survive most programmes. Apply a radiation design margin of at least 2 on TID against the predicted mission dose, because part-to-part and lot-to-lot variation is large. And never accept a datasheet radiation number for a commercial part without lot-specific test data, because commercial processes change without notice and tolerance is not a specified parameter.

References

• Holmes-Siedle, A. and Adams, L. Handbook of Radiation Effects, 2nd ed., Oxford University Press, 2002.
• Barth, J. L., Dyer, C. S. and Stassinopoulos, E. G. “Space, Atmospheric, and Terrestrial Radiation Environments”, IEEE Transactions on Nuclear Science, 50(3), 2003.
• Zeitlin, C. et al. “Measurements of Energetic Particle Radiation in Transit to Mars on the Mars Science Laboratory”, Science, 340:1080, 2013.
• Hassler, D. M. et al. “Mars’ Surface Radiation Environment Measured with the Mars Science Laboratory’s Curiosity Rover”, Science, 343:1244797, 2014.
• Ginet, G. P. et al. “AE9, AP9 and SPM: New Models for Specifying the Trapped Energetic Particle and Space Plasma Environment”, Space Science Reviews, 179:579, 2013.
• NASA-STD-3001, NASA Space Flight Human-System Standard, Volume 1.