Space Radiation Effects Glossary

The vocabulary of radiation hardness assurance is small but unforgiving: a rate is not a dose, a threshold is not a cross section, a null result is not a zero. These definitions are written to be quoted in a design review. Where a term has a tool on this site, the link follows the definition.

Alpha law

Edmonds' empirical rule for how a device's single event cross section depends on the angle and rotation of the incident ion when the sensitive volume is not thin, replacing the cosine law with a fitted angular factor. Used for FinFET and other non planar devices. Alpha law tool

Anneal

Recovery of a device's parameters after irradiation as trapped charge escapes or defects migrate, usually accelerated by temperature. Total dose test methods specify a biased anneal step to check for rebound in some parts.

Bendel model

A one or two parameter fit for proton induced single event cross section versus proton energy, historically the standard form before Weibull fits became common for protons as well.

Bragg peak

The point near the end of a charged particle's range where its energy loss per unit length, and so its LET, is highest. Ions used in testing are normally chosen so the sensitive volume is well before the Bragg peak, where LET is nearly constant. Beam calculator

CREME96

Cosmic Ray Effects on Micro-Electronics, a widely used model of the galactic cosmic ray and solar particle environment and of single event rate calculation, revised in 1996 and now maintained as an online tool. The origin of most quoted LET spectra. SEE rate tool

Criticality class

In a single event effect criticality analysis, the category a function falls into: Error-Critical (effects unacceptable, probability budget), Error-Vulnerable (low rate tolerated, availability requirement) or Error-Functional (masked by mitigation, capacity requirement). SEECA workbench

Cross section

The effective area a device presents to a single event effect: events divided by fluence, in square centimetres per device or per bit. It rises with LET from a threshold to a saturated value. Cross section tool

Derating

Operating a part below its rated stress, most often the drain to source voltage of a power transistor, to avoid a radiation induced failure mode. Heavy ion burnout and gate rupture set derating levels well below electrical ratings. Derating advisor

Displacement damage

Lattice defects created when an energetic particle knocks atoms out of position in a semiconductor crystal, reducing minority carrier lifetime. Degrades optoelectronics, imagers, solar cells and bipolar gain. Scales with NIEL. NIEL tool

Displacement damage dose (DDD)

The non ionizing energy deposited per unit mass, in MeV per gram, obtained by multiplying particle fluence by NIEL. Allows damage from different particles and energies to be added. NIEL tool

Dose rate

The rate at which ionizing dose is delivered, in rad per second. Total dose test methods specify a standard window (50 to 300 rad per second in the high rate condition) because some devices, especially bipolar linear parts, respond differently at low rates. Standards page

EDAC

Error detection and correction: coding added to stored data so that a limited number of upset bits per word can be detected and corrected. Almost always paired with scrubbing so that errors do not accumulate. EDAC tool

Effective LET

The LET of an ion divided by the cosine of its angle of incidence, representing the increased charge deposited along a longer path through a thin sensitive volume. Valid for thin planar structures and misleading for deep or non planar ones.

ELDRS

Enhanced low dose rate sensitivity: the greater total dose degradation some bipolar and BiCMOS parts show at the low dose rates of space than at laboratory rates. Addressed by low rate or qualified accelerated test procedures. Standards page

Fluence

The number of particles per unit area that have passed through a surface, in particles per square centimetre. A heavy ion run to 1e7 ions per square centimetre is the conventional fluence for a null result. Test planning guide

Flux

Particles per unit area per unit time. The integral flux above an LET is the quantity that, multiplied by a cross section, gives a single event rate.

Galactic cosmic rays (GCR)

The steady background of fully ionized nuclei from outside the solar system, from hydrogen to iron and beyond, with energies high enough to penetrate any spacecraft. The source of most heavy ion single event effects. Highest at solar minimum.

Gray

The SI unit of absorbed dose, one joule per kilogram, equal to 100 rad.

Heavy ion

In radiation effects usage, any ion heavier than a proton used to deposit a dense charge track, from helium to gold. Accelerator beams of heavy ions reproduce the cosmic ray environment for single event testing. Facilities

IRPP

Integral rectangular parallelepiped: a rate calculation method that treats the sensitive volume as a box and integrates over the distribution of chord lengths through it, with a cross section curve providing a distribution of critical charges. SEE rate help

LET

Linear energy transfer: the energy a particle deposits per unit path length, in MeV cm2/mg when normalised by density. The variable against which single event cross sections are measured. Protons are below one; the heaviest cosmic ray ions approach one hundred. SEE guide

LET threshold

The lowest LET at which a single event effect is observed in a device, conventionally defined at a fluence of 1e7 ions per square centimetre. Requirements for destructive effects are written as a threshold the part must exceed, commonly 37 or 75. SEL test LET tool

Micro latchup

A latchup in which the supply current rises to a sustained level below the device's maximum rating, so that it may go unnoticed by a coarse current monitor while still requiring a power cycle to clear.

Multiple bit upset (MBU)

More than one bit in the same word upset by a single particle, which defeats single error correcting codes. Distinguished from multiple cell upset, which is several physical cells regardless of word membership. EDAC tool

NIEL

Non ionizing energy loss: the fraction of a particle's energy loss that goes into displacing atoms rather than ionizing them, tabulated by particle and energy. The basis for converting fluences between energies for displacement damage. NIEL tool

Null result

A test run in which no events were observed. It does not give a cross section of zero; it bounds the cross section at about 2.3 divided by the fluence at 90 percent confidence and says nothing about higher LETs than were tested. SEECA workbench

Proton direct ionization

Upset caused by a proton's own ionization rather than by a nuclear recoil, possible only in devices with very low critical charge, which now includes the most advanced commercial nodes.

Rad

The traditional unit of absorbed dose, 100 ergs per gram. Written rad(Si) when the absorbing material is silicon. A krad is one thousand rad.

Radiation design margin (RDM)

The ratio of the dose or fluence a part has been shown to survive to the mission requirement. A factor of two is the conventional minimum for total dose with well characterised parts. Standards page

Radiation hardness assurance (RHA)

The engineering process that ensures a system's electronics will perform in the mission's radiation environment: environment definition, requirements, parts selection and screening, analysis, testing, mitigation and verification. RHA guide

Rectangular parallelepiped (RPP)

The box shaped model of a device's sensitive volume used in single event rate calculation. Its depth decides how strongly the effective LET assumption applies.

Saturated cross section

The limiting value the cross section approaches at high LET, approximately the total sensitive area of the device for that effect. Cross section tool

Scrubbing

Periodically reading every word of a protected memory, correcting any errors, and writing the corrected data back, so that single errors do not accumulate into uncorrectable ones. The scrub period is a primary design lever. EDAC tool

SEB

Single event burnout: destructive failure of a power transistor when a heavy ion triggers the parasitic bipolar structure and the device enters avalanche. Depends strongly on drain voltage; mitigated by derating. Derating advisor

SEDR

Single event dielectric rupture: permanent breakdown of a thin dielectric, in an antifuse FPGA, a capacitor or a high field oxide, caused by a single heavy ion.

SEE

Single event effect: any effect caused by a single ionizing particle, including upsets, transients, functional interrupts, latchup, burnout and gate rupture. SEE guide

SEECA

Single event effect criticality analysis: the systematic, function by function assessment of what each single event effect does to a mission, which class it falls in, what rate or probability is tolerable and what evidence shows compliance. SEECA workbench

SEFI

Single event functional interrupt: a device stops operating normally until reset or power cycled, because an upset landed in control logic rather than data. Common in processors, FPGAs, DRAM and flash controllers.

SEGR

Single event gate rupture: destructive breakdown of a power MOSFET gate oxide by a heavy ion under gate and drain bias. Depends on ion energy and range as well as LET. Derating advisor

SEL

Single event latchup: a heavy ion or proton recoil turns on a parasitic bipolar structure in bulk CMOS, producing a sustained high current that destroys the part unless the current is limited and power is cycled. Worst at high voltage and temperature. SEL threshold predictor

Sensitive volume

The region of a device in which deposited charge contributes to a single event effect. Its dimensions, particularly depth, govern the angular dependence of the cross section.

SET

Single event transient: a temporary voltage pulse on a device output caused by a single particle, which may be harmless or may be captured by downstream circuitry as an error or a reset. Analog SET guideline

SEU

Single event upset: a change of state in a memory cell, latch or flip flop caused by a single particle. Non destructive; corrected by rewriting. EDAC tool

Single hard error (SHE) or stuck bit

A permanent change in a memory cell caused by one particle, usually a threshold shift in the cell's oxide large enough that it can no longer hold one state.

Solar particle event (SPE)

A burst of energetic protons and heavy ions from a solar flare or coronal mass ejection, lasting hours to days, during which single event rates can rise by orders of magnitude and dose accumulates rapidly. Worst day and worst week spectra are used as design cases.

South Atlantic Anomaly (SAA)

The region over the South Atlantic where the inner proton belt dips closest to Earth, so that low Earth orbit spacecraft receive most of their trapped proton dose and proton induced upsets during passes through it.

TID

Total ionizing dose: the cumulative ionizing energy absorbed by a device over the mission, in rad(Si) or gray, which degrades parameters through trapped charge in oxides. Reduced by shielding. TID, DD and SEE compared

Trapped particles

Electrons and protons held in Earth's magnetic field in the Van Allen belts. The inner belt is dominated by protons, the outer belt by electrons; both are the main sources of total dose in Earth orbit.

Upset rate

The number of single event upsets per unit time for a device or bit in a given environment, usually quoted per device per day or per bit per day. SEE rate tool

Weibull fit

The four parameter function (saturated cross section, threshold LET, width, shape) conventionally fitted to cross section versus LET data and used as the input to rate calculation. Cross section tool

Worst day, worst week

Solar particle event design spectra, from the October 1989 event, representing the average flux over the worst single day and the worst week of a large event. Used to size the rate a design must survive during an event.