How often will the parts in your system upset or latch up in orbit?

Heavy-ion rate. The rate is Edmonds' effective flux form, R = ∫ Φ_eff dσ: the Weibull curve σ(L) = σ_sat [1 − exp(−((L − L₀)/W)^s)] is cut into LET bins, 20 per decade from the onset, each bin's step in cross section is multiplied by the environment's effective integral flux at the bin's log midpoint, and the result is interpolated between the two tabulated Z/X ratios that bracket yours. The largest effective flux over all ratios at each LET gives the geometry-independent upper bound shown by the SEU/Proton calculator.

Proton rate. For a full part marked estimated (the default for an onset of 15 MeV·cm²/mg or less), the proton cross section is estimated from the heavy-ion Weibull with the 200 MeV coefficient of the Edmonds and Bendel convolution, 7.13e-6 ∫ σ(L)/L² dL; a part marked known uses the proton cross section entered for it, and a part marked immune gets none. Parts with partial or no data get the estimate for each sampled curve unless they are marked immune. The cross section is applied against the integral proton flux of the environment above a threshold energy tied to the heavy-ion onset.

Monte Carlo, for each of 10,000 iterations. A part with a full Weibull fit keeps its computed heavy-ion and proton rates, times a factor drawn uniformly between 0.8 and 1.2. A part with no data is susceptible with its technology's probability; when it is, the saturated cross section is drawn from a lognormal around 2e-4 cm², the onset from the EHSP trend for that cross section, and W and s from a library of observed pairs. A partial part draws its onset and cross section from lognormals around the measured values. A null result uses the bound 4.02 L_test^−3.2, raised in proportion when the test fluence is below 6.5e5 ions/cm², with a lognormal spread. Every part's rate is multiplied by its quantity and the parts are summed.

Reported numbers. The mean and the 90 percent level of the heavy-ion, proton and total system rate over the iterations, and the 95 and 99 percent levels of the total; the mean time between events is the reciprocal of the 90 percent rate, and the risk class follows from it (low beyond a year, medium beyond 30 days, high beyond 7 days, very high below). The chance of at least one event over T years is 1 − exp(−R × 365.25 T). The derivation, the tables and their history are in the help page, section 3.

Turn a heavy ion cross section curve, or a proton cross section, into an on orbit rate. The tool integrates a Weibull or tabulated cross section against CREME96 and OMERE style environment spectra for GTO, GEO, LEO, polar, ISS and interplanetary orbits, applies the rectangular parallelepiped chord length correction, propagates fit uncertainty by Monte Carlo, bounds latchup with the Ladbury method when only a null result exists, and writes a PDF report. Rates come out per device per day and per mission.

Whose method this is. The heavy ion rate is computed the way Larry Edmonds of the Jet Propulsion Laboratory formulated it: as the integral of an effective flux with respect to the measured cross section, R = ∫ Φ_eff dσ, with the environment and the sensitive volume geometry carried in Φ_eff(L; Z/X) and the maximum over aspect ratio giving a geometry independent upper bound. The engine here descends from an implementation of his recommendations for CREME96 heavy ion rate calculations; the methodology gives the derivation and the references.

Related: fit the cross section first in the SEE cross section tool. For FinFET and other non planar parts the integral RPP model is known to be insufficient; the alpha law tool offers one possible approach, not yet validated. Feed the rate into the SEE Analysis Workbook to compare it with what the mission can tolerate, and read the help and methodology for the equations and their limits.

One header line, then one line per part, comma separated; lines that start with # are skipped. Header names are not case sensitive. Every part takes name, category, technology and quantity; category is full, none, null_result or partial (anything else is read as none).

A full part takes LET_th, sigma_sat, w and s. A none part takes susceptibility_prob (0.5 when blank). A null_result part takes test_LET and fluence (1e6 when blank). A partial part takes measured_LET_th and measured_sigma. Any part can take proton_status (known, immune, estimated or unknown; when blank, a full part with LET_th above 15 is immune and one below is estimated, and other parts are unknown) and proton_xs for a known proton cross section.

The import also accepts part, mfr, tech, qty, let0, sigma, width, shape or power, p_susceptible and measured_let as alternative column names. Export CSV writes name, category, technology, quantity, LET_th, sigma_sat, w, s, susceptibility_prob, test_LET, fluence, measured_LET_th, measured_sigma, proton_status and proton_xs, so an exported list reads back in.

The built-in heavy-ion environments are LET spectra in silicon for Z = 2 to 92 behind 100 mil of aluminum, stored as effective-flux tables for 11 aspect ratios from Z/X = 1 to 0.01 and 120 LET values, with the integral spectrum kept alongside. The proton environments are the integral proton flux above energy from the same transported CREME96 runs; the 1470 km table is a legacy AP-8 table kept as delivered. Each card names its orbit, shielding and source, and the help page gives the run settings.

An imported CREME96 .LET file in the native format (integral LET spectrum with % header lines) is folded in your browser into the same kind of table with the exact chord-length distribution, which takes a few seconds; a two-column file of LET and differential flux gets the isotropic geometric factor only. An imported proton spectrum takes the place of a built-in proton table, which holds the integral flux above each energy per cm² per day: a CREME96 FLUX or TRANS output (.flx or .tfx; the transported .tfx includes the shielding) is read natively, its proton block integrated above each energy and converted from per m² per s per sr to per cm² per day over 4π sr, the way the built-in tables were made; a two-column file is energy in MeV and the integral flux above that energy, in the unit picked next to the import button. Correction, September 2026: earlier versions used an imported proton file as it came, so a file in per cm² per s gave proton rates 86,400 times too low; recompute any proton rate made with an imported file. Imported environments stay in your browser and last until the page is reloaded.

R. Ladbury et al., "Statistical Analysis of Historical SEL Test Data to Provide A Priori Risk Estimates for Use of Unhardened CMOS Parts," IEEE Trans. Nucl. Sci., vol. 72, no. 4, pp. 1094 to 1101, Apr. 2025; and the CERN SEL database trend of Ladbury et al., IEEE TNS 2023.

L. D. Edmonds and F. Irom, "Extension of a Proton SEU Cross Section Model to Include 14 MeV Neutrons," IEEE Trans. Nucl. Sci., vol. 55, no. 1, pp. 649 to 655, Feb. 2008.

A. J. Tylka et al., "CREME96: A Revision of the Cosmic Ray Effects on Micro-Electronics Code," IEEE Trans. Nucl. Sci., vol. 44, no. 6, pp. 2150 to 2160, Dec. 1997.

Extended Historical SEL Priors (EHSP 2026-09): the susceptibility posteriors and the onset trend, derived in the SEL threshold methodology.

The full list is in the references of the help page.