e1012-dd-unc

Use when determine displacement damage (DD) assessment uncertainty factors for a spacecraft radiation analysis under ECSS-E-ST-10-12C §8.7: identify each uncertainty source (environment model, NIEL sc

SKILL.md

GitHub

ECSS Radiation Hardness Assurance — Displacement Damage Uncertainty Budget (space-systems/ecss/e1012-dd-unc)

Use when the task is to determine and document the uncertainty budget for a displacement damage (DD) radiation assessment, per ECSS-E-ST-10-12C §8.7. The procedure identifies each source of uncertainty, assigns a bounding factor, combines the factors into a single overall multiplier, applies it to the nominal proton-equivalent fluence, and checks whether the resulting design fluence is covered by the device's DD withstand capability with the required margin.

Domain quick reference

  • Displacement damage (DD) results from non-ionizing energy deposited by incident particles (protons, neutrons, electrons, heavy ions) that displace atoms from their crystal lattice sites. The damage level is characterised by the proton-equivalent fluence (p_eq/cm²), derived by applying Non-Ionizing Energy Loss (NIEL) scaling factors to each particle species in the environment.
  • §8.7 of ECSS-E-ST-10-12C requires that the DD assessment explicitly accounts for every significant uncertainty source. Four sources must be addressed: (a) environment model — uncertainty in the particle flux/fluence predicted by the space environment model (e.g. AP8/AP9 proton belts, solar proton event statistics); (b) NIEL scaling — uncertainty in the NIEL tables or functions used to convert each particle species to the proton-equivalent basis; published NIEL values carry their own experimental scatter; (c) shielding transport — uncertainty introduced by simplifying the spacecraft geometry or using an approximate particle-transport code; (d) device response data — uncertainty arising from limited lot-to-lot test data, inter-device scatter, or extrapolating published DD thresholds to the actual mission conditions.
  • Each source is assigned a numeric uncertainty factor ≥ 1.0 representing the upper-bound multiplier on the nominal estimate. Two combination methods are used in practice: multiplicative (product of all factors, conservative, used when source independence cannot be demonstrated) and root-sum-square (RSS, applied when sources are statistically independent, yielding a smaller but still bounding combined factor).
  • The combined uncertainty factor is applied to the nominal proton-equivalent fluence to yield the design fluence. The device's DD withstand fluence must exceed the design fluence by at least the required design margin factor (typically 2.0 for qualified parts per the ECSS radiation hardness assurance programme).

Workflow

  1. List every uncertainty source relevant to the assessment. At minimum address the four mandatory sources: environment model, NIEL scaling, shielding transport, and device response data. Reject any source label not in this set before assigning a factor.
  2. Assign a numeric factor ≥ 1.0 to each source, justified by supporting evidence (model validation data, published NIEL scatter bounds, transport code benchmarks, device test scatter). A factor of 1.0 must be explicitly justified; it implies negligible uncertainty from that source.
  3. Select the combination method. Use multiplicative unless the independence of every source pair can be demonstrated; use RSS only when that demonstration is documented. Compute the combined factor by the chosen method.
  4. Multiply the nominal proton-equivalent fluence by the combined factor to obtain the design fluence. This is the fluence the part must survive with margin.
  5. Retrieve the device DD withstand fluence from the radiation test data or the published tolerance level. Compute the margin factor as withstand fluence divided by design fluence.
  6. Compare the margin factor against the required value (default 2.0 for standard qualification paths). If the margin factor is below the required value, the part fails the DD margin check; increase shielding, select a more tolerant device, or reduce mission fluence.
  7. Document the full uncertainty budget: source list, assigned factors, combination method, combined factor, nominal fluence, design fluence, margin factor, and pass/fail conclusion. The budget is a required deliverable under §8.7.

Pitfalls

  • Assigning factors less than 1.0 on the basis that an uncertainty source is "well-characterised". A factor less than 1.0 implies the nominal estimate overestimates the true value, which is not a conservative assumption for a margin check. All factors must be ≥ 1.0.
  • Combining factors using RSS without documenting the independence argument. RSS gives a smaller combined factor than multiplicative; using it without justification unconservatively reduces the design fluence and may hide a margin shortfall.
  • Omitting the NIEL scaling source because it appears small. NIEL tables for protons at energies below ~10 MeV carry non-trivial scatter; omitting this source is not permissible under §8.7 without an explicit justification that its factor equals 1.0.
  • Reading a margin factor ≥ 1 as a pass when the required margin is 2.0. A margin factor of 1.5 means the device survives the design fluence but without the required safety margin; this is a margin non-compliance, not a pass.
  • Using the same uncertainty budget across different device types or mission orbits without re-evaluating the environment model factor. Each orbit profile samples a different particle energy spectrum, and the model uncertainty is spectrum-dependent.

Behavior contract (gate 3)

The factor validation, combination logic (multiplicative and RSS), design fluence computation, margin check, and end-to-end budget assembly are exercised by the gate 3 contract test: scripts/test_e1012_dd_unc.py against scripts/e1012_dd_unc_logic.py (stdlib unittest, offline). Run:

python3 scripts/test_e1012_dd_unc.py

Compliance

  • ECSS standards are freely downloadable (ESA); cite source and paraphrase per standards-map.yaml.
  • compliance: STANDARDS-REF, gated: false.