ECSS Crimping — Wire Stripping and Preparation (space-systems/ecss/q7026-wire-stripping-and-prep)
Use when the Preparation clause of ECSS-Q-ST-70-26 is the task: deciding whether a stripped conductor is fit to be crimped, before any contact is placed on it and before any tool closes.
Domain quick reference
- A crimp is a cold weld, so every defect that will ever limit the joint is already in the wire end before the tool closes. Nothing downstream inspects the strands that the blade cut, because the barrel hides them.
- Strip length is a two-sided band, not a target. Short leaves the barrel part-filled, so the compression the die applies lands on fewer strands than the band assumed. Long leaves bare conductor between the insulation and the barrel, unsupported by either.
- Every number here is per gauge. The strip band, the strand count and both damage allowances move with the wire, so an end is only ever compared with its own gauge entry, and a gauge the table skips has no band. The entry between two tabulated neighbours is not their average.
- Nicked and severed strands are different losses and carry different allowances. A nick is a stress raiser that work-hardens and parts under vibration months later; a severed strand is conductor that the joint never receives at all.
- Birdcaging and contamination are recoverable, and insulation damage and tinning are not. A birdcaged bundle can be re-laid and a dirty conductor cleaned, but damage on the insulation that stays in the harness and solder inside the crimp zone both travel with the wire.
- A solder-tinned conductor is not crimped. Solder creeps under the sustained barrel pressure and the joint that measured correctly at inspection relaxes into a resistive contact in service.
- Rework has a length cost. Cutting an end back and stripping it again consumes wire, so an end can be recoverable in principle and unrecoverable in this harness because the routed length is gone.
- An unqualified stripping method or a lapsed tool calibration voids the measurements rather than failing the wire. The end may be perfect; the record is not evidence.
Workflow
- Validate the gauge table: a two-sided strip band whose maximum is not below its minimum, a positive strand count and two damage allowances that stay below it. An allowance that reaches the strand count is not an allowance.
- Take the entry for the end's own gauge, refusing an untabulated gauge rather than interpolating a band between its neighbours.
- Grade the exposed length against the band, report which side it fell on and where inside the band it sits, and decide rework against reject on the wire margin that a re-strip needs.
- Count the strands in the bundle against the gauge, then the nicked and the severed against their own allowances. Refuse a bundle count that disagrees with the gauge and damage totals larger than the bundle, because both mean the record is inconsistent.
- Read the conductor form and cleanliness: birdcaging and contamination into rework, insulation damage and tinning into rejection.
- Check the stripping method against the qualified set and the tool calibration against its due date; a tool due today is still current.
- Take the worst of the four checks, name every check sitting at that level, prefix each finding with the check it came from, and roll the lot up with its accepted, rework and rejected counts plus the ends the record never covered.
Pitfalls
- Treating strip length as a minimum. An over-long strip passes every downward check and leaves unsupported bare conductor where the harness flexes.
- Carrying one strip band across a mixed-gauge loom. The band that fills the heavier barrel overfills or underfills the lighter one.
- Interpolating a band for a gauge the table skips. Two tabulated neighbours do not imply the entry between them, and the joint would be graded against a number nobody accepted.
- Merging nicked and severed strands into one damage count. They fail by different mechanisms on different timescales and the allowances are deliberately different sizes.
- Reading birdcaging as cosmetic. A bundle that does not enter the barrel round is compressed unevenly, and the strands on the open side never reach the wall.
- Accepting a tinned conductor because the crimp measures correctly. The measurement is taken before the solder has crept.
- Offering rework without checking the wire that is left. An end that can be re-stripped in the abstract cannot always be re-stripped in a harness whose routed length is already cut.
- Reporting an end disposition from a tool whose calibration has lapsed. The wire may be sound; the numbers are not evidence about it.
- Comparing a strip length with a band edge, or a calibration interval with its due date, by bare arithmetic. A value specified to land exactly on the limit can evaluate a few units in the last place past it after a unit conversion, so the comparison absorbs the representation error while the limit stays untouched.
Behavior contract (gate 3)
The gauge table validation and its refusal of an untabulated gauge, the two-sided strip length grading with its rework margin, the separate nicked and severed strand allowances, the conductor form and cleanliness reading, the stripping method and calibration check, the per-end worst disposition and the lot rollup with its completeness flag are exercised by the gate 3 contract test: scripts/test_q7026_wire_stripping_and_prep.py against scripts/q7026_wire_stripping_and_prep_logic.py (stdlib unittest, offline). Run: python3 scripts/test_q7026_wire_stripping_and_prep.py
Compliance
- ECSS standards are freely downloadable (ESA); cite the source and paraphrase per standards-map.yaml.
- compliance: STANDARDS-REF, gated: false.
