What could have been known, record 10
The Warning Before Flight 981
Freeze the DC-10 cargo-door record on 27 June 1972. Choose a rule before the later documents appear, then advance every release to the end. The warning was real. So were the limits on who could act, what paperwork could prove, and what nobody yet knew about TC-JAV.
Memo to accident
calculating
days with the failure chain already described
The interval is computed live from the memo date and the observed accident date. Before the crash, this archive never reaches a binding order that both made false locking impossible and verified enough pressure relief to protect the floor.
The counterfactual is conditional
An FAA order alone is not a proved rescue.
TC-JAV was Turkish-registered. The record does not establish foreign adoption of a U.S. order. A Douglas delivery hold with physical verification would have refused the aircraft because the expected SB 52-37 support bearing was absent.
That scores the rule against the outcome that actually followed. It does not reveal the economic cost, the actions on an unseen branch, or a probability of accident.
Layer one: lock the rule
27 June 1972, end of day
The day on F. D. Applegate's memorandum is the boundary. No time survives, so end of day in California is a declared convention, not a recovered timestamp. The public NTSB recommendations had been adopted internally, but were not issued until 6 July. The firewall therefore keeps them on the later side.
Decision replay
Step 1 of 2: commit
Which rule should govern the next dispatch? The choice is stored only in this page and is never transmitted.
Rule locked
Five records are inside the firewall
The later record is still hidden. Advance it one dated release at a time.
The complete trajectory
Every release is shown below, including the public NTSB recommendations that the scout's 19-record list missed. Nothing is downsampled. The highlight follows the replay; later rows remain visibly present as withheld records.
Layer two: authority is part of the mechanism
Who can stop this aircraft, and what counts as proof?
Three questions are often collapsed into one. A regulator's rule must reach the operator. A manufacturer's hold can reach manufacture and delivery. A physical check can contradict a completion record. Operate those questions separately.
Retrospective physical finding
Schematic, not to scale. The French inquiry found that the extra SB 52-37 support bearing expected to have been fitted before delivery was absent. It also found evidence that application of the modification had only begun.
Not allowed back through the firewall
February 1976 final report: TC-JAV's support bearing was absent; lockpin and switch rigging were incorrect; an improvised shim was present; and the warning could extinguish before the door was locked. These findings explain the false paperwork after the fact. They were not facts Applegate could inspect on 27 June 1972.
What the 1972 record could not have told them
The sharpest unavailable quantity is the false-negative rate of the manufacturer's completion records: how often a factory record saying a modification was incorporated corresponded to hardware that was absent or incomplete. No denominator for that rate appears in this archive. Nor could the cutoff record reveal TC-JAV's later rigging, shim, inspection practice, passenger load, or exact decompression sequence.
The 39-aircraft remediation table
The published anchor breaks under a literal day count.
The House print supplies all 39 ship-level completion rows. That permits a real fleet curve, but it also exposes a one-day problem in the report's best-known aggregate.
Count delays from
Senate issue date: 1972-07-03
Switch to the Federal Register's 8 July bulletin date. The live result is being computed. The report's 5 is reproducible only by pairing the later date with a strict boundary, not by pairing its own 3 July issue date with an ordinary inclusive 90-day window. The page preserves all three quantities.
“Aircraft-days” here means the sum of calendar days from the chosen bulletin date to each listed completion. It is not flight time, passenger exposure, aircraft availability, or an accident-risk denominator.
Open all 39 completion rows
| Operator | Fuselage | Completed | Delay |
|---|
The check
Firewall negative control: a planted 2099 record entered the same pipeline, was rejected as later, and did not move the answer.
Uncertainties
- The source conflict between 3 July and 8 July for SB 52-37 is unresolved.
- The 20-document manifest is a closed set assembled from named official records. It is not every communication that existed.
- The false-negative rate of factory completion records is unknown.
- Foreign adoption and enforcement of a hypothetical FAA order are not established.
- Operational and economic costs of a hold are absent.
- The reproduced private Convair memorandum has no explicit reuse licence in the congressional package. This page paraphrases it.
Free choices
- The cutoff is end of the printed day in California because no memo time survives.
- The frozen rule requires two independently necessary protections.
- Missing proof counts as refusal under the proof-before-service rule.
- The 90-day display offers both inclusive and strict boundaries instead of choosing silently.
- Aircraft-days begin at one of the two documented bulletin dates and end at listed completion.
- The counterfactual paths separate FAA authority, Douglas delivery control, and foreign adoption.
Refusals are results
Bad or missing dates go to the visible undated bucket. An FAA-only rescue is refused because jurisdiction is unproved. A paperwork-only interlock rule reports that TC-JAV would have passed the record check. No passenger-miles or accident probability is inferred.
Primary record and reuse
Sources checked against the files
- U.S. House subcommittee print, December 1974, especially printed pages 11 to 17 and 42 to 43. It reproduces the memo account and the 39-aircraft completion table.
- U.S. Senate committee print, especially printed pages 18 to 33. It records the voluntary agreement, status letters, and a reproduction of SB 52-37 dated 3 July.
- NTSB Windsor investigation CHI72AC099 and the issued A-72-97 and A-72-98 letter. The letter says adopted 23 June and issued 6 July 1972.
- Federal Register, 2 April 1974, pages 11992 to 11993. It identifies the bulletin dates, telegrams, issue date, scope, and effective date.
- French Commission final report, February 1976, especially PDF pages 4, 66 to 67, 80, and 101 to 103. It supplies the occupant anchor, missing bearing, rigging, and failure chain.
U.S. congressional, NTSB, and Federal Register texts are federal government records. The underlying private memorandum's reuse status is unconfirmed, so only a short paraphrase is used. BEA permits noncommercial and pedagogical reuse with integrity, meaning, origin, and publication date preserved; commercial reuse requires permission. No source PDF is redistributed here. The checked transcription and verification notes live in research/the-warning-before-flight-981/.