FOD-Detection.com

FR1879 was fatigue, not FOD. The fracture surface settled it.

The preliminary report landed on 13 August, and it settles the question this site was left holding three weeks ago. Blade No. 10 of the right CFM56-7B26 fractured through the dovetail in its root. About 37 per cent of the fracture surface carries the smooth, curving features of fatigue. The origin sits on the concave side of the blade near the trailing edge, roughly 14mm out from the root end face, and the ratchet marks there indicate not one crack but several, initiating along some 16mm of the contact face. Beside them, heavy fretting on the dovetail and on both faces of the shim it bore against.

Under the electron microscope the striations run from 0.2mm deep out to the boundary at 15.9mm, coarse and unchanging the whole way — which the NTSB reads as high amplitude fatigue: cyclic loading at frequencies above one per flight, at stress amplitudes above the ordinary high-cycle case.

None of that is a foreign object. A blade struck by debris carries an impact site, and a fracture that runs from it — ingestion leaves a mark you can point at. This blade was cracking from inside its own root fixing, in several places at once, for long enough to leave a fatigue region 16mm deep. The debris explanation is no longer the leading one.

We said in July that the fracture surface would say which it was. It has said.

What makes the FOD reading understandable — and this is worth conceding plainly — is the timing. The quick access recorder has vibration on that engine jumping suddenly near the end of the takeoff roll, from about 0.2 cockpit units to 1.7 before the aircraft left the ground. It drifted to 2.0 over the next three minutes, then climbed to 4.3 over the four minutes after that, at which point the crew reduced power and called for the high vibration checklist. When the power went back up, it reached the recorder’s 5.0 ceiling in nineteen seconds, and the bang followed. Something did change on the takeoff roll at Thessaloniki. O’Leary was right about when. He was wrong about why: that is the signature of a crack reaching critical length under takeoff power, not of something being swallowed.

The finding that should worry people has nothing to do with runways. Those fan blades were inspected ultrasonically under CFM service bulletin 72-1033, which exists to look for precisely this — crack indications in the dovetail, concave and convex faces. They were done on 11 November 2025, and again on 24 May 2026, two hundred and fifty-three cycles before the accident flight. Both times, no findings.

That service bulletin was issued in 2018, after a fan-blade-out that killed a passenger. The inspection regime written in response to the last one looked twice at this blade and saw nothing.

There is a foreign object in this file, and it is not on the runway. Flight crews reported four suspected bird strikes to this same No. 2 engine in the twelve months before the accident, with remains found in two. Investigators recovered bird remains from the oil cooler face, from a thrust reverser blocker door drag link and from the fan case, and lifted fluorescing organic traces off five more blades under black light. All of it has gone to the Smithsonian’s feather identification laboratory. Whether any of it bears on the fatigue is not established, and the report is careful to say so.

It matters here only for the vocabulary. Birds are foreign objects; they are not FOD in the sense this site has used the term for sixteen years, which is debris lying on a paved surface that an inspection or a detection system could have found. No runway scanner detects a bird at five hundred feet. The industry has always known this, which is why the bird problem was sold as a separate product — Xsight split it out in 2013, and Moog bolted avian radar onto Tarsier in 2022. When O’Leary said foreign object damage on takeoff at Thessaloniki, he was pointing at the airfield surface. The foreign objects in the file arrived through the air, across a year.

One further finding is being reported in a way that needs care. The report notes three airworthiness directives — 2025-04-01, -02 and -03, covering fastener enhancements on the inlet cowl, fasteners and an external doubler on the fan cowl, and additional brackets on the exhaust nozzle — none of which had been incorporated on this engine. Several outlets have run that as an aircraft flying without the modifications designed to prevent this very accident. The compliance date for all three is July 2028. The aircraft was not overdue, because the deadline has not arrived. Whether two more years is the right allowance for a set of directives aimed at containing blade-out damage is a fair question, and one worth asking. Whether this operator was out of compliance is not a question: it was not.

Corrections to our 25 July piece. Two errors, both ours. We wrote that North Macedonia’s authority was leading, reasoning that the engine let go in its airspace; the report places the event near Polykastro, about four and a half nautical miles inside Greece, and records that Greece’s HARSIA delegated the investigation in full to the NTSB on 16 July. We also called the failure uncontained. The NTSB found no radial uncontainment through the engine cases — the fragments left through the inlet cowl, punctured in two places, rather than through the containment ring. Fan-blade-out with cowl penetration is the accurate description. Our inference from a 39-day ground stop at Kaunas is superseded by the inspection record above, which is the real answer to the servicing question and a good reminder not to read maintenance history off an aircraft’s parking pattern.

So the answer to the question we asked three weeks ago — what was on the runway at Thessaloniki that morning, and was anything watching it — is that there was nothing on the runway, and no detection system built in the last two decades would have altered this outcome. That deserves stating as plainly as the accusation was.

The term was reached for as a shield, and the metallurgy has taken it away. What remains is a blade that cracked in its root fixing over many cycles, an inspection written after the last fatal case that looked twice and passed it, and a set of containment modifications not due until 2028. None of it is a FOD story.

One caveat the report itself insists on: this is preliminary, the investigation is open, and what initiated the cracking is not established. Bird ingestion over the preceding year has not been excluded as a starting point — which would make the origin a foreign object, though still not one on a paved surface, and still not one any runway system could have found.

The industry this site covers should take the lesson anyway, because it is the one that keeps recurring: foreign object damage is the cheapest explanation on the shelf, and it is the first one reached for when the alternative is a maintenance file.

Sources (NTSB preliminary report DCA26FA274, Airliner World, Paddle Your Own Kanoo)