Under 14 CFR 33.76, turbine engines must ingest birds and keep running — no uncontained failure, no loss of safe flight. The large-bird standard uses a single bird weighing between 1.85 and 3.65 kilograms, sized to the engine’s inlet throat area.
Since 2023, turbofans must also pass climb and approach flocking-bird tests. Airframe parts like windshields and tails fall under separate, far less publicly documented rules — and whether the U.S. facility most associated with firing real carcasses still fires at all is unconfirmed.
The cannon gets the headlines. A compressed-air gun that fires chicken carcasses at aircraft hardware is the kind of detail that travels well, and it has been traveling since at least the 1970s, when a facility at Arnold Air Force Base in Tennessee started using a barrel salvaged from an 8-inch naval gun to hurl birds at military canopies.
Here’s the correction most retellings skip: the method is older than the facility, and its earliest uses were about airframe parts — windshields — not engines.
That distinction matters, because the modern version of the question is a regulatory one. What a traveler actually boards is an aircraft certified under a rulebook that specifies bird weights, impact speeds, and pass/fail outcomes — and that rulebook is far more detailed on engines than on the parts of the airframe a bird can also hit.
The gap between what the rules require and what the public record shows is the real story here.
What a bird strike engine test actually requires
Start with the provision doing the heavy lifting. 14 CFR 33.76 governs turbine-engine bird ingestion, and its core demand is blunt: an impact cannot produce an uncontained failure or a loss of safe flight.
For the large-bird case, certification uses a single bird. Its weight is not fixed — it scales with the engine’s inlet throat area, landing somewhere in that 1.85 to 3.65 kilogram band. A widebody engine with a bigger mouth takes a bigger bird.
The rulebook changed recently. FAA amendments finalized in 2023 added flocking-bird core-ingestion tests for climb and approach, so a turbofan has to prove it can swallow birds into the core during the phases where flocks are most likely to be met. The climb test runs at 261 knots true airspeed, according to the Federal Register rulemaking, with one bird at the highest weight the applicable table assigns to that engine’s inlet area.
Manufacturers don’t always fire a real carcass.
Paragraph (a)(5) lets an object the Administrator accepts stand in, and FAA advisory circular AC 33.76-1B — issued April 3, 2023 — sets the bar: the substitute must match a real bird’s mass, shape, and density, and reproduce how it behaves on impact.
The retrieved FAA material, however, contains no current numeric dataset that would show how a synthetic bird’s behavior compares with a real carcass. The allowance exists; the public equivalence evidence is thinner than the rule implies.
| Authority | Regulation or section | Component/test | Bird mass | Impact speed | Pass/fail criterion |
|---|---|---|---|---|---|
| FAA | 14 CFR 33.76 | Turbine-engine bird ingestion; large bird | 1.85–3.65 kg, depending on engine inlet throat area | Specified by the applicable engine-ingestion test; exact large-bird speed not independently verified here | No uncontained failure or loss of safe flight; exact paragraph-specific outcome varies by test category |
| FAA | 14 CFR 33.76(e)(2)(i) | Climb flocking-bird core-ingestion test | One bird at the highest weight specified in the applicable table for the engine inlet area | 261 knots true airspeed | Bird must be directed for maximum core ingestion; the engine must satisfy the applicable continued-operation requirements |
| FAA | 14 CFR 33.76(e)(2)(ii) | Approach flocking-bird core-ingestion test | One bird at the highest weight specified in the applicable table for the engine inlet area | Not independently verified from the retrieved text | Applicable approach-condition ingestion requirements apply; exact performance details are omitted rather than inferred |
| FAA | 14 CFR 33.76(a)(5) | Substitute object or artificial bird | Must represent the applicable bird | Must reproduce the applicable impact condition | Object must be accepted by the Administrator; FAA guidance requires simulation of mass, shape, density, and impact effects |
| FAA / FAR Part 25 as summarized by EASA | FAR Part 25 airframe provisions | General transport-airplane structure | 4 lb | Cruise speed, VC | Structure must withstand the impact without the specified failure condition; exact component text was not retrieved |
| FAA / FAR Part 25 as summarized by EASA | FAR Part 25 empennage provision | Tail or empennage | 8 lb | Cruise speed, VC | Empennage structure must withstand the impact |
| EASA | CS-25 as summarized in EASA’s bird-strike study | Windshield/openable-window protection | 1.8 kg | VC | An openable window is required unless sufficient windshield area remains clear after the strike |
| EASA | CS-25 as summarized in EASA’s bird-strike study | General airframe structure | 4 lb | Cruise speed, VC | General structural requirement applies; CS-25 has no separate empennage clause in the comparison |
| Source: Federal Aviation Administration (FAA); European Union Aviation Safety Agency (EASA); National Transportation Safety Board (NTSB) report repository | |||||
Two rows carry most of the weight. The climb flocking-bird test is the newest and most demanding addition — a single bird at the heaviest table weight, aimed at the core. And the artificial-bird row is the one to watch, because it’s where the rulebook grants flexibility without publishing the evidence that the flexibility is safe.
The airframe side is where the public record goes quiet
Engines dominate the coverage. Airframes don’t, and the asymmetry is visible in the rules themselves.
On the U.S. side, the transport-airplane provisions summarized in EASA’s comparative study set a general structural requirement of 4 pounds at cruise speed, plus a standalone 8-pound empennage requirement — the tail gets its own clause, at cruise speed, all to itself.
Europe’s CS-25 handles the same hazard differently. There is no separate empennage clause in the comparison; the tail folds into the general 4-pound structural rule. CS-25 does add a windshield condition the study does not identify in the parallel FAR provision: after a 1.8-kilogram bird strike at VC, enough windshield area must stay clear — unless the aircraft provides an openable window.
What neither rulebook yields, in the material retrieved, is component-specific pass/fail thresholds for radomes. That’s a real gap, not a rhetorical one. If you’ve read that a 1.8-kilogram bird is fired at 250 to 500 knots at radomes with no radar loss permitted, treat it as unverified — the numbers circulating for airframe components are thinner on sourcing than the engine figures, and this article won’t assert them.
Superdeals cover departures from North America, Europe and Australia.
Flight deals most people never see
Air Traveler Club monitors thousands of routes for pricing anomalies that traditional search engines miss. With ATC Superdeals members save up to 80% on each flight vs. regular fares.
Too good to be true? Check it out yourself. It's free.
Delivered weekly. Unsubscribe anytime.
A certified engine, a flock, and the Hudson
On January 15, 2009, US Airways Flight 1549 — an Airbus A320-214 flying on CFM56-5B turbofans — met a flock roughly two minutes out of LaGuardia. Both engines lost nearly all thrust.
The crew ditched the jet in the Hudson River. All 155 people aboard got out, with four passengers and one flight attendant seriously injured, per the NTSB factual record.
Investigators attributed the initial failure to large-bird ingestion, and the survival of everyone aboard to the crew, the aircraft’s equipment, and the emergency response. It was never a certification-enforcement action, and nothing in the record suggests the engine’s certificate was voided.
That’s the narrower lesson. A certificate is not a guarantee of thrust after a bird strike; it is evidence that the failure mode was tested for.
Why the method outlived the facility that made it famous
The origin story is an airframe story. The earliest bird-impact work was about windshields — how badly they would lose against birds — and that hazard is still handled by a separate set of rules from the engine ones.
Three decades later, the technology got a permanent home. The AEDC S-3 Bird Impact Range at Arnold Air Force Base was first fired in 1972 and used to test military aircraft canopies. Over its service life it fired more than 1,000 shots.
Then the record stops being useful. As of a check on October 13, 2026, the 2025–2026 record is silent on whether the range has fired since 2007 — or whether it has been shut down, handed to another mission, or replaced.
So the facility that anchors the public imagination of bird-impact testing may not be pulling the trigger at all — and the rulebook it once served has moved on without it.
What this means for you
The practical value here is narrow and real. When you board a flight, the engines under you were certified to swallow birds during climb-out and approach — the two phases where the 2023 amendments now demand proof.
The airframe around you was certified too, but against a public evidence base that is thinner and less comparable across regulators. That doesn’t make the aircraft unsafe; it means the documentation you can find is uneven, and the radome numbers in particular aren’t there to check.
One thing to watch: whether the S-3 range’s status gets resolved. If it has gone quiet, the certification regime still functions — but the most recognizable piece of physical infrastructure behind it no longer does.
Reporting by
Key terms
- 14 CFR 33.76
- The FAA regulation that governs bird ingestion by turbine aircraft engines. It requires that an impact cause neither an uncontained failure nor a loss of safe flight, and for the large-bird case it sets the test bird’s weight at 1.85 to 3.65 kilograms, scaled to the engine’s inlet throat area. Its scope stops at the engine, so the airframe parts a bird can also strike — windshields, tails, radomes — are covered by a different set of rules.
- CS-25
- The European certification specification for large transport airplanes, the counterpart to the FAA’s FAR Part 25. According to EASA’s bird-strike study, it sets general structural protection at 4 pounds at cruise speed and adds a 1.8-kilogram windshield-clear-area condition, with no standalone empennage clause matching the FAA’s separate 8-pound tail requirement. That structural difference is why the same hazard produces different checklists depending on where an aircraft is certified.
- Flocking-bird core-ingestion test
- A certification trial in which one bird is fired into an engine’s core to prove the engine can keep running after meeting a flock. Introduced for climb and approach in the FAA’s 2023 amendments, it uses the single highest bird weight the applicable table assigns to that engine’s inlet area. Because flocks, not lone birds, are what aircraft actually meet on departure and arrival, the test targets the phases where an engine is most exposed.
- Artificial bird
- An FAA-accepted substitute object used in place of a real carcass during required bird-ingestion tests. Under 14 CFR 33.76(a)(5), it must match a real bird in mass, shape, and density and reproduce how it behaves on impact, per advisory circular AC 33.76-1B. The allowance is real, but the public record contains no current numeric dataset showing how closely a synthetic bird tracks a real one.
- Uncontained failure
- A failure in which engine debris escapes the casing instead of staying inside it. It is the outcome the engine bird-ingestion rule is written to prevent, and the certification question is binary: after a specified impact, the engine must avoid it and retain safe flight. Airframe rules take a different approach, requiring specified structure to survive a strike rather than demanding that nothing escape.
Questions? Answers.
Can a jet engine survive a bird strike?
It has to, by rule. Under 14 CFR 33.76, turbine engines must pass bird-ingestion tests where impacts cannot cause uncontained failure or loss of safe flight. The 2023 amendments added climb and approach flocking-bird core-ingestion tests, with the climb test run at 261 knots true airspeed.
How much damage can a bird do to a plane?
Enough that airframe rules exist for it: a 4-pound bird at cruise speed for general structure, and a separate 8-pound requirement for the empennage in FAR Part 25. The worst documented case is Flight 1549, where large-bird ingestion drained nearly all thrust from both engines.
Has a bird ever flown into a plane engine?
Yes — routinely, and one case is definitive. On January 15, 2009, US Airways Flight 1549, an Airbus A320-214 with CFM56-5B turbofans, struck a flock shortly after departing LaGuardia. The NTSB attributed the loss of nearly all thrust on both engines to large-bird ingestion, and all 155 occupants evacuated after the ditching.
Is the AEDC S-3 Bird Impact Range still operating?
Unconfirmed. The range was first fired in 1972 and fired more than 1,000 shots over its service life, but as of October 13, 2026, the 2025–2026 record is silent on whether it has fired since 2007 or been shut down, handed to another mission, or replaced.
Can manufacturers certify an engine using artificial birds instead of real carcasses?
Yes. 14 CFR 33.76(a)(5) allows an object accepted by the Administrator to substitute for a bird, and FAA guidance requires it to match a real carcass in mass, shape, and density and to reproduce its impact behavior. What the retrieved record does not provide is a modern dataset establishing numerical equivalence between the two.
