I am the pilot and owner ofAT3 19-4226 which was consumed by fire at Somersby 11 Feb 2021, and I am submitting this report to clarify and correct some commentary which has been made about the event.
the engine run-up was standard for a Rotax 912 i.e. 4000rpm, not 'a full power runup' as suggested on 20 April
the engine did not 'cut out' during the 'engine run' - it was shut down after the throttle had been retarded to the idle setting
the pilot did not 'manage to scramble to safety...after the engine had burst into flame' - once the engine was shut down, and the pilot noticed smoke observed coming from the engine cowl he vacated the airplane. Fire commenced several minutes later.
Other points:
a thorough pre-flight inspection, had been conducted....including engine & airframe, all items including all hoses, wiring, battery, etc secure and in good condition; hydraulic fluid, coolant & oil levels all at correct levels.
All maintenance requirements were met and the airplane was in excellent condition.
the airplane had 48 litres fuel ( 98 octane car fuel ) at start-up.
The engine started normally, and all indications ( oil temp, oil pressure, coolant temperature etc ) were normal.
10 minutes after start the oil and water temperatures were warm enough for an engine run-up....
the ignition check: engine smooth at 4000rpm, ignition drop 125rpm each ignition system ( 300 allowed ).
the idle check: Upon reducing throttle to achieve 3000rpm the engine ran slightly rough...reducing further to 2500 rpm then 1800rpm the engine ran very rough. Back to 2500rpm, still rough so I closed throttle, the engine on verge of cutting out so I switched off both ignitions, and switched off electric master switch. Engine stopped.
I thought, momentarily, that there may be a fuel supply problem, hence rough operation.
I then looked ahead at the stationary propeller and noticed that light, thin smoke was emanating from front of engine cowl.
I opened the canopy, released my harness, got out of plane and hurried to the engine cowl, noting that smoke only coming from the front of the engine ( note: airplane parked into wind, strength 5-8 knots ).
looking inside the engine cowl, from the rear inspection hatch, there was no evidence of smoke behind the engine, or any flames. Looking at the nose again I cd see that the smoke was thickening and darkening in colour and rising from the air inlets mounted on top of the engine cowl.
( I did not have a fire extinguisher in the airplane; had I had one I would have fired it into the engine cowl at this time ).
I rushed back to the cockpit, and pulled the fuel shut-off valve lever ( the valve is below the fuel tank, which is ahead of the instrument panel, but behind the firewall and the engine bay ).
However, once the engine had ceased rotating there cd be no fuel pumped by the engine-driven pump, and the electric pump had been switched off, prior to the engine run-up. So there had been no pressure feeding of fuel into the fuel lines once the engine had been shut down.
I then walked back to the front of the airplane, noted the smoke becoming more dense, then in an instant, flames erupted from the nose cowl inlet.
The flames rapidly expanded, consuming the composite engine cowl ( top & bottom halves ), the coaming above the instrument panel, the fuel tank aft of the firewall, the perspex canopy and then the combustible materials in the cockpit.
After 10-15 minutes the metal keel of the fuselage melted and the nose fell forward due the weight of the engine; later the centre-section and mainwheel attachments melted and the airplane fell onto its belly.
The fire burnt for 30 minutes ( the Fire services arriving 20 minutes after they were called, reasonable considering then 25 kms they had to travel ).
After all combustible materials ( fabric, fibreglass, perspex, carbon-fibre, 48 litres 98 octane fuel ) were consumed, the fire extinguished itself. What remained was a pile of ash with metal wing-tips and the metal tail sticking out !
The engine was charred and rendered unusable.
The firewall ( aft of engine bay, ahead of the fuel tank ) was scorched but intact
The battery was not Lithium-Ion, and had been replaced approximately 12 months before the accident. It exploded during the fire, as did two of the three tyres.
Possible causes of the fire ?
An electrical problem ? - such as ignition short-circuit / spark / failure ? Would have had to be instantaneous as engine only ran roughly for less than 30 seconds after engine run-up and before shutdown ( engine guage indications remained normal throughout ). Fire fuelled by airplane fuel , oil, combustible materials ? Probably the engine ran rough due to fuel disruption..... possibly a fuel hose rupture, a fuel hose coming loose or a carburettor flooding due some internal jamming ( note: all hoses within 5-year life, were well-secured and were in excellent condition ).
Note that each carburettor had a drip tray, with a drain tube secured on the nose wheel leg at the bottom of the engine cowl.
Unfortunetly, no rubber item, including hoses, survived the 30 minute fire, so they cannot be inspected to test hypotheses. However, the left hand carburettor was mostly intact wheras the right hand carburettor was partially melted.
So, what did I learn from this accident ?
Airplanes burn very quickly ( due availability of large quantities of fuel, and possible 'supercharging' by forced airflow, if airborne ).
Airplane fires are rare, so existing designs and mechanical practices are obviously successful in ensuring safety.
Install a fire extinguisher accessable in the cockpit. It is not required by regulation and weighs several kilograms, but its use could extinguish a fire or at least, forestall its spreading.
Have all fuel & oil hoses sheathed in fire-resistant materials.
If the engine cowl is of flammable material, line the interior surface of the engine bay with fire-resistant material, if allowed by regulation.
Possibly an engine bay temperature monitor / continuity wire to detect fire ? Could be difficult to achieve, without extensive experiment; might cause unnecessary engine shutdowns, but if developed to a high level of reliability and accuracy, could give early fire warning to the pilot. So, worthwhile.
Fuselage-mounted fuel tanks possibly pose a fire risk to the pilots & passengers in an accident and occupy fuselage space which could be used as stowage areas for luggage or equipment; wing-mounted fuel tanks are well away from the cabin and the engine ( on single-engined airplanes ), and also unload the wing spar, so are better from the structural or safety point-of-view ( see item 2 above ).
In summary, this fire was unexpected ( as the airplane was in excellent condition ) and unusual ( as the Rotax engine and LSA's in general have a very good safety record ). It is not possible to conlusively ascertain the cause of the fire.
AT3 destroyed by fire just before takeoff Somersby Airfield
in Aircraft Incidents and Accidents
Posted
I am the pilot and owner of AT3 19-4226 which was consumed by fire at Somersby 11 Feb 2021, and I am submitting this report to clarify and correct some commentary which has been made about the event.