enginehistory.org

[hfriedman]

Pipes, Hoses and Fires-Follow- up

This posting on this site is justified by the supposition that all happenings forward of the firewall belong to us.
At 8:00 Am on September 1, 1920 the mail plane on it scheduled run from New York to Chicago limped past Morristown, NJ, at an estimated 1,000 feet. Spectators attracted by the sound of its BMW IIIa engine backfiring “with an ominous thundering noise,” watched it dive into the ground trailing flames.
What made the accident notable at a time of many crashes was that the pilot, Max Miller, killed along with his mechanic, was a star of the airmail service assigned that particular airplane because of his trouble free record .
The airplane was one of the US Post Office’s eight Junkers F-13s bought for $20,000 each and called J-L-6s in the US to deemphasize their German origin. Their all metal cantilever winged design, enclosed cabin and large carrying capacity and range was expected to revolutionize the airmail service, permitting nonstop flights between New York and Chicago.
Then on September 14th another of the Post Office’s F-13s exploded after a forced landing near Pemberville, Ohio, killing Walter Stevens, another highly experienced mail pilot. Witnesses reported that prior to the accident the engine “sounded as if the machinery were disabled.” On February 9, 1921 two more airmail fliers were killed. The newspapers reported: “As the plane came down to about 500 feet an explosion was heard. Then a second blast was heard, and the machine took a sudden nosedive and burst into flames.”
All told, five of the Post Office’s eight F-13s were destroyed by fire with 6 killed, and the airmail service grounded the F-13. The type would go to on help launch airlines throughout the world but in the US it was back to the DH-4 open cockpit biplanes.
Explanations

The explanations are many, varied, and mutually inconsistent :

Antony L. Kay in Junkers aircraft and Engines writes that: “It seems that the [fire} was caused by a modification . . . in which a fixed fuel pipe was replaced with a flexible one.”

The EADS website flatly states, “the high-octane [sic] U.S. gasoline corroded the rubber collars of the fuel lines and thus caused fires and explosions.”

Another website notes that “[T]he fact, that the aircraft were flown with petrol instead of benzol in the U.S.A. was pointed out as a possible reason for the accidents.”

Herschel Smith wrote in Aircraft Piston Engines: “In Germany, the use of synthetic fuel, and perhaps of an early form of synthetic rubber, allowed the use of rubber tubing for fuel lines. Those sent here though, employed the then-standard copper tubes. The vibration of the big six in a metal airframe caused fatigue failures (then sometimes called ‘crystallization’) of the copper, and in flight fires resulted.”

Dean C. Smith, an airmail pilot of the time, wrote in his 1961 autobiography: “Unfortunately the engine installation and fuel system proved totally unsuitable and had a habit of catching fire. Their engines used benzol which would not only eat through our rubber hose connections but would freeze at our common winter temperatures and clog the lines and strainers with ice.”

In sort, the accidents happened because rubber was substituted for metal but also because metal was substituted for rubber- because benzol was used and also because it was not used.

The BMW IIIa was the engine powering the late marks of the Fokker DVII but its operation did require, at first, a little care. Using a 20% benzol fuel it could be operated at the then high compression ratio of 6.4 but only at altitude. Bill Gunston opines that if the pilot used full power at sea level “the engine would have blown apart.” He notes that airplanes fitted with that engine sometimes had three throttles to be opened sequentially during the climb out. Could some of the accidents have resulted from pilots, unfamiliar with their new mounts, over boosting the engine low down?. Could the airmail service have been tring to operate these high compression engines on straight gasoline with resultant distructive detonition and pre-ignition?