Page 206 - Failure Analysis Case Studies II

191
rear corner of the rear ADF window on Yoke Peter, and that this grew to failure after only 1286
flights.
It has been reported that Mr Ronald Bishop, the Chief Designer at de Havilland during this
period, felt that the mistake made was to allow rivets and bolts to be used to assemble the windows
and reinforcements onto the aircraft skin. Other parts of the aircraft were glued using “Redux”,
but the tooling required was thought to be too difficult to achieve, and too expensive for these cut-
out areas. Other riveted areas, some wing skin scctions for example, were known to be susceptible
to fatigue crack growth from the rivet holes, and the use of riveting to fix such thin-section aluminium
sheet in the vicinity of cut-outs was probably more damaging than the shape of the windows. In
fact, none of the cracks in the body or wings of the test Comet emanated from the cut-outs directly,
but came from rivet or boltholes near cut-outs, and the initial failure site on Yoke Peter was from
a bolthole rather than the edge of the ADF window.

4. CONCLUSIONS

The de Havilland Comet was a truly novel aircraft. It had a number of new features which are
now accepted as part of modern aircraft design, but at the time set a completely new trend. A
number of technical advances had to be made to enable the aircraft to fly, and these stretched the
scientific knowledge of the time to the limit. However, as with all pioneers, the first to enter a new
field are the first to encounter the problems, and this is especially so in commercial aviation, where
failure can be spectacular and high-profile.
The failure of the pressure cabin was due to fatigue crack growth from defects which were
probably present from the construction of the aircraft and had not been a problem in earlier designs
of aircraft, as the required cabin pressure had been lower. That this problem was not detected by
the rigorous testing undertaken by de Havilland was probably due to an unfortunate set of cir-
cumstances with regard to the order in which the tests were performed, and could not easily be
foreseen at the time. The knowledge gained from these unfortunate accidents enabled scientific
knowledge to advance, and testing procedures to be instigated which ensured the increased safety
of future civil aircraft.
All the observed cracks in the pressure cabin [l, 21 emanated from bolt or rivet holes near the
cut-out areas. It was probably not the shape of the cut-outs that was so damaging to the fatigue life
of the cabin, rather the method of fixing the windows and doubler plates onto the pressure cabin.
Had the windows not been square then the “Redux” glueing method might have been applied to
these areas, and the failure avoided.
After the problems of the Comet I, de Havilland produced the Comet IV, which was larger,
carried 80 passengers, and had a greater range. This aircraft entered history as the first commercial
jet aircraft to cross the Atlantic on 4 October 1958, and inaugurated a route which has carried many
millions of passengers since. However, 3 weeks later, a Pan American Boeing 707 flew the same
route carrying 120 passengers, and indicated the supremacy of the American airline industry. The
Comet continued to be built until 1962, by which time 113 had been made, showing the quality of
a design commenced in September 1946, and has entered history as the first commercial jet airliner
and the first to operate a scheduled service across the Atlantic.

REFERENCES
I. Cohen, Baron L. of Walmer, Farren, W. S., Duncan, W. J. and Wheeler, A. H., Report of the Court of Inquiry into the
Accidents to Comet G-ALYPon 10 January, 1954 and Comet G-ALYY on 8 April, 1954. HMSO, London, 1955.
2. Royal Aircraft Establishment, Report on Comet Accident Investigation, Accident Report 270. Ministry of Supply, London,
1954.
3. Green, A. E., Proceedings ofthe Royal Society A, 1945, 184,231-252.
4. Material Specification, Alurniniumcoated high tensile aluminium alloy for sheet and coils. DTD 546B, Ministry of
Supply, HMSO, London, 1946.