Page 201 - Failure Analysis Case Studies II

186
airliner, and propelled civil aviation into a new era. The de Havilland DH106 had been conceived
in 1943 by Sir Geoffrey de Havilland, and design work had begun in September 1946. The prototype
first flew on 27 July 1949, by which time agreements to supply 14 aircraft to the British Overseas
Airways Corporation (BOAC) and two to the Ministry of Supply had been signed. On entering
service, the aircraft could carry 36 passengers at a cruising speed of 450 mph (200 m s-I), with a
range of 2500 miles (4000 km).
To enable the payload to be sufficiently large for commercial viability, the weight of the aircraft
and fuel had to be kept to a minimum. The construction techniques used were a mix of old and
new, rivets being used in certain areas as well as a method of glueing the aircraft skin and stringers,
called “Redux”. This new technique had been pioneered by de Havilland, in the Hornet and Dove
aircraft, to reduce the weight of the structure whilst maintaining the strength. The power for the
aircraft was delivered by four Ghost turbofan engines built by the de Havilland Engine Company
Limited. To enable these engines to run as efficiently as was practicable, this aircraft was expected
to fly at 40,OOOft (10.7 km), or double the cruising altitude of the then commercial airline fleet. At
this cruising altitude, the passengers and crew require an artificial oxygen supply, and it was decided
to pressurize the cabin at the equivalent to a comfortable 8000ft (2.4km), which gave a pressure
differential across the aircraft skin of 8.25psi (56kPa) at cruising altitude. This was double that
which had been previously employed, and de Havilland conducted many tests to ensure the integrity
of the cabin.
As well as the four turbofan engines, there were a number of other new features, including high-
pressure refuelling, the hydraulic actuation of the control surfaces, and an air-conditioned cabin,
which altogether made this a completely new aircraft.
The Comet I was seen as the new hope of the British aircraft industry, but a number of crashes
tarnished the image of this graceful airliner. There were a number involving take-off, which culmi-
nated, on 3 March 1953, in the death of the crew delivering Comet CF-CUN to Canadian Pacific
Airlines. These were ascribed to the unfamiliarity of the pilots with the new aircraft. The mid-air
break-up of Comet G-ALYV 50 km north-west of Calcutta, exactly 1 year after the inaugural flight,
was found to be due to excessive stresses in the airframe due to a tropical storm in the area.
However, there then followed two accidents under similar conditions in the space of 3 months,
which could not be so easily explained. The first of these was on 8 January 1954, and involved
Comet G-ALYP (Yoke Peter) approximately half an hour after take-off from Ciampino airport in
Rome bound for London on the last leg of a journey from Singapore. Yoke Peter was climbing to
27,000 ft (8.27 km) in good weather conditions when it was seen to crash into the sea near Elba in a
number of pieces, some of which were in flames. The Comet fleet was grounded, and the possible
causes examined, a process which was not assisted by the inspection of the wreckage, as most of
this was on the seabed at the time. A number of recommendations were made, resulting in improve-
ments to the Comet I, and the fleet re-entered service on 23 March 1953.
On 8 April 1954, Comet G-ALYY (Yoke Yoke) took off from Ciampino airport bound for Cairo.
After approximately 30min, when Yoke Yoke would have been reaching the top of its climb to
35,000 ft (10.6 km), all contact was lost, and wreckage was later found in the sea near Naples. The
operator of the Comets (BOAC) again withdrew all Comets from service, and on 12 April the
Ministry of Transport and Civil Aviation removed the Certificate of Airworthiness from the Comet.

2. THE INVESTIGATION
Following these accidents, the Secretary of State for Civil Aviation requested a full investigation
into their causes by the Royal Aircraft Establishment (RAE) at Farnborough, and a Court of
Inquiry was established [I]. This investigation encompassed a number of lines of approach, but two
aspects of particular interest are the reconstruction work on G-ALYP (Yoke Peter), and the
accelerated simulated flight testing of Comet G-ALYU (Yoke Uncle).
Comet Yoke Uncle had beem obtained from BOAC after flying for 3539 h and undergoing
1221 cabin pressurizations [2]. The accelerated simulated flight testing took the form of cabin
pressurization using water, and wing loading using hydraulic rams (Fig. 2). Water was chosen to
pressurize the cabin as it is reasonably incompressible, and any failure would not result in the
complete loss of the pressure cabin due to the stored energy. If air had been used, any failure of the