The trouble with British Aircraft pre 1956 (including the Valiant) was the double whammy of building aircraft with a design method that could not guarantee safety in a catastrophic failure [Safe Life] with materials [like DTD683] that could fail catastrophically. This fact was coupled with the fact that not all the metallurgical data was known when these aircraft entered service, for example it wasn't known until 1968 (4 years after the Valiant was scrapped) that for DTD 683, water or water vapour increased crack growth rates by a factor of 10. This was due to the Oxygen in the water rapidly oxidising the "fresh" aluminium releasing Hydrogen at high pressure which was enough to cause the crack to grow exposing more un-oxidised aluminium so producing more high pressure hydrogen and so on, in an auto-catalytic process.
Due to a lack of money cost cutting methods like using samples that had already been used in a stress test to measure fatigue strength led to an over estimation of the fatigue strength of a material by a factor of up to 100. This was due to the fact that pre-stressing these materials increased their fatigue strength. (
See Section 4 para 2 of this ARC Paper). In fact this method (pre stressing) was used in a failed attempt to improve fatigue resistance.
This was aggravated by the lack of consideration for fatigue failure which was rare in the older materials.
DTD683 was known to be a troublesome material at least as early as 1951. Here is Mr Gardner of Vickers explaining the problems
structural problems | flight structural | structural efficiency | 1951 | 2503 | Flight Archive
Mr. Gardner then turned to a consideration of materials, first
touching on the newer aluminium-zinc-magnesium alloys,
D.T.D. 363 and D.T.D. 683, which, used as extrusions and
forgings, made appreciable weight-saving possible.
two difficulties had arisen: (a) distortion after machining, and
(b) variation of strength across the section. The distortion problem
with this alloy had become of general importance...
... this was an unsatisfactory aspect of the new alloy.
The second effect, which gave low core properties, was one which
needed to be known before design-values for the material were
agreed.
And Mr Black of Vickers Supermarine in 1953
light alloys | cold bending | permanent distortion | 1953 | 0935 | Flight Archive
The alloys considered were those
of specifications D.T.D. 363A and D.T.D. 683....
Mr. Black stated, the increased strength
of the materials was accompanied by a lowering of ductility as
measured by the elongation obtained from a tensile test. A low
elongation value was undoubtedly undesirable in an aircraft
material, because only small amounts of permanent distortion
could take place before a fracture occurred, and large amounts of
cold work could not be withstood without fracture.
... the alloys had a reduced capacity to
absorb permanent distortion, a low ductility and a low ratio of
fatigue strength to tensile strength ...
... a high normal stress level in use, and the result was a
greatly increased sensitivity to stress concentrations resulting
from bad design or surface notches ...
... residual internal stress would reach a high level, and was very undesirable.
So the problems were well known, but instead of giving up with the materials they soldiered on thinking the problems were 'solved' after all the Ministry of Supply told them they had to use this material and the UK Aircraft Industry was pretty arrogant about its ability to solve problems having produced war winners like the Spit' and Lanc'
But the Industry had not solved all the problems and it began to dawn in 1955 that their approach was flawed. This is well evidenced in the technical papers of the time available in the Aerade catalog. The research switches from attempts to improve fatigue resistance to research into understanding the process of crack formation, growth rates and ways to stop the failures from being catastrophic.
Airframe Fatigue 1955
On a comparative stress
basis, the new alloys such as DTD.363, 364 and 683 had no
better fatigue properties than the earlier alloys. Thus, for structures
of equal static strength, a reduction in fatigue life occurred,
and one example given by Rhode showed about a fivefold reduction
in fatigue life in transferring from 24 S-T to 75 S-T
(75 S-T is the American designation for DTD683)
The problem of working out a Safe Life was virtually impossible given the large scatter in the fatigue data
cycles | endurance limit | salt spray | 1955 | 0363 | Flight Archive
... tests on 57 specimens give lives ranging from 430,000 cycles to 117,423,000 cycles with a mean of 23,324,000; such results underline the magnitude of the problem ...
Things came to a head in 1956 with this lecture given by a Lockheed structures engineer to the RAeS
In it he delivers the "coup de grace" to Safe Life and states that DTD683 was the worst choice for fault tolerant structures
1956 | 0396 | Flight Archive
Putting this another way the big question is: —
. . . . . . . . .. Laboratory (or predicted or recorded) life...
Safe Life = ----------------------------------------—--------------
.. . . . . . . . . . . . . . . . . . . . . ? .
On the determination of this ? factor hinges the adequacy of the
safe-life method.
Given the scatter in the data above, the nominator [Laboratory (or predicted or recorded) life] in the above equation was also largely guess work. Effectively the equation above becomes.
.. . . . . . . . . . ?
Safe Life = ----------
.. . . . . . . . . . ?
You don't need to be a mathematical genius to see the problem using an equation like that to design 'safe' aeroplanes.
All of this was made even more problematic with the publication also in 1956, of this paper in the Journal of the Institute of Metals by a Birmingham metallurgist team, which condemns DTD683 as it was too unstable (hence the wide scatter in the data).
Journal of the Institute of Metals
Mr Gardner of Vickers had noted the unstable nature of the alloy in his 1951 lecture.
The lack of stability shown both in extrusions
and forgings was an unpleasant feature in production.
DTD683 was removed from use 3 years later; 1959, 5 years before the Valiant was. Its use post 1956 was limited to components in compression, such as under-carriage components.
In 1956 several things happened many Safe Life designs built with the new alloys were either scrapped or re-designed like the Argosy or Shackleton, but not the Valiant. Also in 1956 Vickers began flying a Valiant deliberately into turbulence they measured the strains and the experiments produced the disturbing result that the Valiant fleet had a remaining Safe Life of 70 hours, later revised to 300 hours (presumably under some pressure from MoD and Whitehall Mandarins). The problem with low flying was the increased frequency of gusts that would exceed the limits on the airframe. Exceeding the limits was potentially the initiating event for a later fatigue failure, as described in the Birmingham paper and as implied by Mr Black of Vickers Supermarine in his 1953 talk.
... only small amounts of permanent distortion could take place before a fracture occurred.
Also in 1956 Macmillan said the following in a memo to PM Eden about defence expenditure.
"When the story of the aeroplanes finally comes out it will be the greatest tragedy if not scandal in our history"
1956 was the year in which the Government launched a review of the Aircraft Industry, which later resulted in the industry's restructuring. The last 6 Valiants were cancelled in 1956.
According to Flight magazine (corroborated in Eric Morgan's book) only 50 of the 104 (108) Valiants were still in service when the scrapping order came in 1965. Using the (incomplete) data available the average life of a Valiant was 7.6 years with an average of just over 300 flying hours per year each.
In todays money (price of a Valiant in 1956 = ~£500,000) and using a back of the envelope calculation, the cost to the Tax payer was ~£154,000 per hour of Valiant flight time, that number excludes the actual running costs such as wages, fuel etc and just uses the rough purchase cost.
That was Macmillan's "scandal". The "tragedy" was the number of fatalities in non-combat accidents, which reached a peak of nearly 1 a day in 1954.