|
|
||
|
|
|
|
From Farnborough AAIB Bulletin No. 2/96, Ref: EW/A95/6/1, Information Source: Farnborough, UK, AAIB Field Investigation
"The failure had developed from a 100 mm (4 inches) long region of pre-existing circumferential cracking in the membrane gore panel adjoining the inboard edge of the doubler. The rupture subsequently propagated as rapid tears from either end of this pre-existing crack region. The rapid tears initially ran in an approximately circumferential direction before turning radially downwards (ie toward the centre of the bulkhead), due almost certainly to stiffness changes in the gore panel itself as the tears propagated, and the influence of radial frames and stiffeners. Having started to run radially inwards, toward the centre of the dome, both tears were then arrested by the first of the circumferential anti-tear straps which they encountered."
Proposed aft pressure bulkhead failure sequence (From AAIB Bulletin No. 2/96, Ref: EW/A95/6/1):
"The absence of any significant corrosion pitting or other surface defects at the fatigue initiation sites, together with the presence of fatigue origins on both sides of the bulkhead, suggested that the fatigue cracking had been caused by stress reversals of a kind which would not normally occur on a pressure bulkhead of this type. In this connection, the presence of an oil-can type deformation [nonlinear "snap-through" type of buckling] immediately outboard of the fatigue initiation region was considered significant, insofar as it provided a mechanism whereby the bulkhead in the affected region would tend to adopt an 'S' shaped cross-sectional profile when relaxed (unpressurised), changing to a convex profile when pressurised. Such flexural changes would cause a ripple to run across the adjoining area as the oil-can deformation 'popped out ' under pressure loading, which in turn would have created stress reversals in the gore panel of the kind implicated by the fatigue growth. On the available evidence, therefore, the oil-can (nonlinear buckling) deformation of the bulkhead appeared to offer the most plausible explanation for the fatigue cracking which subsequently led to the [local] bulkhead rupture."
Page 124 / 256