The purpose of the A-7D quality assessment was to establish the
manufacturing quality (ai)
of the A-7D aircraft. This was
accomplished using the Equivalent Initial Quality Method. The method was applied to a sample problem
involving an A-7A wing fatigue test failure.
Next, specimens were cut from an A-7D production aircraft and tested to
failure under a selected block loading.
The fracture surfaces were then fractographically examined and the
equivalent initial quality was established.
A photograph of the failure
area of a full-scale fatigue test of an A-7A wing was used as a sample
problem to check out the Equivalent Initial Quality Method. The wing had been subjected to a 10-level,
blocked, low-high stress spectrum.
Fractographic measurements were taken from the photograph (Figure 3.2.3), making it possible to generate a large
portion of the crack growth curve.
Crack propagation analyses were performed using the computer routine
EFFGRO and the Wheeler Retardation Model until the analytical crack growth
curve correlated well with the fractographic test data. This correlation is presented in Figure 3.2.4, which indicates that the manufacturing
quality of the test hardware at the failure location was equivalent to an
initial crack of length ai =
0.00109 in. This excellent correlation
of the analytical crack growth prediction with the fractographic test supported
the validity of the Equivalent Initial Quality Method for this particular
problem.
Figure 3.2.3. A-7A Wing Fatigue Test Fracture Surface
[Rudd & Gray, 1978]
Figure 3.2.4. Equivalent Initial Quality Results for A-7A Wing [Rudd &
Gray, 1978]
The Equivalent Initial Quality Method was next used to
establish the A-7D quality assessment. This assessment was accomplished using test
specimens cut from the lower wing skin of an A-7D production aircraft
that had been used as a gunfire target.
Because this particular aircraft had low
flight time (691.9 hours), the probability of cracking in the wings was very
low. The location of each
specimen in the lower wing skin is illustrated in Figure
3.2.5. Each specimen was made of
7075-T6 aluminum and contained multiple holes.
The geometric details for each specimen are presented in Table 3.2.1, indicating that the thickness ranged from
approximately 3/16 in. to 1/4 in. and the nominal values of the width and hole
diameter were 3 in. and 1/4 in., respectively.
The specimens contained two types of holes – countersunk holes (wet-wing
region) and straight-shank holes (dry-wing region).
The test specimens
were subjected to a fatigue stress spectrum consisting of 5,000 cycles with a
maximum stress of 20 ksi and a stress ratio of 0.1 followed by 100 cycles with
a maximum stress of 30 ksi and a stress ratio of 0.1. The block spectrum was chosen because it produced test lives of
reasonable length (less than 20 blocks) and fracture surfaces that were readily
readable.
Table 3.2.2 contains a summary of the
number of fastener holes involved, the number of flaws detected, the number of
flaws fractographically examined, the crack length range at the time of specimen failure (af), and the range of the equivalent initial quality (ai). All but two of the 44 holes contained double flaws. One of these two holes contained one crack, while no crack was
detected in the other hole. This resulted in a total of 85 flaws, of
which 44 were examined fractographically. The flaws were arbitrarily chosen for fractographic examination at
magnifications ranging from 30x to 400x using a universal measuring
microscope. The equivalent initial
quality range for all the holes was found to be 0.00015 - 0.0022 in. A statistical distribution of the A-7D
equivalent initial quality was obtained.
Table3.2.1. Geometric Details of A-7D Quality
Assessment Specimens [Rudd & Gray, 1978]
Specimen
|
Thickness a
|
Width a
|
Hole Diameter a
|
101
|
0.226
|
2.93
|
0.253 b
|
201
|
0.226
|
2.93
|
0.253 b
|
301
|
0.217
|
3.00
|
0.253 b
|
401
|
0.231
|
3.00
|
0.253 b
|
501
|
0.183
|
2.90
|
0.253 c
|
502
|
0.176
|
3.00
|
0.253 c
|
601
|
0.263
|
3.00
|
0.253 c
|
602
|
0.264
|
3.00
|
0.253 c
|
a Dimensions in inches
b Countersunk hole
c Straight-shank hole
Figure 3.2.5. A-7D Quality-Assessment Specimen Locations [Rudd & Gray,
1978]
Table
3.2.2. A-7D Quality Assessment
Test Results [Rudd & Gray, 1978]
Specimen
|
No.
Holes
|
No.
Flaws
|
Range a
|
Flaws Tracked
|
ai a
Range
|
101
|
7
|
14
|
0.05-0.75
|
14
|
0.0004-0.0022
|
201
|
6
|
12
|
<0.01-1.10
|
12
|
0.0004-0.0012
|
301
|
4
|
8
|
0.01-0.65
|
1
|
0.0003
|
401
|
3
|
6
|
0.02-0.50
|
1
|
0.0002-0.0014
|
501
|
8
|
14
|
0.00-0.60
|
1
|
0.0007
|
502
|
8
|
16
|
<0.01-0.62
|
6
|
0.0006
|
601
|
4
|
8
|
0.02-0.50
|
8
|
0.00015-0.0009
|
602
|
4
|
7
|
0.00-1.05
|
1
|
0.0006
|
Total
|
44
|
85
|
|
44
|
|
a
Dimensions in in.
The fractographic examinations revealed the origins of the
flaws for both the straight-shank holes and the countersunk holes as
illustrated in Figure 3.2.6. There is equal possibility of flaw
occurrence along the bore of the hole for the straight-shank hole, while the
most frequently occurring flaw location for the countersunk hole is at the
inside radius of the small-diameter portion of the hole. Typical flaw origins for each type of hole
are shown on the fracture surfaces of Figure 3.2.7. Also illustrated in Figure
3.2.7 is the readability of the fracture surfaces for the selected stress
spectrum, with the dark marking bands resulting from the application of the
high-load (maximum stress of 30 ksi) portion of the specimen.
Figure 3.2.6. A-7D Flaws Origins [Rudd & Gray, 1978]
Figure 3.2.7. Fracture Surfaces for Countersunk (Left) and Straight-Shank
(Right) Holes [Rudd & Gray, 1978]
Metallurgical
investigations of the A-7D flaw origins revealed that the flaws were the result
of two different sources-anodize
pitting and mechanical sources. The
majority of the flaws (86.4%) initiated from anodize pits in the
following manner. Insoluble
microconstituents were exposed along the bore
of the hole during the hole-drilling operation. The anodizing ate away the microconstituents and caused
pitting. The exposed pits were then
filled with aluminum oxide, resulting in flaw initiation. The remaining flaws (13.6%) were due to the
mechanical damage. Although anodizing
provided corrosion protection, it also resulted in the majority of the fatigue
cracks.
All
but two of the holes contained double flaws, of which none were
through-the-thickness flaws. The
selected stress spectrum marked the fracture surfaces extremely well, making it
possible to determine the crack length
within each loading block. Hence, it
was possible to fractographically determine the equivalent initial
quality for each flaw examined.
Figure 3.2.8
presents the probability density of occurrence versus the equivalent initial
quality for the A-7D aircraft. It
should be noted that the A-7D equivalent initial quality was determined by
fractography alone, since it was possible to measure the crack length during
the application of the first block of loading.
Figure 3.2.8. Probability Density of Occurrence of A-7D Equivalent Initial
Quality [Rudd & Gray, 1978]
The probability density of occurrence (Figure
3.2.8) was used to determine the cumulative probability of occurrence for
the A-7D aircraft. Figure 3.2.9 presents the cumulative probability of
occurrence versus the equivalent initial quality for the A-7D and F-4 C/D
aircraft. Also presented in Figure 3.2.9 is the cumulative probability of
occurrence with 95% confidence for each aircraft. For example, Figure 3.2.9 indicates
that with 95% confidence, 99.9% of the A-7D flaws have an equivalent length
less than 0.007 in. This means that one
out of a thousand flaws have an equivalent length greater than 0.007 in.
Figure
3.2.9. Cumulative Probability of
Occurrence of A-7D Equivalent Initial Quality [Rudd & Gray, 1978]