GLARE teardowns from the MegaLiner Barrel (MLB) fatigue test R.J.H. - - PowerPoint PPT Presentation

Nationaal Lucht- en Ruimtevaartlaboratorium – National Aerospace Laboratory NLR

GLARE teardowns from the MegaLiner Barrel (MLB) fatigue test

R.J.H. Wanhill, D.J. Platenkamp, T. Hattenberg A.F. Bosch and P.H. de Haan ICAF 2009, Rotterdam, the Netherlands, May 2009

ICAF, Rotterdam, the Netherlands, May 2009 2

GLARE: Generic illustration of GLAss REinforced aluminium laminate

ICAF, Rotterdam, the Netherlands, May 2009 3

GLARE: commercially available examples

All based on aluminium alloy 2024-T3 and high strength S2 silica glass fibres

Type GLARE 2A GLARE 2B GLARE 3 GLARE 4A GLARE 4B GLARE 5 *RD = Rolling Direction for aluminium layers Prepreg lay-up between two Al layers 0° / 0° 90° / 90° 0° / 90° 0° / 90° / 0° 90° / 0° / 90° 0° / 90° / 90° / 0° Prepreg layer thickness 0.25 mm 0.25 mm 0.25 mm 0.375 mm 0.375 mm 0.5 mm Typical applications Unidirectionally loaded parts with RD-Al* in loading direction (stiffeners) Unidirectionally loaded parts with RD-Al* perpendicular to loading direction (butt straps) Bi-axially loaded parts with ratio 1:1 of principal stresses (fuselage skins, bulkheads) Bi-axially loaded parts with ratio 2:1 of principal stresses with RD-Al* in main loading direction (fuselage skins) Bi-axially loaded parts with ratio 2:1 of principal stresses with RD-Al* perpendicular to main loading direction (fuselage skins) Impact-critical areas (floors and cargo liners)

ICAF, Rotterdam, the Netherlands, May 2009 4

GLARE: standard notation examples

GLARE 2B – 7/6 – 0.4 each aluminium layer 0.4 mm thick numbers of aluminium (7) and interleaved prepreg (6) layers type of GLARE, see slide 3 (butt straps) GLARE 4A – 5/4 – 0.4 each aluminium layer 0.4 mm thick numbers of aluminium (5) and interleaved prepreg (4) layers type of GLARE, see slide 3 (fuselage skins)

ICAF, Rotterdam, the Netherlands, May 2009 5

General objectives of GLARE teardowns from MLB fatigue test

Verification of teardown capabilities Crack locations and shapes Fatigue “initiation” and crack growth behaviour Provide data to check GLARE crack growth models

ICAF, Rotterdam, the Netherlands, May 2009 6

Airbus MLB fatigue test article

45,402 simulated flights with

conservatively high fatigue loads

ICAF, Rotterdam, the Netherlands, May 2009 7

“Opened out” MLB view with GLARE teardown locations

ICAF, Rotterdam, the Netherlands, May 2009 8

F4 GLARE window area

Skin + one window area doubler: GLARE 3–7/6–0.3/0.4

7 2024-T3 layers

ICAF, Rotterdam, the Netherlands, May 2009 9

F4 teardown procedure

NDI fastener holes

• Fastener removal; eddy current rotor inspection
• Window frame removal; eddy current pencil probe inspection
• f GLARE skin and Al window frame

Optical fractography

• NDI-indicated cracked holes opened up and mapped

Field Emission Gun Scanning Electron Microscope

(FEG-SEM) fractography

• Largest fatigue cracks in either window frame and GLARE skin

ICAF, Rotterdam, the Netherlands, May 2009 10

F4 largest fatigue crack locations (window C65-C66)

More crack indications in GLARE, see slide 11 Most cracks in fastener hole bores rather than countersinks: influence of secondary bending, see slide 12

• largest crack(s) 0.91–0.95 mm
• largest crack 2.3 mm ×

1.7 mm GLARE skin aluminium window frame

ICAF, Rotterdam, the Netherlands, May 2009 11

NDI verification and analysis

Al window frame 5 3 GLARE skin 19 22 Fastener hole fatigue cracks

• Area C64–C65
• Area C65–C66

Statistics for GLARE skin fatigue cracks

• 4 false calls for 216 inspected

holes

• Mean POD curve:
• for 90% probability

+50% confidence level the detectable crack length is 0.25 mm, i.e. most cracks in GLARE will be detected during complete teardown

ICAF, Rotterdam, the Netherlands, May 2009 12

F4 fatigue crack shapes and largest cracks

GLARE skin Al 7175-T73 window frame

0.5 mm

fatigue

• rigin

0.2 mm

fretting scar largest “readable” crack: 0.91 mm

ICAF, Rotterdam, the Netherlands, May 2009 13

F4 fractographic analysis (FEG-SEM)

“Readability” of MLB load spectrum (not derived to

provide crack growth markers)

• largest window frame crack

Fatigue “initiation” lives and crack growth behaviour

• largest window frame crack
• largest “readable” GLARE skin crack

Comparison of GLARE skin and window frame cracking

N.B.: Largest “readable” GLARE crack too small to check GLARE fatigue crack growth models: < 1 mm length, see slide 12

ICAF, Rotterdam, the Netherlands, May 2009 14

MLB load spectrum “readability” at F4 location: window frame crack

Repeated blocks of 2150 simulated flights 8 basic flight types Identifiable patterns due to peak loads in severest flights (A, B, C) Conclusion: good “readability”

C C B A C

ICAF, Rotterdam, the Netherlands, May 2009 15

Fractographic analysis of largest window frame crack

Fretting scar at fatigue origin

explains effectively zero “initiation” life and the indication of an initial crack size

Persistent retardation beginning

at a*=0.6 mm could be due to termination of the well-known “short crack effect”

ICAF, Rotterdam, the Netherlands, May 2009 16

Fractographic analysis of largest readable GLARE aluminium layer crack

Back-extrapolation of a versus N unfeasible da/dN “plateaux”: GLARE at about 50% of plateau for window frame crack

ICAF, Rotterdam, the Netherlands, May 2009 17

F4 GLARE window area summary

Complete teardown enabled NDI detection of 0.25 mm

cracks in GLARE with 90% probability and 50% confidence

Crack shapes and distributions in window frames and

GLARE skin indicate local secondary bending

Fractographic “readability” of MLB fatigue load spectrum

was excellent for largest window frame crack

• Fatigue “initiation” life ~ 0, owing to fretting between

fastener and hole

Fractographic “readability” was less for largest GLARE

crack, owing to much lower growth rates

ICAF, Rotterdam, the Netherlands, May 2009 18

F7 GLARE door beam area (detail)

Beam: GLARE 3–9/8–0.4 skin + seven doublers

34 2024-T3 layers

failure location of rectangular sample

ICAF, Rotterdam, the Netherlands, May 2009 19

F7 teardown procedure

NDI fastener holes (Airbus Deutschland)

• fastener removal; eddy current rotor inspection
• largest crack indication 7 mm, at fastener hole 33

Rectangular sample removed and pulled to failure

(see also slide 18)

• breaking load exceeded Limit Load requirement

Optical and FEG-SEM examination

• fractography of fastener hole 33 failure
• fastener hole bores 31–35

ICAF, Rotterdam, the Netherlands, May 2009 20

F7 fastener hole 33 optical fractography

Crack shapes show influence of secondary bending owing

to pressurization-induced bulging of door cut-out area

largest fatigue crack 6.55 mm

ICAF, Rotterdam, the Netherlands, May 2009 21

F7 fastener hole 33 fractographic analysis (FEG-SEM): I

Uniformly spaced fatigue

striations

Conclusion: predominance

• f pressurization loads at

F7 location: but see slide 22 for minor influence of severe simulated flights

• crack length 6.55 mm

2 μm

ICAF, Rotterdam, the Netherlands, May 2009 22

F7 fastener hole 33 fractographic analysis (FEG-SEM): II

Poor hole quality explains

effectively zero “initiation” life: corrected trend line compensates for minor influence of severe simulated flights Comparison of crack growth rates for the F4 and F7 locations. Note effectively constant da/dN for door beam crack. This is encouraging for GLARE crack growth models

ICAF, Rotterdam, the Netherlands, May 2009 23

F7 fasteners hole bores: fatigue and hole quality: hole 34

severe cracking

• glass fibre layers

starting to protrude and crack less severe cracking

• fatigue crack “initiation”
• ften at aluminium layer

corners

• ccasional heavy scoring
• n aluminium surfaces

leading to crack “initiation” away from corners

200 μm 500 μm 500 μm 200 μm

ICAF, Rotterdam, the Netherlands, May 2009 24

F7 GLARE door beam area summary

Breaking load of rectangular sample > Limit Load despite fatigue cracks at several fastener holes

Largest crack NDI-indication, 7 mm, close to actual crack

length, 6.55 mm

Fractographic “readability” excellent for largest crack

• Fatigue “initiation” life ~ 0, owing to poor hole quality
• Effectively constant fatigue crack growth rate

ICAF, Rotterdam, the Netherlands, May 2009 25

F6 GLARE stringer coupling area (schematic)

Stringer coupling: 3 thicknesses of GLARE 2A–2/1–0.3 2 2024-T3 layers

ICAF, Rotterdam, the Netherlands, May 2009 26

F6 teardown procedure: NDI only

Fastener removal and complete disassembly Eddy current rotor inspection of fastener holes for

skins, butt strap, stringers and stringer couplings

Eddy current pencil probe inspection of fastener hole

faying surfaces (not possible for aluminium skin/stringer faying surfaces because adhesively bonded as well as mechanically fastened)

ICAF, Rotterdam, the Netherlands, May 2009 27

F6 NDI-indicated crack classification

Many crack indications for GLARE components, but all < 4.5 mm Aluminium stringers had longer cracks

ICAF, Rotterdam, the Netherlands, May 2009 28

General conclusions

NDI teardown capabilities very good (F4 and F7 locations) Fatigue cracks characterized for F4 and F7 locations

• crack locations and shapes
• fatigue “initiation” and crack growth behaviour:

F7 door beam crack with effectively constant crack growth rate is encouraging for GLARE crack growth models

MLB GLARE components demonstrated high Damage

Tolerance –

component largest aluminium layer crack F4 skin < 1 mm F7 door beam < 7 mm F6 stringer coupling components < 4.5 mm – after more than 45,000 simulated flights with a conservatively high fatigue load spectrum

ICAF, Rotterdam, the Netherlands, May 2009 29