paper 1

International Journal of Fatigue 23 (2001) S337–S347www.elsevier.com/locate/ijfatigue

Fatigue and corrosion in aircraft pressure cabin lap splices

R.J.H. Wanhill*, M.F.J. KoolloosNational Aerospace Laboratory NLR, P.O. Box 90502, NL-1009 BM Amsterdam, The Netherlands

Abstract

Aircraft structures are susceptible to fatigue and corrosion damage, notably at joints. There may be interactions between fatigueand corrosion, especially as aircraft age. Longitudinal lap splices from several types of transport aircraft pressure cabins weredisassembled and investigated for Multiple Site Damage (MSD) fatigue cracking and corrosion. The results are compared withNASA data for a full-scale fatigue test. Broadly speaking, pressure cabin lap splices have either a fatigue problem or a corrosionproblem: MSD fatigue cracking is not initiated by corrosion. The results are discussed with respect to lap splice MSD fatiguemodelling, testing and analysis. 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Fatigue; Corrosion; Aircraft pressure cabins

1. Introduction

NASA [1], Fokker [2] and the NLR have obtainedfractographic data on Multiple Site Damage (MSD)fatigue cracking in longitudinal lap splices of transportaircraft pressure cabins. These data are supplemented byIAR [3] and NLR examination of severely corroded lapsplices. The lap splices came from five service aircraft,namely three Fokker F28s, a British Aerospace BAC 1-11 and a Boeing 727-100, and full-scale tests on aFokker 100 and the forward fuselage of a Boeing 747-400 [4]. The histories of all seven pressure cabins aresummarised in Table 1.

This paper covers the results so far, considering (1)MSD fatigue crack initiation and early crack growth, (2)corrosion and any associations between corrosion andMSD, and (3) the consequences for lap splice MSDfatigue modelling and analysis.

2. Lap splice positions and configurations

Figs. 1–3 show the positions of the investigated lapsplices, specifying the predominant damage, MSDfatigue cracks or corrosion. Fig. 4 shows the configur-

* Corresponding author. Tel.:+31-527-248-294; fax:+31-527-248-210.

E-mail address: [email protected] (R.J.H. Wanhill).

0142-1123/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved.PII: S0142-1123 (01)00147-5

Table 1Service or test histories of the pressure cabins

Aircraft type Flights Flight hours Simulatedflights

F28-4000 (1) 43,870 28,694(2) 43,323 unknown

F100 126,250a

BAC 1-11 75,158 52,000B 747-400 60,000b

B 727-100 53,424 59,848F28-1000 34,470 unknown

a At 110% design load.b At 100% design load.

ations of the lap splices investigated by the NLR. Thepositions and configurations differed widely, and the B747-400 lap splice was different again: a four rivet rowdesign with a Z-stiffener along the lower inner row [1].The single common feature of all the lap splices was thesheet material, aluminium alloy 2024-T3 Alclad.

3. Fatigue

Fatigue assessment was to determine the MSDlocations, any environmental (corrosion) influences onfatigue crack initiation and growth, the crack initiationlives and early crack growth rates. As expected [5], themost MSD-susceptible rivet row was the upper one in

S338 R.J.H. Wanhill, M.F.J. Koolloos / International Journal of Fatigue 23 (2001) S337–S347

Fig. 1. Positions of the MSD fatigue cracked lap splices investigatedby the NLR.

the outer sheet of each lap splice. However, other rivetrows were also susceptible, notably the lower one in theinner sheet, see Fig. 1 and Piascik and Willard [1].

Fig. 2. Positions of the MSD fatigue cracks in the B 747-400 lap splice investigated by NASA [1].

Fig. 3. Positions of the corroded lap splices.

3.1. MSD fatigue initiation characteristics

Fig. 5 shows the characteristic MSD fatigue cracklocations and shapes. These aspects are discussed next,followed by a summary:

1. F28-4000 service aircraft: The total lengths of MSDwere 530 and 330 mm for aircraft (1) and (2), respect-ively. The cracks initiated at many sites along the fay-ing surface edges of the outer sheet dimpling cones,see Figs. 4 and 5. The cracks were mechanically

S339R.J.H. Wanhill, M.F.J. Koolloos / International Journal of Fatigue 23 (2001) S337–S347

Fig. 4. Configurations of the lap splices investigated by the NLR, looking FORWARD.

induced, with no evidence that corrosion aidedinitiation. The large number of initiation sites sug-gests that fatigue cracking began soon after the air-craft entered service.

2. BAC 1-11 service aircraft: MSD was along about500 mm of port and starboard lap splices. The cracksinitiated at a variety of locations, see Fig. 5, thoughmainly at faying surfaces close to the rivet holes(types A, C and D). There was no evidence that cor-rosion aided initiation.

3. F100 full-scale test (indoors): MSD extended overseveral frame bays having poor adhesive bonds. Thecracks initiated from the faying surfaces, mostly atmultiple sites close to the rivet holes, see Fig. 5. Therewas no evidence of corrosion, as expected from atest indoors.

4. B 747-400 full-scale test (outdoors): MSD extendedover several frame bays, see Fig. 2. The cracksinitiated at a variety of locations, see Fig. 5, thoughmainly at faying surfaces (types A, A�) and rivethole/faying surface corners (types C, D) [1]. Therewas no evidence of corrosion, even though the testwas outdoors.

In summary, bearing in mind that the F28 dimpled lap

splices are uncustomary, the in-service and full-scale testMSD fatigue initiation occurred mostly from faying sur-faces near or at the rivet hole corners. This means thatthe faying surface condition (cladding, anodising, prim-ing, interfay sealant, adhesive bonding) and rivet holecorner quality are important. Also, fretting must play arole, if not always during crack initiation, then probablyduring early crack growth [6].

3.2. Secondary fatigue initiation characteristics: B727-100 aircraft

Secondary fatigue cracks in the B 727-100 lap spliceinitiated usually from intergranular corrosion cracking(exfoliation corrosion and stress corrosion) which hadalready spread away from the rivet positions. However,12 small fatigue cracks were found at six rivet hole pos-itions, without any evidence of fatigue initiation due tolocal corrosion (pitting), for example Fig. 6. Fig. 7 showsthe locations and shapes of these fatigue cracks at therivet holes. Ten of the cracks occurred inside the rivetholes in the inner sheet. This behaviour is distinct fromMSD, where the majority of cracks initiated from fayingsurfaces, and is discussed later in the paper.


Fig. 5. MSD fatigue crack locations and shapes.

Fig. 6. Intergranular corrosion cracking (exfoliation and stresscorrosion) with a secondary fatigue crack initiating from a rivet hole:B 727-100 lap splice outer sheet. The arrow points to the fatigue origin.

3.3. Early fatigue crack growth: environmental effects

Unlike fatigue initiation, there was evidence of (local)environments affecting the service aircraft fatigue frac-ture surfaces, either by post-cracking corrosion or duringcrack growth. The MSD and secondary fatigue fracture

Fig. 7. Secondary fatigue cracks at or near rivet holes for the B 727-100 lap splice.

surfaces from the F28-4000, BAC 1-11 and B 727-100service aircraft showed some post-cracking corrosion,making it difficult to measure fatigue striation spacingsnear the crack initiation sites. However, this post-crack-ing corrosion was generally mild, since otherwise thefatigue fracture surfaces would have been obliterated [7].

Of more interest are environmental effects during


Fig. 8. Multiple site MSD fatigue initiation and early crack growth in an F28-4000 lap splice.

fatigue crack growth. Evidence for such effects wasobserved:

1. Fig. 8 shows a particularly clear example of fracturesurface ‘beach marks’ . These sometimes occurredduring early MSD crack growth in the F28-4000 andBAC 1-11 lap splices, and were usually present onthe secondary fatigue fracture surfaces from the B727-100 lap splice. The beach marks may indicateperiodic changes in the local environment [8,9]. Atleast for the F28-4000 aircraft, it is certain that thebeach marks were not due to variable amplitude load-ing, since the pressure cabins of these aircraft experi-enced the maximum pressure differential in eachflight.

2. Fig. 9 shows a secondary fatigue fracture surface fromthe B 727-100 lap splice, with special features(arrowed) similar to those found — albeit more abun-dantly — for corrosion fatigue at low stress inten-sities [7].

Fig. 9. Secondary fatigue crack early growth in the B 727-100 lapsplice. Arrows point to narrow elliptical and semi-elliptical featureswith major axes in the local crack growth direction.

Presently the significance of these observations, otherthan being diagnostic for environmental fatigue crackgrowth, is limited. Fatigue striation spacing measure-ments for the light and dark bands comprising the beachmarks have not shown systematic differences in crackgrowth rates. This could mean that modelling and pre-diction of early MSD crack growth need not account forenvironmental effects beyond that of normal air, butmore work needs to be done, notably tests at low cyclefrequencies of the same order as in-service cabin press-ure cycling, and at overall crack growth rates similar tothose for early MSD crack growth in the lap splices.

3.4. Early MSD crack growth rates

1. Transverse (through-thickness): observation of trans-verse fatigue crack growth near the rivet holes wasoften hampered by fretting products and, for serviceaircraft, fracture surface corrosion. Fig. 10 summar-ises the data obtainable from fatigue striationmeasurements on fracture surfaces from the F28-4000and BAC 1-11 service aircraft and the F100 full-scaletest. Also shown is an estimate for the F100 test at100% design load. This estimate is derived from thefactor (100/110)7.1, where 7.1 is the ‘Paris Law’exponent in a piecewise linear fit to log da/dN versuslog �K data for 2024-T3 in the same range of crackgrowth rates [10]. Fig. 10 shows two noteworthy fea-tures: (a) through-thickness crack growth rates weremore or less constant for each aircraft type, and (b)the data and estimate indicate crack growth rates allabove 10�8 m/cycle.

2. Longitudinal: Fig. 11 summarises the data obtainedfrom fatigue striation and marker band measurementsfor the BAC 1-11, F100 and B 747-400 aircraft,together with an estimate for the F100 test at 100%design load, as discussed above. The BAC 1-11 andF100 data are presented more fully in [2,10], and theBAC 1-11 lap splices are still being investigated.

As before, the data and estimate indicate crack growth


Fig. 10. Summary of transverse (through-thickness) fatigue crack growth rates.

Fig. 11. Summary of longitudinal fatigue crack growth rates: B 747-400 data from NASA [1].

rates above 10�8 m/cycle. Also, the crack growth ratesfor different aircraft types were fairly similar in the rangeof crack sizes 0.5–2 mm. This is remarkable in view ofthe differing lap splice configurations, see Fig. 4 and [1].

These features of early MSD crack growth will bereturned to in the section on MSD fatigue modellingand testing.

4. Corrosion

The lap splice samples from severely corroded areasof the B 727-100 and F28-1000 service aircraft wereexamined for any cracking along the rivet rows and ator near the rivet holes.

4.1. B 727-100

The lap splice sample was provided by the IAR afternon-destructive examination using D sight, eddy currentand X-ray techniques [3]. Corrosion was visibly concen-trated internally along the upper side of the lapsplice/stiffener connection (R.H. side in Fig. 4), but Dsight showed corrosion-induced ‘pillowing’ to be presentthroughout the lap splice: corrosion-induced pillowing isbulging of the lap splice between the rivet positions, andis caused by local build-ups of corrosion productsbetween the outer and inner sheets.

Fig. 12 summarises the results of lap splice disas-sembly and examination by the NLR, combining theinformation from the outer and inner sheets. Almost 95%


Fig. 12. Schematic of the results of the B 727-100 lap splice disassembly: •=intergranular cracks; ↑=fatigue cracks at rivet holes; BS=Body Station.The numbers at the ends of the rivet rows refer to the rivet positions.

of the cracking, in terms of crack locations, was inter-granular owing to exfoliation corrosion and stress cor-rosion, the latter especially as corrosion built up betweenthe sheets to cause locally high stresses [11]. Some of thelarger intergranular cracks were associated with small,secondary fatigue cracks. These fatigue cracks usuallyinitiated from intergranular cracking, but sometimesdirectly from rivet holes, as mentioned earlier in thepaper with reference to Figs. 6 and 7. Also, Fig. 12shows fatigue cracking at two rivet positions (96 and 99)that had not undergone intergranular cracking but wereclose to severe corrosion revealed by lap splice disas-sembly.

The secondary fatigue cracks at — and especiallyin — the rivet holes most probably initiated owing to acombination of normal in-service stresses and additionalstresses caused by corrosion product build-ups betweenthe outer and inner sheets.

4.2. F28-1000

The lap splice sample was a 165 cm long strip con-taining about 200 rivets. The sample had been removedfrom the position shown in Fig. 3, owing to excessiveexternally visible corrosion. This had previously led tothe lap splice outer sheet being ‘cleaned up’ to a depthof about 0.25 mm, reducing the sheet thickness from1.2 mm, Fig. 4, to less than 1 mm.

Disassembly and detailed examination, includingopening-up selected outer sheet rivet holes by tensiletesting, showed no evidence of any pre-existing cracks.This result was unexpected, since the position of thesample was known to be MSD-susceptible. In fact, a450 mm long MSD fatigue crack was found in the sameaircraft at the same location on the left-hand side of thepressure cabin.

From these results we may draw two conclusions:

1. Excessive external corrosion in an MSD-susceptiblearea need not be associated with fatigue cracking,

despite stress increases in the lap splice outer sheetowing to cleaning up and thinning it.

2. Even in an MSD-susceptible area it is possible toaccumulate many flights (34,470, see Table 1) withoutfatigue crack initiation.

5. MSD fatigue modelling and testing

5.1. General remarks

Fatigue of pressure cabin longitudinal lap splices isdetermined by locally complex stress distributions whichare very difficult to determine [5,10,12]. In particular,this means realistic stress intensity factor solutions areunavailable for cracks smaller than the sheet thicknesses;and one may doubt the usefulness, through lack of accu-racy, of solutions for cracks with dimensions less thanthe rivet hole diameters.

Nor does it seem possible to model fatigue initiationby stress analysis. This is not only because of the com-plex stress fields, but also because MSD fatigue initiationin lap splices depends strongly on the local behaviourof various materials (aluminium alloy matrix, claddingand anodised layers, primers and sealants) at the fayingsurfaces and near or at rivet hole corners [10,13]. Notethat models based on fatigue initiation at corrosion pits[14] or inclusions in the aluminium alloy matrix [15]are most probably inappropriate for 2024-T3 Alclad lapsplices: the F28-4000 and BAC 1-11 service aircraftsamples showed no evidence of corrosion pitting, andtests have shown fatigue cracks initiate in the cladding,not the 2024-T3 matrix [12,16,17].

Instead — at least for now — recourse must be madeto empirical modelling that describes the actual earlyMSD fatigue behaviour of lap splices from service air-craft and full-scale tests. One such model, by Eijkhout[2], is described next.


5.2. Empirical model of early MSD fatigue behaviour

Eijkhout’s model [2] is based on the following obser-vations:

1. MSD fatigue cracks tend to initiate at several sitesnear or at rivet hole corners, and grow in directionsvarying gradually from transverse (through-thickness)to longitudinal, see the F100 sketch in Fig. 5.

2. The transverse crack growth rates are nearly constant,e.g. Fig. 10.

3. The transverse and longitudinal crack growth rates aresimilar for cracks smaller than twice the sheet thick-ness, compare Fig. 10 with Fig. 11 for crack lengthsup to about 2.5 mm (sheet thicknesses 1.2–1.25 mm,see Fig. 4).

Fig. 13 is a schematic of the model. There are threemain assumptions, the first two being derived from theforegoing observations: (a) constant crack growth rate inthe transverse direction, generally equal to the initialcrack growth rate in the longitudinal direction, i.e.dc/dN=AeBai; (b) crack depth c=0 at ai, the ‘ initiationlength’ ; and (c) quarter-circular crack fronts in the tran-sition from transverse to longitudinal crack growth.These assumptions are convenient but not essential. Forexample, the model can be used for non-constant dc/dNand for crack initiation at rivet hole corners, i.e. ai=0in Fig. 13. Also the model can be used for both non-countersunk and countersunk lap splice sheets [2].

Fig. 13. Eijkhout’s model illustrated for a non-countersunk sheet and multiple fatigue initiation sites.

Examples of the model’s use are given in Eijkhout’sreport [2]. The model provides estimates of the fatiguecrack growth lives and hence the fatigue ‘ initiation’lives, Ni in Fig. 13, whereby Ni is taken to be the fatiguelife beyond which there is a regular process of crackgrowth.

Two further aspects are of interest. Firstly, the modelenables estimates of the lives at which MSD fatiguecracks become through-thickness, which information ispotentially useful for in-service non-destructive inspec-tion. Secondly, the model can be made compatible withmarker bands on the fatigue fracture surfaces from full-scale tests, as in the case of the B 747-400 [1]: whendetermining the striation-based crack growth equations,da/dN=AeBa, for individual cracks, one can check theequations’ compatibility with the distances betweenmarker bands, adjusting the equations as necessary.

5.3. MSD fatigue initiation lives

Table 2 gives estimates of the MSD fatigue initiationlives (lives to first crack initiation) for the F28-4000,F100, BAC 1-11 and B 747-400 aircraft types. The F28-4000 estimate is only a reasonable guess. The F100 andBAC 1-11 estimates were obtained with Eijkhout’smodel, described above. The B 747-400 estimates weremade from marker band analyses [1].

There is considerable variation in the estimates,though the F28-4000 lap splices are not of general inter-est, and one would expect the shorter fatigue lives for


Table 2Estimates of MSD fatigue initiation lives in fuselage longitudinal lapsplices

Aircraft type MSD rivet row Lives to first crack initiation

Flights Simulatedflights

F28-4000 Outer sheet upper row A fewthousand?

F100 Outer sheet upper row 60,000a

F100 Inner sheet lower row 70,000a

BAC 1-11 Outer sheet upper row 60,000BAC 1-11 Inner sheet lower row 50,000B 747-400 Outer sheet upper row 5000–15,000b

a At 110% design load.b At 100% design load.

B 747-type aircraft, which were designed for a servicelife of 20,000 flights [18]. Actually, the total variationin initiation lives for the F100 and BAC 1-11 aircraft isgreater: for the F100 full-scale test the estimates rangedfrom 60,000 to 97,000 simulated flights, and for theBAC 1-11 the estimates ranged from 50,000 to 74,000flights.

5.4. MSD fatigue crack growth behaviour and itssignificance for testing

From the information provided by Figs. 10 and 11 itis apparent that the transverse and longitudinal fatiguecrack growth rates were above 10�8 m/cycle for cracksizes ranging from 30 µm to 5 mm. For 2024-T3 thismeans only long crack growth behaviour should be oper-ative [10,19], which facilitates crack growth modellingand agrees with Eijkhout’s approach [2].

On the other hand, these relatively high early crackgrowth rates in actual pressure cabin longitudinal lapsplices make questionable the usefulness and relevanceof sub-scale specimen tests. Uniaxial specimens ‘simu-lating’ the F100 pressure cabin longitudinal lap spliceshad early crack growth rates far too low compared to thefull-scale test results [20]. This was also true for biaxialspecimens [21]. This situation may change whenimproved stress analyses become available and are usedfor improving sub-scale specimen design and testing.However, it may turn out that reliance will still have tobe made mainly on full-scale tests [22,23].

6. Lap splice fatigue analyses

6.1. Analysis methods

Fig. 14 shows the ‘ traditional’ and developing fatigueanalysis methods for transport aircraft pressure cabin lapsplices. In the ‘ traditional’ methods, the inspection thres-

Fig. 14. Fatigue analyses for transport aircraft pressure cabin lapsplices.

hold is established using fatigue life S–N data, cumulat-ive linear damage analysis (if deemed necessary) andscatter factors. Subsequent inspection intervals are basedon a safe fatigue crack growth period using da/dN versus�Keff long crack growth data, crack growth models forspectrum loading (if deemed necessary) and scatter fac-tors.

In the developing methods, the inspection threshold isintended to be established using fatigue crack growthanalysis and tests, whereby it is assumed that the struc-ture contains initial flaws (Initial Quality Flaw Sizes,IQFS). Subsequent inspection intervals are based on asafe fatigue crack growth period that accounts for MSDessentially as a refinement — however important — tothe ‘ traditional’ analyses.

6.2. Analysis problems: actual versus predictedbehaviour

Fig. 15 is a schematic of the probable differencesbetween actual early MSD crack growth behaviour andpredictions using macroscopic (long) crack growth mod-elling, whereby the IQFS values are obtained either fromactual data for manufacturing flaws or by back-extrapol-ation using a long crack growth model.

For both types of prediction the long crack growthmodel is used to make an empirical fit such that thefatigue life is represented as a regular crack growth pro-cess, beginning as soon as the aircraft enters service.


Fig. 15. Schematic differences between actual and predicted earlyMSD fatigue crack growth behaviour in transport aircraft pressurecabin longitudinal lap splices: the predictions are fitted to the macro-scopic observations and the IQFS values are obtained from actualmanufacturing flaws or by back-extrapolation using a macroscopic(long) crack growth model.

This premise is incorrect: crack initiation is a physicalreality, however uncertain its definition (as mentionedearlier, we take ‘ initiation’ to be the fatigue life beyondwhich there is a regular process of crack growth). Thefitted model therefore has limited transferability, and ingeneral should not be used for ‘blind’ predictions ofcrack growth in other structural areas with differing lapsplice geometries and — most importantly — differingfaying surface conditions. Nor should the fitted modelbe used for predicting crack growth at different (local)stress levels. This latter point is significant for tworeasons:

1. The stress-dependence of fatigue initiation life willprobably be very different to that of fatigue crackgrowth, e.g. [24].

2. Actual fatigue crack growth rates could be in a differ-ent Paris Law regime, i.e. the exponent m in therelation da/dN=C(�K)m could be different. Anexample of a surprisingly high but realistic exponent,m=7.1, was mentioned earlier, with reference toFig. 10.

There would seem to be no solution to the above prob-lems so long as it is assumed that crack growth beginsas soon as the aircraft enters service. One alternative isto further investigate the usefulness of Eijkhout’s model.This possibility has much to recommend it. The modelis based on physical reality: it takes account of actuallap splice MSD fatigue initiation and crack growthbehaviour, and the present paper has shown much com-monality in this behaviour for several aircraft types anddifferent positions of the lap splices. Also, as mentionedearlier, it may turn out that MSD fatigue modelling willhave to rely mainly on full-scale fatigue testing. If so,then Eijkhout’s model provides a way of describingcrack growth, notably the most important early crackgrowth through the sheet thickness, and a way of esti-

mating the fatigue initiation life. Of course, since themodel is empirical, the parameters in the model haveto be determined for each type of aircraft, and also —possibly — for fuselage areas where the design stresslevels are significantly different, e.g. varying by morethan 10% from the average.

7. Conclusions

The characteristics of MSD fatigue and corrosion intransport aircraft pressure cabin longitudinal lap spliceshave been investigated using samples from five serviceaircraft, namely three Fokker F28s, a British AerospaceBAC 1-11 and a Boeing 727-100, and full-scale test datafor the Fokker 100 and the forward fuselage of a Boeing747-400 [1]. The following conclusions are drawn:

1. There is no primary association between MSD fatigueand corrosion for 2024-T3 Alclad lap splices. How-ever, there was evidence of local environmentaleffects during early MSD fatigue crack growth in lapsplices from service aircraft. It is uncertain whetherthe fatigue crack growth rates in service are signifi-cantly different, owing to environmental effects, fromthose determined by testing in laboratory air.

2. There was a strong tendency for MSD fatigue cracksto initiate at faying surfaces near or at rivet hole cor-ners. However, for the severely corroded B 727-100lap splice secondary fatigue cracks initiated usuallyfrom intergranular cracks due to corrosion and stresscorrosion, but sometimes occurred directly from rivetholes, albeit without any evidence of fatigue initiationdue to local corrosion.

3. A distinction should be made between MSD fatigueinitiation and fatigue crack growth: it is physicallyincorrect to consider lap splice fatigue solely as a reg-ular crack growth process that begins as soon as theaircraft enters service.

4. There are indications of considerable variation inMSD fatigue initiation lives, both in the range of livesfor each aircraft type and between aircraft types.

5. The most MSD-susceptible rivet row was the upperone in the outer sheets of the lap splices. However,other rivet rows were susceptible, notably the lowerone in the inner sheets.

6. Early MSD fatigue crack growth rates, for crack sizes30 µm–5 mm, were above 10�8 m/cycle (or flight),which means one should not expect any differencebetween short and long fatigue crack behaviour in thelap splices. However, the relatively high crack growthrates make the usefulness and relevance of sub-scalespecimen tests questionable.

The foregoing conclusions should be taken into


account during further development of fatigue analysismethods for transport aircraft pressure cabin lap splices.

References

[1] Piascik RS, Willard SA. The characteristics of fatigue damage inthe fuselage riveted lap splice joint. National Aeronautics andSpace Administration Langley Research Center, Hampton, VA:NASA Technical Publication NASA/TP-97-206257, 1997.

[2] Eijkhout MT. Fractographic analysis of longitudinal fuselage lap-joint at stringer 42 of Fokker 100 full scale test article TA15 after126250 simulated flights. In: Fokker Report RT2160. Amster-dam: Fokker Aircraft Ltd, 1994, (November).

[3] Chapman CE, Fahr A. Nondestructive examination of aircraft lapjoint specimens for the Netherlands National Aerospace Labora-tory, Memorandum LM-ST-787, Institute for AerospaceResearch, National Research Council Canada, Ottawa, February1997.

[4] Gopinath KV. Structural airworthiness of aging Boeing jet trans-ports — 747 fuselage fatigue test program. In: 1992 AerospaceDesign Conference, AIAA 92-1128. Washington, DC: AmericanInstitute of Aeronautics and Astronautics, 1992, (February).

[5] Eastaugh GF, Simpson DL, Straznicky PV, Wakeman RB. A spe-cial uniaxial coupon test specimen for the simulation of multiplesite fatigue crack growth and link-up in fuselage skin splices.In: Widespread Fatigue Damage in Military Aircraft, AGARDConference Proceedings 568. Neuilly-sur-Seine: Advisory Groupfor Aerospace Research and Development, 1995, (pp. 2-1–2-19).

[6] Rooke DP. Fracture mechanics analysis of short cracks at loadedholes. In: Behaviour of Short Cracks in Airframe Components,AGARD Conference Proceedings No. 328. Neuilly-sur-Seine:Advisory Group for Aerospace Research and Development, 1983,(pp. 8-1–8-6).

[7] Wanhill RJH, Schra L. Corrosion fatigue crack arrest in alu-minium alloys, Quantitative Methods in Fractography, ASTMSTP 1085, editors BM Strauss and SK Putatunda, AmericanSociety for Testing and Materials, pp. 144–165, Philadelphia,Pennsylvania (1990).

[8] Darvish M, Johansson S. Cyclic change in the humidity of theenvironment during fatigue crack propagation and its effect onfracture surface appearance. Scandinavian Journal of Metallurgy1992;21:68–77.

[9] Mussert KM. Formation of beach marks on Alclad 2024-T3sheet. MS Thesis, Department of Chemical Technology andMaterials Science, Delft University of Technology, Delft,December 1995.

[10] Wanhill RJH. Fractography of MSD in fuselage lap splices. BriteEuram Project No. BE95-1053, SMAAC: Structural Maintenanceof Ageing Aircraft, Document No. SMAAC-TR-1.3-02-1.3/NLR.NLR Technical Report TR 98310, National Aerospace Labora-tory, Amsterdam, July 1998.

[11] Komorowski JP, Bellinger NC, Gould RW. The role of corrosionpillowing in NDI and in the structural integrity of fuselage joints,ICAF 97, Fatigue in New and Ageing Aircraft, editors R Cook

and P Poole, Engineering Materials Advisory Services Ltd, Vol.1, pp. 251–266, Cradley Heath (1997).

[12] Muller RPG. An experimental and analytical investigation on thefatigue behaviour of fuselage riveted lap joints. Delft: Delft Uni-versity Press, 1995.

[13] Wanhill RJH. Effects of cladding and anodising on flight simul-ation fatigue of 2024-T3 and 7475-T761 aluminium alloys, NLRTechnical Report TR 85006 L, National Aerospace Laboratory,Amsterdam, January 1985.

[14] Wei RP, Harlow DG. Corrosion and corrosion fatigue of alu-minium alloys — an aging aircraft issue. In: Wu X-R, Wang Z-G, editors. Fatigue ’99, Proceedings of the Seventh InternationalFatigue Congress, vol. 4. Cradley Heath: Engineering MaterialsAdvisory Services; 1999. p. 2197–204.

[15] Laz PJ, Craig BA, Rohrbaugh SM, Hillberry BM. The develop-ment of a total fatigue life approach accounting for nucleationand propagation. In: Wu X-R, Wang Z-G, editors. Fatigue ’99,Proceedings of the Seventh International Fatigue Congress, vol.4. Cradley Heath: Engineering Materials Advisory Services;1999. p. 833–8.

[16] Forsyth PJE. The effect of cladding condition on the stages offatigue crack formation and growth. In: Branger J, Berger F, edi-tors. Problems with Fatigue in Aircraft, Proceedings of the EighthICAF Symposium. Emmen: Swiss Federal Aircraft Establishment(F+W), 1975, (pp. 2.5/1–2.5/23).

[17] Schijve J, Jacobs FA, Tromp PJ. The significance of cladding forfatigue of aluminium alloys in aircraft structures, NLR TechnicalReport TR 76065 U, National Aerospace Laboratory, Amster-dam, July 1976.

[18] Spencer MM. The Boeing 747 fatigue integrity program. In:Advanced Approaches to Fatigue Evaluation, Proceedings of theSixth ICAF Symposium, NASA SP-309. Washington, DC:National Aeronautics and Space Administration, 1972:127–78.

[19] Newman JC Jr., Edwards PR. Short-crack growth behaviour inan aluminium alloy — an AGARD Cooperative Test Programme.In: AGARD Report No. 732. Neuilly-sur-Seine: Advisory Groupfor Aerospace Research and Development, 1988.

[20] Schra L, Ottens HH, Vlieger H. Fatigue crack growth in simu-lated Fokker 100 lap joints under MSD and SSD conditions, NLRContract Report CR 95279 C, National Aerospace Laboratory,Amsterdam, June 1995.

[21] Vlieger H, Ottens HH. Results of uniaxial and biaxial tests onriveted fuselage lap joints specimens, NLR Contract Report CR97319 L, National Aerospace Laboratory, Amsterdam, April1997.

[22] De Jong GJ, Elbertsen GA, Hersbach HJC, Van der Hoeven W.Development of a full-scale fuselage panel test methodology,NLR Contract Report CR 95361 C, National Aerospace Labora-tory, Amsterdam, May 1995.

[23] Vercammen RWA, Ottens HH. Full-scale fuselage panel tests,NLR Technical Publication TP 98148, National Aerospace Lab-oratory, Amsterdam, March 1998.

[24] Schijve J. Fatigue life until small cracks in aircraft structures.Durability and damage tolerance. In: Harris CE, editor.FAA/NASA International Symposium on Advanced StructuralIntegrity Methods for Airframe Durability and Damage Toler-ance, 1994:665–81, (NASA Conference Publication 3274, Part 2,Hampton, VA).

paper 1

Documents