Applied Mechanics and Materials Vols 3-4 (2005) pp 197-204 Online: 2006-08-15

© (2005) Trans Tech Publications, Switzerland

doi:10.4028/www.scientific.net/AMM.3-4.197

Spar Corner Radius Integrity for the A400M Wing

T. Edwards and J. Thompson

Atkins Aerospace and Defence Systems, 650 Aztec West, Almondsbury, Bristol BS32 4SD
England. tim.edwards@atkinsglobal.com; jeremy.thompson@atkinsglobal.com

Keywords: composite; inter-lamina; corner radius; opening moment; failure stress.

Abstract: The paper focuses on the structural integrity of the corner radius of the carbon fibre
composite, ‘C’-section spar for the Airbus A400M wing. The corner radius is subject to opening
moments generated by internal wing box fuel pressures. The low inter-lamina strength of
composites makes de-lamination of the corner of prime concern.

The paper describes initial development of analytical techniques to calculate the through-thickness
tensile stresses and inter-lamina shear stresses developed in a corner radius under applied bending
moments and transverse shear forces. A test programme is also described, aimed at the
determination of the failure moment of curved laminates under pure bending moments. Using the
analytical expressions developed, a through-thickness failure stress is calculated from the failure
moments. A variation of the failure stress with specimen thickness is indicated, showing that
thicker specimens fail at higher inter-lamina stresses – a characteristic that must be exploited in the
design of the spar.

Using finite element analysis of the test configuration, in conjunction with virtual crack extension
techniques, it is demonstrated that, at the failure load, a constant rate of strain energy release
accompanies inter-lamina crack growth in the different test specimens. A critical energy release
rate for uncontrolled crack growth is thus established, which is used, in conjunction with further
finite element analysis, to predict the failure stress of specimens with different values of thickness
and corner radius.

It is concluded that this fracture mechanics approach to integrity can be applied to the A400M spar
corner and to similar aircraft structures. Recommendations for further testing and correlation with
analysis are proposed to strengthen the theoretical basis for such integrity assessments.

Introduction
The Airbus A400M aircraft breaks new ground for its manufacturer in a number of areas, but
principally in the use of laminated carbon fibre reinforced plastic (CFRP) composite construction in
the main wing box. The use of laminated composites in preference to traditional aluminium alloys
enforces the adoption of radically different structural assessment techniques to address familiar
structural problems, and presents new difficulties not encountered with metals. One difficulty
arises from the poor through-thickness (inter-lamina) strength of laminated composites.

For thin-walled aircraft structures, stresses in a direction normal to the panels are not normally of
concern: stress levels in this direction are usually very low, and metallic materials have good
through-thickness strength. Laminated composites, by contrast, have a very modest strength
through the thickness, but, for most areas of the structure, stresses remain well below allowable
stress values, so integrity is not compromised. For panels with a high curvature, however, through-
thickness stresses can become quite large.

The A400M wing box is a fuel tank, and the sharp curvature of the spar corner radii attracts
significant through-thickness stress under wing box fuel pressure loads. Such stresses, negligible
for metallic designs, drive up both the laminate thickness and the radius of the corner in a laminated
composite design, resulting in increased weight for the spar. It is thus essential that both the

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans
Tech Publications, www.ttp.net. (ID: 128.226.37.5, State University of New York at Binghamton, Binghamton, USA-12/05/15,15:42:04)
198 Advances in Experimental Mechanics IV

stresses and the material strengths are calculated with precision so that weight efficiency of the spar
may be maximised.

Wing Box Structure and Loading

Figure 1 illustrates a typical cross section for the A400M wing box. The box comprises carbon
composite skins and spars (ribs are metallic), the spars being of ‘C’ section bolted to the skins. The
wing box is divided into a number of fuel tanks, the skins, spars and some ribs acting as fuel tank
boundaries. The skins and spars are stiffened by bonded composite stiffeners, which allow the fuel
tank boundaries to resist fuel pressures.

M CFRP Skins

M
CFRP Spar

Figure 1 – Wing Box Cross Section and the Effect of Internal Fuel Pressure

A critical load case for the A400M design is a refuel overflow case, during which a fault causes the
tanks to be over-filled, creating a large rise in the tank internal fuel pressure. The pressure load
induces out-of-plane bending in the spar webs and skin panels. Although in general the structure is
well able to resist such loads, the spar corner, which, as Figure 1 illustrates, is subject to an opening
moment, develops significant through-thickness tensile stresses between the plies within the
laminate.

Calculating the magnitude of the opening moment and the transverse shear force arising from the
fuel pressures can be achieved either through hand calculation or finite element analysis techniques.
The calculation of the inter-lamina tensile and shear stresses arising from these loads is more
complex than for a homogeneous material, but a simplified calculation approach has been
established.

Corner Radius Stresses

In order to assess the integrity of the spar corner when the wing box is subject to fuel pressure, it is
necessary to calculate the inter-lamina stresses. A moment generates tensile stresses, whilst a
transverse shear load induces shear stress between the plies. Lehknitskii [1] developed expressions
for the inter-lamina stresses arising in a homogeneous anisotropic material, but such an expression
is not directly applicable to inhomogeneous laminated composite materials. Ko and Jackson [2]
extended Lehknitskii’s approach to address laminated materials, although the complexity of this
exact, stress function solution is significant. To avoid such complexity, a method based on
Engineer’s Bending Theory was derived, which enforces the simplifying assumption of “plane
sections remaining plane”. Using this approach, in-plane strains in successive plies are calculated,
in-plane stresses in the plies are found using the ply stiffness, and finally, by applying radial
equilibrium to each ply, the resulting inter-lamina stress can be computed.

Figure 2 shows the through-thickness tensile stress predicted in a laminate with an applied opening
moment of 1500 N.mm/mm using this theory. The corner in this calculation has an internal radius
of 10mm, whilst the laminate has a thickness of 5mm with 20% 0o, 60% +/-45o and 20% 90o fibres
following the ply sequence: {45, -45, 0, 0, -45, 45, 90, 90, -45, 45}sym. Note that the 0o direction
lies parallel to the spar flange. Homogeneous theory predicts a peak stress of 36MPa for this
 Applied Mechanics and Materials Vols. 3-4 199

configuration, whereas the inhomogeneous approach predicts 44MPa. This example analysis is
typical of the analysis of any laminated composite - that the inter-lamina stresses arising are
dependent on ply sequence, and are rather different from those found in homogeneous materials. In

Homogeneous vs Inhomogeneous

50

Radial Stress (MPa)

40

30
Lekhnitskii
Inhomogeneous
20

10

0
0 0.2 0.4 0.6 0.8 1
Eta (y/t)

Figure 2 – Inter-Lamina Tensile Stresses, 20/60/20 Laminate

this example, the peak tensile stress is predicted to be 20% greater than would be predicted using
the homogeneous formulation. Therefore, when assessing the integrity of the A400M spar corners,
it is vital to employ the revised formulation in order to accurately predict the highest inter-lamina
stresses. Comparisons between the derived expressions and the exact values of Ref.[2] showed a
maximum over-estimate of inter-lamina stress of 2%, justifying the use of the simplified approach
whilst also providing a measure of conservatism appropriate in aerospace design.

From the geometric, loading and material strength data that were available, it was clear at an early
stage that transverse shear played little part in the integrity of the spar corner: failure is dominated
by the opening moment applied to the corner and the strength of the corner depends on the through-
thickness tensile strength of the material resin.

The Results of Strength Testing

Having developed a satisfactory means by which inter-lamina stresses could be calculated, a value
of the resin strength was required to allow an integrity assessment to be carried out. In devising a
test to measure this strength, it was decided that:

1. the test configuration should, as far as possible, represent the configuration for the spar corner
under pressure loading;
2. artificial damage should be introduced into the test specimens to simulate the type of defect that
might be present in a real aircraft structure.

The testing arrangement shown in Figure 3 was adopted. The test configuration conforms, broadly,
to ASTM D6415 [3], a small divergence from that standard being made to accommodate Airbus
standards on maximum defect size and minimum defect spacing. The angle specimens were cut to
52mm lengths from long angles manufactured both from Hexcel M21/T700 and from Cytec 977-
2/HTS (these materials being approximately equivalent in terms of stiffness and strength). The
specimens had simulated damage in the form of Teflon film inserted between plies during the
manufacturing process at locations of predicted maximum through-thickness stress. Table 1 shows
the test matrix, eight specimens being tested for each data point.
200 Advances in Experimental Mechanics IV

Loading
Direction
Defect in Corner

Loading Bars on
Test Specimen Roller Bearings

Figure 3 – Angle Bend Test Configuration

Thickness (t),
Lay-up Defect
Material Outer Radius, (R)
(%0/%45/%90) (mm)
(mm)
10/80/10 3x12
M21/T700 t=5, R=10
20/60/20 3x52
977-2/HTS t=10, R=25
30/40/30 Impact

Table 1 – Test Matrix for Angle Bend Specimens

Strain gauge instrumentation on the inner and outer laminate surfaces provided an indirect measure
of the moment generated in the angle. The gauges showed that, in the original test configuration,
excessive friction was apparent at the (fixed) loading bars, so these were changed to rollers in
subsequent tests.

The tests were carried out both at room temperature in the “as received” condition and hot (700C).
Completion of the testing yielded a failure load against each specimen. Failure moment at the
corner of the angle could be derived from the test geometry, correlated with strain gauge output.
The corresponding through-thickness stress at failure could then be calculated from this moment
using the theoretical approach previously described. In this way, a failure stress could be assigned
to each test. The results showed that the through-thickness strength for the 10mm thick specimens
(mean value 25MPa) was, apparently, greater than for the 5mm thick samples (mean value 18MPa).
A typical specimen, post-test, is shown in Figure 4.

A Fracture Mechanics Approach to Laminate Strength Assessment

The apparent increase in the through-thickness strength of the laminate resin with laminate
thickness had not been anticipated, but was welcome in view of the enhanced structural
performance that might be gained for thicker laminates. However, in order to capitalise on the
potential increase in strength, it was necessary to match the empirical results with a robust theory
that would explain the increase.
 Applied Mechanics and Materials Vols. 3-4 201

De-Lamination

Figure 4 – 10mm Thick Angle Post-Test, showing Through-Thickness De-lamination

The resin failure adjacent to the inter-lamina defect was recognised as a Mode Ι opening of a crack
in classical linear elastic fracture mechanics (transverse shear is zero in the corner radius), as
appreciated by Martin [4]. It was therefore postulated that, by correlation of the test results with
finite element modelling, a critical energy release rate could be determined, and hence a failure
stress for angles of other thickness could be predicted by matching this critical value. Therefore,
detailed finite element models of 5mm thick and 10mm thick specimens under the test conditions
were constructed, the 10mm thick model being shown in Figure 5.

This model, which simulated the in-homogeneity of the specimen, represented the defect by means
of disconnected nodes at the corner radius. The refinement of the model in the corner made it
possible to modify the length of the defect (a “virtual crack extension”), and facilities within the
NASTRAN finite element code allowed a prediction to be made of the total strain energy within the
specimen. The test rig itself was assumed to be rigid.

Figure 5 – Finite Element Model of Test for 10mm Thick Specimen

Using these models, the following plan was proposed:

• Analyse the model in the test configuration at the failure load and calculate the total strain
energy in the model;
202 Advances in Experimental Mechanics IV

• Increase the modelled length of the de-lamination by a short distance and re-analyse the model
under the same displacement of the rollers;
• Calculate the strain energy in the model under the revised configuration.

Through Thickness Strength vs Angle Thickness

30

Failure Stress (MPa) 25

20

15

10

0
2 3 4 5 6 7 8 9 10 11 12 13 14 15

Angle Thickness (m m )

Figure 6 – Variation of Through-Thickness Strength with Laminate Thickness

In this way, the critical rate of release of elastic strain energy at the failure load could be predicted
using the model. This approach is the equivalent of finding the resin fracture toughness.
From such analysis, it was shown that the mean energy release rate for the 5mm thick specimens at
failure (0.076 J/mm) matched that for the 10mm thick specimens (0.074 J/mm) closely. This
release rate was matched for models of other specimen thickness values (2mm and 15mm), and the
predicted failure stress recorded. Figure 6 shows the results of the extrapolation of through-
thickness strength, indicating a curve that, when corrected for B-basis (a value that at least 90% of
specimens will exceed with a 95% confidence), “hot/wet” properties for use in aerospace design,
can be used for corner analysis on the A400M spars.

Conclusion: Strength Assessment of the A400M Spars

The angle bend testing has shown that a variation in the through-thickness failure strength of
A400M spar corner radius with laminate thickness should be expected. This variation is consistent
with the principles of linear elastic fracture mechanics, and inter-lamina tensile strength for a given
thickness may be determined through the use of virtual crack extension techniques in conjunction
with the finite element method and test correlation. Further testing to explore the extremes of
thickness is recommended to avoid significant extrapolation of data.

References
[1] Lekhnitskii S.G., ‘Anisotropic Plates’, 1968.
[2] Ko W.L. & Jackson R.H., ‘Multilayer Theory for Delamination Analysis of a Composite Curved
Bar Subjected to End Forces and End Moments’, NASA Technical Memorandum 4139,
September 1989.
[3] ASTM Standard, ASTM D6415.
[4] Martin R.H., ‘De-Lamination Failure in a Uni-Directional Curved Composite Laminate’, NASA
Contractor Report 182018, April 1990.
Advances in Experimental Mechanics IV
10.4028/www.scientific.net/AMM.3-4

Spar Corner Radius Integrity for the A400M Wing

10.4028/www.scientific.net/AMM.3-4.197