Advances in the Bonded Composite Repair o f Metallic Aircraft Structure phần 2 potx

Chapter 2. Materials selection and engineering 23

If the problem causing the need for the repair was fatigue or corrosion, it may

be more appropriate to use a composite for the repair as these materials are

effectively immune to these problems (composite repair layups generally have fibre

dominated properties which are immune to fatigue whereas layups with matrix

dominated properties may be susceptible to fatigue). The repair material chosen

can also be important where subsequent inspections are required and in many

cases the use of boron/epoxy composites is advantageous as eddy current methods

can be used to readily detect the crack underneath the repair. This is usually more

difficult if a metallic or graphite fibre patch is used due to the fact that these

materials are electrically conducting. Metallic materials will require the use of

stringent surface preparation and surface treatment processes to obtain a durable

bond, however, if a corrosion inhibiting primer is used, these processes could be

conducted elsewhere and the patch stored prior to use. Composite repairs using

thermosetting matrices such as epoxies are comparatively easier to prepare for

bonding, although the processes required are still important [2]. Thermoplastic

composites are in general harder to bond to than the more commonly used

thermoset composites. Finally metals lend themselves best to relatively flat repair

locations due to the difficulty in accurately forming a metallic sheet to a curved

profile. This is one of the strengths of composites where the desired shape can be

formed into the repair during cure.

Further considerations for the selection of a metallic material may include

corrosion and patch thickness. To avoid galvanic corrosion problems between

dissimilar metals, a sensible choice would be to use the original material for the

repair material as well. Where this is not possible, a check should be made to ensure

that different repair materials would not be susceptible to corrosion. For example,

repairs to a graphite/epoxy component will often be performed with a graphite/

epoxy material as well. Use of an aluminium material in this situation would be

unusual as the aluminium will readily corrode if in galvanic contact with the

graphite fibres. The adhesive should serve as an electrically insulating layer,

however, the more usual alternative to a graphite patch in this situation would be

titanium which will not corrode should the insulation break down.

In situations where the thickness of the repair is critical (on an aerodynamic

surface for example) consideration may be given to either steel or titanium to repair

aluminium. The greater stiffness of these materials should permit the design of a

thinner patch than would be possible with aluminium. Again consideration should

be given to possible galvanic coupling and potential corrosion problems in this

situation and it is possible that the choice of a composite may be preferable.

Laminated metallic materials have been developed in the Netherlands which

consist of layers of composite sandwiched between thin aluminium alloy sheets [3].

Where the composite used is kevlar (or aramid) the laminate is referred to as

ARALL (aramid reinforced aluminium laminate) and if the composite used is glass

fibre, the laminate is referred to as GLARE (Chapter 14). The fundamental idea

behind the development of these materials is to combine the traditional advantages

of both metals and composites. The composite component confers increased fatigue

strength and damage tolerance to the structure, while the aluminium allows the use

24 Advances in the bonded composite repair of metallic aircraft structure

of conventional metallic forming, fastening and manufacturing processes for

reduced cost.

GLARE has been proposed as a possible material for use in bonded repairs and

in particular has been used as a material for the repair of damage to the fuselages of

transport aircraft. The principal advantage of GLARE in this situation is the high

coefficient of thermal expansion. Work by Fredell et al. [4] and Chapter 14, has

shown that for repairs to thin fuselage skins which will mostly see pressurisation

loads at cruising altitudes (-55 "C), the higher coefficient of thermal expansion of

GLARE provides structural advantages compared with composite alternatives (see

Section 2.6 for further discussion). On the other hand the low specific stiffness of

GLARE results in a much thicker patch than for a high modulus composite

material, and this needs to be carefully considered in the design to ensure that

bending effects due to neutral axis offset are not excessive and that high stresses at

the ends of the patch are alleviated by tapering for example.

Finally, it may be possible to use nickel as a repair material in some specific

circumstances for example where geometry is complex. The repair of a crack in the

comer of a bulkhead pocket is a good example. Nickel can be electroformed to

replicate the surface of a mould with very high precision, and therefore it should be

possible to produce an electroformed nickel patch which will fit precisely into the

pocket. As mentioned above, the isotropic nature of the nickel would be an

advantage in this situation, although care needs to be made to ensure that the

electroforming process does not produce planes of weakness within the electro￾form. Work is underway to evaluate this method as a repair option for a damaged

army gun support structure [5]. In situations such as this where a certain degree of

rough handling can be expected, the hard, damage resistant surface of the nickel

provides another important advantage over a fibre composite repair.

2.2.2. Non-metallic materials

The two main non-metallic materials used are boron/epoxy and graphite/epoxy

composites. Glass fibre composites are not used due to their low stiffness and

kevlar composites while strong and stiff in tension have relatively poor

compression performance.

Boron fibres were first reported in 1959 and were the original high modulus fibre

before the development of graphite fibres in the 1960s. Boron composites were used

to produce aircraft components such as the skins of the horizontal stabilisers on the

F-14 and the horizontal and vertical stabilisers and rudders on the F-15. The use of

boron composites in large-scale aircraft manufacturing has largely stopped now

due to the development of more cost-effective graphite fibres. The production

process for boron fibres is time consuming and does not lend itself to mass

production in the same way as modem methods for producing graphite fibres. For

this reason the price of boron fibres has not dropped as significantly as that of

graphite fibres which are now at around I/lOth the cost. Boron fibres are

manufactured individually by chemically vapour depositing boron onto a heated

tungsten wire substrate from boron trichloride gas in a reactor. The fibres are

Chapter 2. Materials selection and engineering 25

available from Textron Speciality Materials in 100 and 140 micron diameters and

commercial pre-pregs are available with either 120°C or 175°C curing epoxies. The

fibre diameter is significantly larger than normal graphite fibres due to the presence

of the tungsten core. Attempts have been made in the past to use a carbon filament

precursor to reduce the production costs, however, these boron-carbon filaments

have generally not had the high level of strength that can be produced with the

tungsten filament precursor.

Boron fibre is an extremely hard material with a Knoop value of 3200 which is

harder than tungsten carbide and titanium nitride (1800 -1880) and second only to

diamond (7000). Cured boron composites can be cut, drilled and machined with

diamond tipped tools and the pre-pregs are readily cut with conventional steel

knives. In practice the knives cannot actually cut the hard fibres, however, gentle

pressure fractures the fibres with one or two passes. “Snap-off’ knife blades are

commonly used as the cutting edge is rapidly worn by the hard fibres. Although it is

possible to cut complex shapes with the use of templates, laser cutting has been

shown to be the most efficient way to cut a large amount of non-rectangular boron

plies. Circular patches, for example, are readily cut using a laser cutter with the pre￾preg supported on a backing material such as Masonite.

The combination of very high compressive stiffness, large fibre diameter and high

hardness means that boron fibres can readily penetrate skin and care must be

exercised in handling boron pre-preg to reduce the chance of splinter-type injuries.

If a fibre does enter the skin, it should be removed very carefully with he tweezers.

Trying to squeeze the fibre out must be avoided as the fibre may fracture into

smaller segments.

The stiffness and diameter of boron fibres also restricts their use in small radius

corners. The 100 micron diameter fibre can be formed into a radius of 30 mm, but

this is about the limit than can be comfortably achieved. The smaller diameter of

graphite fibres makes it the choice for smaller radii situations. In most other

aspects, boron pre-pregs handle and process in a similar fashion to the more

common graphite pre-preg materials.

As a repair material, boron/epoxy composites have a number of advantages [ 1,6]

including;

0 an intermediate coefficient of thermal expansion which helps to minimise the

level of thermally induced residual stress which results from an elevated

temperature cure. This contrasts with graphite fibres mentioned below.

0 relatively simple NDI is possible using eddy currents through the repair patch to

detect the extent of the defect. This is possible due to the non-conducting nature

of the fibres.

0 no galvanic corrosion problems when bonded to common airframe materials.

0 a good combination of high compressive and tensile strength and stiffness (the

compressive strength of a unidirectional B/EP composite is 2930 MPa compared

with 1020MPa for HMS GR/EP)

Graphite fibres are now available in a very wide range of properties and forms and

improvements in manufacturing processes has seen the cost of the fibres reduce

over the past 25 years. Although the fibres are not as hard as boron, the cured

26 Advances in the bonded composite repair of metallic aircraft structure

composites are very abrasive and diamond tipped tools are normally used for

cutting or machining. The fine graphite laden dust from such operations is believed

to be a health hazard and so measures to control this hazard must be taken. This

electrically conducting dust can also cause problems with electrical equipment if it

is not removed and filtered from the room air. Graphite pre-pregs are commonly

available as 120°C and 175°C curing systems and lower temperature cure resins are

also available now for use in repair situations.

Graphite fibre is an unusual material in that it has a slightly negative coefficient

of thermal expansion, which means that the fibres contract slightly in the axial

direction when heated. This results in relatively high levels of thermally induced

residual stress if the cured composite is bonded to the structure with an elevated

temperature curing adhesive. As well, the fibres are electrically conducting and will

cause galvanic corrosion of aluminium if the two are in electrical contact. Due to

the electrical conductivity it is more difficult to use eddy-current NDI methods with

these materials to check the position of a crack under the patch for example.

Graphite composites are significantly cheaper than boron composites and are

available from a very wide range of suppliers. They offer a wide range of properties

for design and with epoxy resin matrices are readily processed and can be cured to

complex shapes to suit the damaged structure. If a repair is required to a tight

comer with a small radius, graphite fibres would be preferred to boron as

mentioned above.

Repairs to aircraft are usually weight critical and so the specific properties of the

various repair materials are therefore of interest. Table 2.2 compares the mechanical

and thermal properties of some candidate patch or reinforcing materials. This

comparison includes boron/epoxy (b/ep) and graphite/epoxy (gr/ep), the metal/

composite laminates GLARE and ARALL and typical high-strength aluminium

and titanium alloys - which also represent the metals to be repaired.

2.2.3. Patch material selection

Many of the criteria for selection of a successful repair material have been

discussed in the above two sections. The reader is referred to Sections 2.1 and 2.2

for a complete discussion of the issues and in this section a summary of the main

points is given referring to the four main repair materials and some of the main

design issues that are commonly faced.

0 Patching efficiency: High tensile stiffness is required to minimise the crack

opening displacement after repair and therefore keep the stress intensity and

crack growth down. The fibre composite materials are naturally more efficient

than either the conventional or laminated metallic materials (refer Table 2.2 for

specific stiffness i.e. modulus divided by destiny).

0 Operating temperature: For sustained high temperature operation over 1 50"C, a

titanium patch may prove to be the best solution. Conventional aluminium

alloys and the laminated metals would need to be carefully investigated as there

are a range of upper temperature limits depending on the alloy and heat

treatment involved. In general, most aluminium alloys could withstand extended

Chapter 2. Materials selection and engineering 21

Table 2.2

Relevant materials mechanical and physical properties for component and patch materials.

Thermal

Shear Critical Fatigue expansion

Modulus modulus strain Strain Density coefficient

Material GPa GPa x 10-~ x 10-~ (g/cm3) oc x IOP

Aluminium alloy

Aluminium alloy

Titanium alloy 6

Boron/epoxy b/ep

(unidirectional)

Graphite/epoxy gr/

ep (unidirectional)

Aluminium laminate

GLARE 2

Aluminium laminate

ARALL 3

Electroformed

Nickel

1015 T6

2025 T3

A1/4V

12

12

110

208 max

20 min

148 max

12 min

65

68

201

21

21

41

1

5

na

na

16

6.5 3.3

4.5 3.3

8.8 6.8

7.3 1.0

13 12.0

5.2 3.3

8.9 3.3

1.7-3.4 na

2.8

2.8

4.5

2.0

1.6

2.5

2.3

-9

23

23

9

4.5 min

23 max

- 0.3 min

28 max

- 15

- 16

... 13

Notes: (a) Maximum modulus and minimum expansion coefficient are in the fibre direction, other values

are for the transverse direction, (b) shear modulus values for the composite are for through-thickness

deformation, (c) critical strain refers to failing strain for the composites and yield strain for the metals,

(d) fatigue strain refers to approximate strain for crack initiation at lo6 cycles, R - 0.

periods at 120°C, which is slightly higher than the normal operating temperature

of 105°C for a 175°C curing composite pre-preg. Higher temperature curing

resins are available for composites, although the availability is not as high and

depending on the system involved, processability may be reduced.

0 Residual stress: If a repair (cured at elevated temperature) is likely to see

extended service at low temperatures (for example a fuselage repair to a transport

aircraft - [4]), the best choice may be either a conventional or laminated metallic

material where the coefficient of thermal expansion is more nearly matched to

the structure. In this situation, graphite/epoxy repairs and to a lesser extent

boron/epoxy repairs will result in higher levels of thermally induced residual

stress [7].

0 Cost: Although not usually a major driver, conventional metallic materials

would offer the lowest material costs, followed by the laminated metals, graphite

composites and the boron fibre composites are the most expensive. Analysis of

repair costs need to be done carefully as often a composite repair may prove to

be cheaper than a metallic repair despite greater material costs. This is largely

due to the excellent formability of composites and the reduced time required to

form the repair patch to the desired shape.

0 Inspections: If full use is made of the benefits of bonded repair technology and

the defect is left in the structure under the repair, it is likely that future non-